LOS ANGELES -- A child is on the witness stand, testifying.

He is slight, this child, with dark combed hair and a narrow face so small that the microphone before it nearly hides him from the room. He sits straight in the chair, or twists it gently from side to side; when he hesitates, considering an answer, he cocks his head and looks in the direction of his mother, but his eyes register nothing except momentary concentration. "Yes," he says. "No." "I don't know." His last name has been deleted from the court record, in deference to what it is the child has come to say, and in the printed transcript he is A., for Answer. Q: Did Ray ever put a part of his body inside your mouth? A: Yes. Q: When you were drawing that picture, why did you draw it? A: To help me not be afraid. Q: Can you explain to us how you were feeling when you drew that picture? A: Sick, and dirty.

The courtroom is windowless and extremely quiet. In the back a reporter and an insurance man type notes into portable computers. The woman writing the book and screenplay balances a stenographer's pad on her lap. No television cameras will attend the testimony of Child Witness Number Five; Child Witness Number Five will not talk about cats being stabbed to death, as did Child Witness Number One, or about being tied up and photographed naked, as did Child Witness Number Three.

He will not testify, as did one child witness at a preliminary hearing, that he was driven to a circus house and molested after seeing someone in an elephant costume. He will not testify, as did a second child witness at the preliminary hearing, that he was driven to a cemetery and made to help dig out a coffin and watch a body being cut with knives. He will not testify, as did a third child at the preliminary hearing, that the McMartin Preschool in Manhattan Beach once featured a game called Lookout, in which small children climbed to the top of the jungle gym to watch for approaching parents while teachers molested the children inside.

It is February 1988, the eighth month of the criminal trial in the longest-running major child molestation case in American history, and here is what Child Witness Number Five has to say:

He says Raymond Buckey, the 30-year-old man who is one of the two remaining defendants, sexually molested him at the McMartin Preschool. He says he saw Buckey's mother Peggy, the other defendant, "and she had her bra on." He says he saw Raymond Buckey kill a horse by hitting it with a bat, and that when the horse died he saw blood come from its body, from "the top of its leg."

"How many times did he hit the horse before it died?" the defense attorney asks.

"I don't know," says Child Witness Number Five.

"Did the horse ever jump up and buck around when Ray hit it with the bat?" the defense attorney asks.

"I don't know," says Child Witness Number Five.

He is 11 years old. At breaks he climbs down from the witness chair and walks to his mother, arms limp at his sides. Sometimes he smiles when his mother puts her arm around him. He was barely 7 when Raymond Buckey was arrested, and over the last five years Child Witness Number Five has talked to therapists, police officers, prosecutors, investigators, defense attorneys, judges and the 23 Los Angeles County grand jury members who in March 1984 handed down the indictments that gave formal shape to the extraordinary proceedings known across the nation by now as the McMartin case.

In August 1983, when a local police investigator examined the first sexual abuse accusation against Raymond Buckey, no modern American precedent offered any premonition of the phenomenon that was about to beset Manhattan Beach. Everything about the McMartin case, from the day Manhattan Beach police officers sat in their headquarters building and wondered how to ask 200 children whether they had been touched by Mr. Ray, has defined a numbing kind of investigatory frontier: there was a single suspect, and then there were dozens, and then there were so many preschools under investigation and so many sexual abuse reports that in this small curve of beach towns at least 500 children, it was said, had told parents or therapists about the men and women who molested them.

There were mass meetings in church recreation halls, and telephone hot line volunteers who fielded calls from fathers wanting nothing more than someone to listen while they wept, and parents who drove their children along the streets of Manhattan Beach so the children could point and say, That house, they took me there. And there were conferences, after a while, in which the parents of Manhattan Beach children could travel to other states to talk to parents of children from West Point, N.Y., or Malden, Mass., or the other American communities in which large groups of small children had begun for the first time to describe both sexual molestation and another kind of abuse, one that FBI agent Kenneth Lanning, for lack of a more descriptive single label, calls "bizarre."

No single agency formally tracked these investigations around the country, but Lanning is a supervisory agent specializing in child sexual abuse cases; local investigators tend to contact him, he says, about cases involving numerous very young children, groups of adults, and reports that suggest the bizarre. By April of this year, Lanning had examined enough of these cases to have lost exact count -- "100 to 200," he says.

Before August 1983, he had never heard of them. Before August 1983 no mail issued forth from groups called Society's League Against Molestation, or Parents Against Child Abuse, or Affirming Children's Truth. Nobody before 1983 had taken out a full-page newspaper advertisement, as did a group calling itself the Friends of the McMartin Defendants, in which large black letters recalled the principal year of the Salem Witch Trials.

"SALEM, MASSACHUSETTS 1692," read the advertisement. "MANHATTAN BEACH, CALIFORNIA 1985." The advertisement appeared in an August 1985 edition of the Easy Reader, a local weekly; when it ran, the preliminary hearing in the McMartin case was in its 14th month and still had five more to go. It was the longest preliminary hearing in California history, and by the time it ended in 1986, a foray into the Manhattan Beach preschool investigations had become so disorienting an enterprise that the arguments were no longer about the guilt or innocence of suspects charged in court. Something had happened in Manhattan Beach, something dreadful and massive and utterly new, but nobody -- not the parents, not the baby sitter whose garden was dug up in a search for buried bones, not the police officer who began seeing a therapist so he might stop lying awake at 4 in the morning with the words of the children all banging around inside him -- nobody in town could prove exactly what it was.

Possibly it was Satanists.

Possibly it was child pornographers.

Possibly it was "hysteria," a word favored by the defense attorneys, or "social contagion." Certainly it was a case without a tangible thing to grab and examine for fingerprints. No bodies turned up on the roadside in a crime like this; no open wound or missing jewelry could show without equivocation that a crime had taken place at all. Medical evidence would be offered by the prosecution and attacked by the defense, but in the end it was the words of the children on which everything had to rest. If you believed the children, a phrase that became a kind of fraternal cry within certain communities of Manhattan Beach, then one great crime stretched plainly before you, and the charges against Ray and Peggy Buckey only began to address it. It is carried still, this credo, so that at the small town crossing where the cheerful shopping boulevard meets the Pacific, station wagons pass from time to time with the printed words, "I BELIEVE THE CHILDREN," affixed in bumper stickers to the backs of their cars.

The Facts On Aug. 12, 1983, a woman named Judy Johnson went to the Manhattan Beach police with a report of one abused child.

There are not a great many undisputed facts in the chronology of the Manhattan Beach preschool investigations, but that is one of them. All sides also agree on the following:

That Judy Johnson told police her 2 1/2-year-old son had told her Buckey had molested him, and that two examining physicians found evidence that the boy had in fact been sodomized.

That a Manhattan Beach police officer telephoned five families in late August, advising them that the department was investigating "improper conduct" at the McMartin Preschool, and asking the parents to ask their own children whether anything had happened to them at school.

That one of these mothers testified that she immediately telephoned the school and asked Raymond Buckey's mother what the police officer was talking about. "She was really shocked, or acted like she was," the mother testified, "and immediately dropped the phone, and said, 'Ray!' in a kind of shrill scream. 'Oh, my God, Ray!' "

That the Manhattan Beach police department arrested Raymond Buckey on Sept. 7, and the next day sent a letter to the families of 200 current and former McMartin Preschool children. "Please question your child to see if he or she has been a witness to any crime or if he or she has been a victim," the letter read. "Our investigation indicates that possible criminal acts include: oral sex, fondling of genitals, buttock, or chest area, and sodomy, possibly committed under the pretense of 'taking the child's temperature.' Also, photos may have been taken of children without their clothing. Any information from your child regarding ever having observed Ray Buckey to leave a classroom alone with a child during any nap period, or if they have ever observed Ray Buckey tie up a child, is important."

That the responses to this letter prompted the Los Angeles district attorney's office to refer about 400 McMartin children to a Los Angeles diagnostic and therapy center called Children's Institute International, and that over the next four months CII staff members, led by a therapist named Kee MacFarlane, conducted videotaped interviews with the children.

That MacFarlane and the other therapists, who declined to be interviewed while the case is still in trial, told parents after approximately 360 of those interviews that their children had confirmed some sexual molestation, the descriptions of which included sodomy, oral copulation, nude photography, and group games that were generally referred to by particular names, like Horsey, or Naked Movie Star.

That on the basis of those sessions, subsequent interviews with the children themselves and reports of physicians who said their examinations had found medical evidence of sexual abuse, the district attorney's office brought 18 former McMartin students before a grand jury to describe what the children said had happened to them; and that on March 22, 1984, the grand jury indicted Raymond Buckey and six female McMartin teachers, including Buckey's sister, his mother Peggy McMartin Buckey, and his grandmother Virginia McMartin, who owned the school and was then 76.

That the indictment charged the seven defendants with 115 counts of child molestation, each of which they emphatically denied, and that when their preliminary hearing was concluded 22 months later, the newly elected district attorney dropped charges against all the defendants except Raymond Buckey and his mother, declaring the evidence against the other five, the early bulk of it gathered under the direction of his predecessor, "incredibly weak."

That Raymond and Peggy McMartin Buckey have insisted from the beginning that they are innocent and being victimized by children who have been led by adults to believe things that are not true.

That two attorneys left the prosecutorial team, and that one of them, a young prosecutor named Glenn Stevens, subsequently signed a book and movie contract that resulted in his laying out, in taped conversations added to the court record and partially reprinted in California magazine, his version of a prosecution he had become convinced was utterly without merit.

That in Manhattan Beach and the towns to the south and east, the McMartin school was the first of seven preschools in 1984 to close following allegations of sexual abuse; that in 1984 an employee of one of the other schools was tried for sexual abuse and released after the jury in his case was unable to reach a verdict; and that a Los Angeles County Sheriff's Department task force assigned to the Manhattan Beach area sexual abuse reports spent more than a year in investigations and interviews with non-McMartin children who had described sexual molestation either in their preschools or locations, sometimes including other preschools under investigation, to which the children said they had been transported.

That when the task force disbanded, no charges were filed as a result of its investigation.

That on July 13, 1987, a Los Angeles criminal jury heard the first opening statement in the trial of Raymond and Peggy Buckey, who are charged with 100 counts of child molestation for acts allegedly committed between the years 1978 and 1983.

That in February Raymond Buckey was for the first time granted bail; that the bail was set in the amount of $3 million, the bond for which Buckey was unable to raise; and that for nearly all of the 4 1/2 years since he was arrested, Raymond Buckey has been incarcerated in the Los Angeles County Men's Central Jail.

'The Judas' "I had no doubt in my mind," Glenn Stevens says, "that these people were all guilty. Not only that, I relied so heavily on the expertise -- I shouldn't say 'expertise,' I should say 'opinion' -- of Kee MacFarlane, and all those in the case, that I believed the number of suspects was much greater than seven. Much greater. At one point, there were over 40 suspects."

In March 1984, seven months after the arrest of Raymond Buckey, Glenn Stevens was asked to join the newly reconstituted prosecutorial team of what was promising to be one of the biggest and most heavily publicized cases the Los Angeles district attorney's office had ever undertaken. He was 30 when he joined the McMartin case, a prosecutor who had worked principally with gang investigations, and he knew that then-District Attorney Robert Philibosian was mounting a summer reelection campaign as the Manhattan Beach preschool investigations unfolded before hundreds of parents and an intensely interested public.

From the day he stepped into the McMartin case, Stevens says, it was plain to him that this prosecution would be like no other he had encountered before. Prosecutors rely on evidence, he says, evidence that on its face can help convince jury members about the crime they are considering and the people who committed it. "McMartin wasn't like that," he says. "Everything -- every little thing -- had to be explained. Including the tapes. Starting with the tapes."

By spring 1983 the CII videotapes had filled drawer after drawer in the locked file cabinets of the prosecutors' evidence room, and Stevens began watching them, one at a time. He had never seen interviews like these, exchanges between a small child and an adult questioner who spoke sometimes in the voice of Mr. Detective Dog and sometimes in the voice of Mr. Pac-Man. When a child said something about abuse, about touching or naked games, the therapist often praised him, Stevens says; when the child said nothing had happened, the therapist rephrased the question, or asked it again, or told the child about other children who had already said the bad things happened to them.

It seemed to Stevens that in any ordinary criminal case these were textbook examples of what are called "leading questions," and he had already heard the defense attorneys assail the taped CII interviews as leading to the point of bullying. But this was not an ordinary criminal case. Interviews with small children are not like interviews with adults, especially if the working assumption is that the children may be harboring secrets they are terrified to reveal. Stevens says he had been warned about that, and he knew also that Kee MacFarlane and the other therapists were not and had never been trained criminal investigators. When therapists work, they traditionally carry no constraints about what may or may not be admissible later as evidence in court.

Also there was medical evidence on some of the children; physical examinations, many of them conducted by a physician working at CII, had found what the examining doctors reported was scarring or other signs consistent with sexual abuse. That was complicated too, Stevens later learned; much of the medical evidence was documented by photographs taken with the aid of a colposcope, a magnifying device whose usefulness for child sexual abuse investigations has only recently been explored. There is no national medical agreement on exactly what colposcopic pictures show -- on what kinds of magnified scars, for example, are unquestionably the product of sexual abuse.

It was the videotapes, though, that seemed to Stevens to form the backbone of the McMartin investigation; it was in those interviews that most of the children had given what were recorded as the first molestation reports. "The child could use a puppet to nod yes or no," Stevens says. "The question would come, 'Well, if you were tied up at school, how would they have done it?' And she would hand a child a piece of string. And the child would take the string, at random, and tie up one of the dolls. This is to signify how the child was tied up at school. This is what I'm watching on the tapes."

There is a nickname for Glenn Stevens among some of the parents and therapists who have spent the last 4 1/2 years with children they believe were molested at Manhattan Beach preschools. He is called "the Judas." He works now at a real estate office on the edge of Beverly Hills, his ties to the district attorney's office formally severed two years ago, and his critics, including the current McMartin prosecutors, suggest that there are serious credibility problems with someone involved in a book and movie whose success can only be enhanced by controversy.

"I should have lambasted this thing a year before," Stevens says. He says he had lost faith in the McMartin case before he ever considered saying so publicly. He knew early on that children were adding to their molestation reports items of information that seemed to him instantly to strain belief; when he began himself to interview the children he wanted to put on the stand, they told him they had been driven to a farm, that someone had taken them to a car wash, that Mr. Ray had molested them as the car was being washed.

"Confabulation" was the term the prosecutors used, and it was, Stevens says he was told, a problem they could manage. Frightened children can be tricked, or become confused. Perhaps the children had imagined certain flourishes; perhaps molesters had made them believe wild things, understanding how thoroughly this would undermine the stories the children told later. In any case, the "core event" remained intact. The children remembered the touching or the sodomy, the core event, and they remembered that even as other stories seemed to mount around it. A sorting job, Stevens says, confronted the prosecutors early on: The molestation details go here, the potentially less plausible accounts go here, and in this group we put reports like the one of the child who says strangers took him hang-gliding off the cliffs of Palos Verdes.

But the preliminary hearing, a grueling undertaking that went on for 19 months, included cross-examination of the children. There were seven defendants at the hearing, Ray Buckey and the six McMartin school women, which meant many defense attorneys were available to stand before 6- and 7-year-olds and ask them, their voices generally kept gentle enough to avoid antagonizing anybody too much, about all the things that happened while they were at the McMartin school.

Some of the testimony was strong, from a prosecutor's perspective, and some was not. Children named the parts of their bodies they said had been touched or penetrated; one child said Raymond Buckey had cut the ears off a rabbit to frighten children into silence; one child said he had been taken to a church and made to drink rabbit's blood. A 10-year-old boy left the stand after selecting from some photographs the faces of strangers he said had molested him; among the faces were those of the actor Chuck Norris and some nuns who had been photographed in the 1940s for a convent school yearbook. "The core event hasn't changed," Stevens would say, when the reporters pressed around him after court, but it began, he says now, to feel lame when he said it.

Also there was the matter of Judy Johnson. It was Judy Johnson's abuse report that had first set off police investigation of the McMartin school, and Judy Johnson, as far as Stevens could see, was mentally disturbed. The prosecution had in its possession a small collection of handwritten memoranda -- notes Stevens says he dug from the boxes of information that awaited him when he entered the McMartin case -- in which Johnson, or investigators interviewing her, had recorded what she said were the details of her conversations with her son, naming in some instances the teachers against whom charges were later dropped.

Feb. 22, 1984: "{Johnson's son} feels that he left LAX {Los Angeles International Airport} in an airplane and flew to Palm Springs. He described the airplane as like one used by Federal Express, only it had windows." Feb. 16, 1984: "The person who buried {him} is Miss Betty. There were no holes in the coffin. Babs went with him on a train with an older girl, where he was hurt by men in suits ... {he} was hurt by a lion. An elephant played with the lion. Squirted H20."

"When I read it, I thought, 'Oh, my God, the woman has flipped out because her kid was abused,' " Stevens says. Johnson telephoned him once from Seattle, saying she was in a hospital but did not know how she had come to be there; she said an AWOL Marine was following her but that she was not sure who he was; she said her dog had been sodomized, and that her estranged husband had molested her son. Judy Johnson never testified at the McMartin preliminary hearing. She was found in her home in December 1986, dead at 42 of what the coroner's office listed as "fatty metamorphosis of the liver," and other physical problems often associated with alcoholism.

By the autumn of 1985, Stevens says, he had become so uneasy about the McMartin case that the sight of the women defendants in court made him uncomfortable. He says there were indications already that some in the prosecutor's office thought the cases insubstantial against all but Ray and Peggy Buckey, but the decision had been made to complete the preliminary hearing before making any changes in the charges. "You're seeing these women looking at you, and they're paying for a lawyer, and they're watching their reputations go down the toilet, and you know you're not going to trial with these women," he says. "And one day, I just got fed up, and frustrated. And I said what I said."

What Glenn Stevens said, with a variety of reporters taking notes as he said it, was, "Kee MacFarlane could make a 6-month-old baby confess to being molested." He meant a baby that had never been molested at all, an implication that was not lost on the Stevens superiors who read the quote in the newspapers and also discovered that he had been the unnamed source -- and then lied about it afterward, as he now acknowledges -- in some articles discussing prosecutorial conflicts in the McMartin case.

Stevens resigned from the district attorney's office in January 1986. Chief Deputy District Attorney Gilbert Garcetti says Stevens had been offered the chance to resign rather than being fired, which Stevens disputes; he says he was never told a firing was imminent. Garcetti says it was not Stevens' opinions but his lying that cost him his job, and he also says Stevens appeared during most of the proceedings to be wholly committed to the prosecution. "Unfortunately, when it comes down to the McMartin case," Garcetti says, "I think his credibility is utterly lacking."

In fact there was debate among the prosecutors; one of Stevens' colleagues was transferred from the McMartin case at her own request after she said she was not convinced the evidence for conviction was sufficient against any of the defendants except Raymond Buckey. The prosecutors who have stayed with the case say internal disagreement should not be surprising, that from the start this has been a memorably complex and difficult case, but that they believed overwhelming evidence lay before them -- the children's testimony, the medical records, the numbers of children with similar stories.

"Our evaluation to proceed to trial with the two defendants that are remaining was an independent evaluation made of all the evidence, which included just hundreds of hours, thousands of hours, of investigation time, Manhattan Beach police department, lawyer time -- and, frankly, of analysis time," says Dan Murphy, special operations director for the district attorney's office. "We're aware of the complaints. We've heard them constantly from the defense lawyers. The fact is that after our evaluation of the evidence we've decided that this is simply not a case we're going to walk away from."

Four months after Glenn Stevens resigned, he signed a contract -- he declines to discuss the financial details -- with the movie producer Abby Mann and his wife Myra Mann, who were at work on a book and motion picture that will describe the McMartin prosecution as a massively misguided effort built on the stories of children coaxed into fabrication. Stevens says he did not entirely believe this when he resigned, that the medical evidence had persuaded him the prosecution had a convincing case against Raymond Buckey.

It was during the following months, Stevens says, as he talked to the Manns and learned more about the debate over the medical evidence itself, that he lost faith even in the Raymond Buckey case. Stevens began to think of the board game children play, the one in which a single pushed lever sets off a complicated chain reaction that at its finale drops a trap over a plastic mouse. He began to think, he says, about a small community in which 200 sets of parents are asked by mail to question their children about sodomy and oral copulation, and then talk about it in large group meetings, and then send their children to the same therapists and the same elementary schools.

Glenn Stevens says he began to think about a mentally unstable woman in the midst of this, and to wonder when the mental instability might in fact have begun. "Judy Johnson is Mousetrap," he says. "She's the foundation. That's what I call it. My Mousetrap."

--------- Ritual Abuse: A Perplexing Pattern of Reports-------------

In February 1985, two dozen investigators and prosecutors gathered in a conference room at the Quantico FBI academy to trade information on a kind of case that had left each of them groping both for explanation and for guidance on how to proceed. The FBI man in charge was a behavioral sciences expert named Kenneth Lanning, and in the months since the 1983 onset of the McMartin case, Lanning had come across a growing collection of cases in other parts of the country that for the first time were confronting police with some of the same elements: Groups of young children, groups that expanded as the cases unfolded, were making sexual molestation reports that swelled with passing weeks to include descriptions of candles, robed adults, chanting, blood, feces and sometimes human dismemberment.

Nothing concrete linked these cases to the Manhattan Beach investigations. All Lanning knew was that police and prosecutors from jurisdictions thousands of miles apart were suddenly calling him with questions he had never encountered before the summer of 1983.

"They were allegations where we couldn't prove them or disprove them," Lanning says. "I was baffled by it, because the children in some cases seemed to be alleging things which didn't seem to be true. Some of these kids were describing injuries for which there were no scars, and murders for which there were no bodies ... After the first half dozen or so cases, I started to ask, 'Well, gee, what in the world is going on here?' "

Glenn Stevens, the Los Angeles prosecutor who later defected from the McMartin case, went to Quantico, as did investigators from San Francisco, New York, Tennessee, Virginia, Illinois and Minnesota, where the town of Jordan had already been consumed by charges of ritual sex and murder that were filed against a group of adults and then withdrawn. The word "ritual," in the absence of some more specific description, had become a link for many of these investigators; in the cases before them, as well as what Lanning estimates by now to be at least a hundred others he has studied himself since the first McMartin arrest, children under the age of 6 were describing what seemed to be elements of something that looked like satanic ceremony -- with almost no adult corroboration or physical evidence to prove it.

"Everybody ended up leaving the meeting probably in one sense a little more confused than they were when they walked in, but also saying, 'At least someone is dealing with it,' " says Sandi Gallant, a San Francisco police officer who has examined adult murders and other crimes that are clearly complicated by some sort of satanic worship. It is partly that background, Gallant says, that leads her to think there may be substance to what some of the recently formed local parent groups are now insisting -- that these children are describing real events involving real adults who are using groups of children for some ritual as well as sexual purpose.

Where but in actual experience, these parent groups have asked, could small children obtain such graphic and frightening detail? "A 5-year-old being able to know what color the lungs are inside a person," says Leslie Floberg, a Manhattan Beach homemaker who has listened to many families' accounts and believes her own son's description of ritual abuse was based on his actual experience. "How does a 5-year-old know that?"

Perhaps, Kenneth Lanning suggests, modern 5-year-old children -- particularly 5-year-old children who live in communities undergoing a highly public child abuse investigation -- know a great deal more about sexual abuse and possible macabre variations on it than their parents think they do. "Parents talk about it," he says. "Parents say, 'Well, we always talk about these things late at night, when kids aren't listening.' Well, you think they're not listening."

Some of the stuff of these accounts, Lanning says, is available to small children in the world around them. "You walk into any video rental store in this country, you go to the horror thriller section, and you can probably find 10 to 30 movies where there's a strong witchcraft-occult theme," he says. And after the widely publicized disappearance of 6-year-old Adam Walsh, he says, a police officer described to him the responses small children had given the officer when they were asked what they thought might actually happen if a bad stranger got hold of them; the horrors they listed, the officer said, were appalling.

Although Lanning's Quantico conference was one of several held over the last three years to try somehow to draw these cases together, there appears to have been no psychological study to examine what besides actual experience might be generating these reports. There are documented instances of community mass hysteria that are much more recent than the adolescent girls claiming to have been bewitched in 17th-century Salem; psychologists have examined modern-day factories and schools, for example, where groups of people began fainting, complaining of illness for which no physical cause could be found, or insisting they were being bitten by bugs that proved not to exist.

But if mass hysteria is to be blamed even in part for the abuse reports appearing around the country, there is no consensus at all on what might be generating the hysteria and stocking it with such vivid and morbid details. "These are not part of the ordinary fairy tale life of a child," says Lenore Terr, a University of California at San Francisco psychiatry professor who has studied the way traumatized children recall their experience and communicate it to others. Terr has found that children can pass on elements of their own frightening experience to others -- after the famous 1976 kidnaping of the busful of schoolchildren from Chowchilla, Calif., for example, she found kidnaper-and-victim games played by local children who had not themselves been part of the abduction.

"Usually at the bottom of these things you can find a victim," Terr says. "I think someone probably was exposed to satanic rites. I don't think you can make this up if you were just exposed to incest." Terr says she has also never encountered children who appeared to be absorbing other children's traumas by insisting it actually happened to them.

Earlier this year, as part of a newspaper series on ritual abuse cases, two reporters from the Memphis Commercial Appeal examined 36 cases from around the country and found some of the same details surfacing in communities hundreds or thousands of miles away: a one-armed man was reported in Jordan, Minn., and in Memphis; children reported being taken to cemeteries or funeral homes in California, Michigan, and Washington state; children from nearly every section of the country described the stabbing, murdering or burning of babies. The Commercial Appeal series raised the suggestion of "urban legends," the apparently foundationless stories, like the one about the woman who was supposed to have dried her poodle in the microwave oven, that sociologists have studied as they travel around the country.

But so volatile is this entire subject by now that believers in satanic abuse, many of whom gather periodically for meetings of their own, see the disparate accounts as further evidence that the stories are real; children that far apart could not possibly have passed information back and forth, they say, so perhaps similar details on both ends of the country simply prove that abusers follow similar ritual.

"I don't want anybody to misunderstand me here -- I've been involved in many, many cases where people dabbling in Satanism committed murders," Lanning says. "This is the most polarizing issue I've ever been involved in, and it divides people into two extremes ... What I believe is that in these cases there is a middle ground, and just because a group of children did not participate in a satanic murder doesn't mean they weren't sexually abused. What happens in these cases is the children allege some of these things, some of it's pretty far-fetched, some of it's a little extreme, and you check it out and it's not there, so somebody wants to take the whole case and flush it down the toilet." Cynthia Gorney.

Tomorrow: Among the believers.