In a characteristically dramatic gesture, Gov. William Donald Schaefer exercised one of the most awesome powers of his office last month -- the power to reverse court-ordered penalties for murderers.

"After a careful and comprehensive review" of each case, Mr. Schaefer commuted the sentences of eight women he described as victims of continuing abuse by their mates -- seven of whom killed the men. He became only the second governor in the nation to free a group of women said to suffer from "battered spouse syndrome," a psychological disorder that can purportedly drive people in violent relationships to kill their abusers.

"It was not one of those things that you just jump in and say, 'Well, I'll do this.' " the governor declared at a news conference Feb. 19 in which he also pressed for legislation that would allow evidence of battered spouse syndrome to be heard in Maryland courts.

Yet an examination by The Sun of the seven murder commutation cases indicates that the process by which those cases were reviewed and prepared for the governor's action was not comprehensive. Nor are the facts of the cases as unequivocal as the governor suggested.

The documentation of abuse, prepared by the House of Ruth, an advocacy group for battered women, accompanied by some checks by state public safety and parole officials, overlooked or ignored readily available evidence that contradicts the contentions of at least three inmates set free:

* Bernadette Barnes, 45, of Baltimore was represented as having murdered her abusive husband after a rape and severe beating.

Unbeknown to the governor and the staffers who prepared the information he was given, Barnes was actually convicted of hiring a hit man to kill her husband. In a plot developed weeks before the murder, co-conspirators were assured that she would collect on a $22,000 life insurance policy.

"It was a contract killing, pure and simple," says Stuart O. Simms, the Baltimore state's attorney, who said that neither he nor his office was contacted by state officials looking into the case for a recommendation on commutation.

* Patricia Ann Washington, 33, of Hyattsville, was represented to the governor as a victim of 10 years' abuse by her husband. Yet in her statements to police after the murder, to a jury on the witness stand and in repeated psychological evaluations after trial, Washington consistently maintained that she had never been hit by her husband before the night of the murder.

"He never hit me," she testified.

The governor was also told that Washington killed her husband "during attack," but no independent evidence could be found in the police investigation or trial record to corroborate any abuse by her husband -- either on the night of the murder or in the years before.

* Virginia J. Johnson, 27, of Salisbury, killed her boyfriend during an attack, according to the case review memo presented to the governor. But witness statements and physical evidence in the legal record indicate otherwise.

Moreover, a review of the case would raise the possibility of a violent nature outside of the abuse syndrome.

While out on bail awaiting her murder trial, Johnson was arrested again after she allegedly pulled a knife on a potential witness. and threatened: "Bitch, I'm gonna kill you. I done already killed one person, and I'll kill you, too."

"The governor has the power and authority to do what he did,said Davis R. Ruark, Wicomico County state's attorney. "But I'm concerned that neither I nor any investigator has had input."

Confronted by the contradictions between the governor's portrayal of the cases and the legal record, Paul J. Davis, chairman of the Maryland Parole Commission, and Bishop L. Robinson, secretary of public safety and correctional services, acknowledged that they did not review trial testimony or other court records.

State officials did not query any of those involved in the original investigation, prosecution or defense of the women's cases. Nor did they attempt on their own to corroborate the histories of abuse cited by the women.

The Domestic Violence Legal Clinic at the House of Ruth, a shelter for domestic violence victims and an advocate for legal recognition of battered spouse syndrome, gathered the information.

Attorneys at he House of Ruth did not review the case records before sending the women's versions of events to state officials, but the material they did submit persuaded the governor to commute the women's sentences five weeks after he visited five inmates in prison--women he later described as having been "pushed absolutely to the brink."

In commuting the sentences en masse, Mr. Schaefer followed the lead of then-Gov. Richard Celeste of Ohio, who had issued similar commutations to a group of women some months before.

After the prison visit, the governor's office sent 12 names to the public safety secretary and the Parole Commission for their review, and of those, eight were recommended for early release. One of the eight was convicted for battery; the remaining seven were convicted of varying degrees of homicide.

Officials in the governor's office and the public safety department, as well as the House of Ruth advocates, say they remain confident that the commutation screening process was sufficient. But they will show little of the evidence.

Public safety officials said that whatever evidence they have to substantiate that the women murdered as a result of abuse is protected by both executive privilege and privacy statutes.

"In all cases we recommended, we found what we thought was evidence sufficient to convince us that they were battered women," said Mr. Davis of the parole commission. "This [review] is not dissimilar to what goes on in a courtroom. There are always going to be people who are not happy with a decision."

But privately, officials in the governor's office acknowledged that they were unsettled to learn from a reporter that Bernadette Barnes was involved in a contract murder with a profit to be had and that Virginia Johnson had allegedly threatened a potential witness in her case.

Informed about the profit motive in the Barnes killing, Nancy J. Nowak, executive director of the Governor's Office of Justice Assistance, at one point told a reporter that she had been assured by Barnes' lawyer at the House of Ruth that the woman didn't even know a life insurance policy existed until her husband's funeral.

This is contradicted by Barnes' own testimony at trial that she secured the services of a contract killer weeks before the murder promising a share of the $22,000 life insurance payment she eventually received.

Secretary Robinson indicated that he was aware of an insurance policy but viewed it as practically irrelevant to the murder, suggesting incorrectly to a reporter that Barnes had been paid only $5,000 in life insurance benefits and that most of that money had been spent on funeral expenses.

Officials in the governor's office who had the most direct hand in presenting the cases to Mr. Schaefer said that public safety officials were responsible for checking the women's stories.

Mary Ann Saar, the governor's director of operations and public safety, and Ms. Nowak, of the Office of Justice Assistance, continue to express confidence in the review process.

"I have faith in those who were charged with this process," Ms. Saar said.

Ms. Nowak acknowledged she had some initial reservations about the subjective nature of the House of Ruth's information, which was based largely on the convicted women's own accounts:

"It's a lot of self-report stuff, and I struggled with that. But knowing that cases like these generally involve self-reporting and a lack of documentation . . . I felt confident with the process."

However, she added, "it was my understanding they were to look at all available information."

Secretary Robinson and Mr. Davis said they based their recommendations to the governor on a review of the women's prison and parole files, which include police incident reports, pre-sentence investigations and brief prosecutorial summaries of the criminal cases, coupled with the histories of abuse prepared by the House of Ruth attorneys.

Mr. Robinson, a former Baltimore police commissioner, said he felt no need to check the women's accounts against the existing legal record. He said the inmate files contained enough information about the cases, contending that the full legal record was not consulted because information about domestic abuse is often excluded from Maryland trials.

"Sufficient elements are contained in these documents," said Mr. Robinson, tapping a folder whose contents he refused to disclose. "I followed the process that I thought was adequate."

Mr. Robinson and other state officials suggested that apparent discrepancies in the women's accounts could themselves be the result of the battered spouse syndrome, a post-traumatic stress disorder in which women can block out details of the murder or contradict themselves in their own statements.

Experts agree on this. Battered women may deceive themselves and others about their status as abused women or their role in the death of their mates because they wish to avoid humiliation and confrontation, said Dr. Ellen McDaniel, a psychiatrist well-versed in the syndrome.

Women suffering from the syndrome "are eager to please. They're thinking, 'Let me get through this next moment of stress, let's get through the moment.' They do on the [witness] stand what they do in their lives," Dr. McDaniel explained.

However, Dr. McDaniel, who has testified in favor of the governor's legislation, said that an investigator confronted with a woman said to be suffering from the syndrome could and should go to "all kinds of outside, objective sources" to check the woman's version of events.

If contradictions exist, Dr. McDaniel said, an investigator familiar with the syndrome might further examine the victim to determine whether those contradictions are caused by the syndrome or some other mental disorder, or whether the woman is simply lying.

In Maryland's commutation cases, state officials acknowledge that this did not occur.