Montana Attorney General Mike McGrath is getting knocked around for statements he made recently in the high-profile case of Jimmy Ray Bromgard, who was released from prison in 2002 after DNA tests exonerated him in a Billings rape case.

Last last month, the Chicago Tribune ran a withering article in which forensic and legal experts castigated McGrath for suggesting, under oath, that Bromgard may not be innocent -- despite the DNA evidence.

The statements were made during a deposition of McGrath last fall by lawyers for Bromgard, who is suing the state and Yellowstone County for damages because he was wrongly imprisoned for 15 years.

Yet while the news coverage focused on McGrath's statements about the Bromgard case, much more of the deposition focuses on something else: Whether the state, under McGrath's direction, adequately investigated the questionable work of former state crime lab analyst Arnold Melnikoff.

It was Melnikoff's testimony on hair comparisons that helped convict Bromgard in 1987 -- testimony later described by a panel of national forensic experts as "completely contrary to generally accepted scientific principles."

Melnikoff left Montana's crime lab in 1989 and went to work for the Washington State Patrol's crime lab.

In the wake of revelations about the Bromgard case in 2002, Washington state officials audited Melnikoff's work there and found errors in 30 of the 100 cases they examined.

Washington ended up firing Melnikoff in 2004, citing his "incompetent and inaccurate" testimony in yet another Montana rape case in 1990, involving hair and blood analysis that helped lead to the conviction of Paul Kordonowy.

Kordonowy was exonerated of that crime by DNA evidence in 2003, although he remains in prison on a separate rape conviction.

Melnikoff also did hair analysis in yet another Montana criminal case since called into question: The murder conviction of Barry Beach, who says he's innocent in the 1979 killing of Kim Nees in Poplar. Beach's request for clemency was heard last week before the state Board of Pardons and Parole.

Melnikoff identified a pubic hair found on the victim's sweater as belonging to Beach. Melnikoff's analysis was never admitted as evidence at Beach's trial, but it was mentioned at trial by the prosecutor Marc Racicot, later elected attorney general and governor.

After the Bromgard exoneration in 2002, McGrath oversaw an investigation of Melnikoff's Montana crime-lab work from the 1980s and late 1970s. The state crime lab is part of the state Justice Department, which McGrath leads as attorney general.

Attorneys for Bromgard, including Peter Neufeld, co-director of the New York-based Innocence Project, have criticized McGrath for conducting what they see as an inadequate, limited investigation.

In the Sept. 29 deposition, McGrath, under questioning by Neufeld, attempts to explain why he declined to hire independent, outside specialists to direct the investigation and why his office scrutinized only Melnikoff's hair analysis cases, rather than other work.

The investigation also did not examine potential effects of Melnikoff's training of other crime-lab workers, did not scrutinize cases where defendants had pleaded guilty, didn't re-examine the hair samples analyzed by Melnikoff, and didn't try to find the "root cause" enabling Melnikoff to repeatedly provide testimony and analyses later found to be seriously flawed.

"What was your reason for rejecting the recommendation of the peer review committee (of national forensic experts) with respect to bringing in outside experts?" Neufeld asked McGrath at one point.

"I didn't think that any purpose would be served," said McGrath, who is now running for chief justice of the Montana Supreme Court. "It's also a very expensive process."

"Did you make any inquiry as to what the cost would be of bringing in one or more experts?" Neufeld asked.

"No," McGrath replied.

"But nevertheless, without doing any research, and getting any numbers, or asking any experts what they would charge, you concluded it would be too expensive; is that a fair statement?"

"That was one of the factors that I considered," said McGrath.

As McGrath went on to explain, his office focused on whether Melnikoff's hair-analysis work may have led to wrongly convicting someone.

The office, with the help of crime-lab personnel, identified about 250 criminal cases where Melnikoff had "generated a report."

The review then zeroed in on some two dozen cases in which the defendant was in prison and Melnikoff's hair report had a "positive association" linking the defendant to the victim of the crime -- just as he had done with Bromgard, erroneously. Melnikoff also testified in some of these cases.

McGrath said he did not scrutinize 13 of these cases where the suspect pleaded guilty.

Ultimately, McGrath, with the help of investigators and staff in his office, determined that Melnikoff's testimony was not the deciding factor in the 12 cases where someone was convicted and sent to prison.

"None of the individuals listed in the attached report were wrongly convicted based on the work or testimony of Arnold Melnikoff," he said in a January 2004 memo on the review.

McGrath rejected a request by Neufeld and the Innocence Project to audit all of Melnikoff's work with an independent "oversight committee" and outside forensic experts -- regardless of whether the case involved hair analysis or whether the suspect had pleaded guilty or was in prison.

The Innocence Project notified Washington state in 2002 of Melnikoff's errors in Montana, and urged a review of his work.

The Washington State Patrol responded by auditing 100 of 1,300 cases on which Melnikoff had worked from 1998-2002. Like Montana, it did an internal review, without calling in outside experts.

However, unlike Montana, the Washington State Patrol did not publicize its 2003 audit findings. It took an investigative report by a Seattle newspaper a year later to bring the findings to light. Prosecutors and defense attorneys involved in the cases where Melnikoff's work was questioned had not been informed of the findings.

Meanwhile, U.S. Magistrate Carolyn Ostby of Billings, who is presiding over the Bromgard lawsuit, last week issued a temporary "gag order" on all parties to the case, forbidding them from talking to the media.

The gag order came in the wake of complaints from the state's lawyers, who said Neufeld's release of the McGrath deposition and other depositions to the media had violated a "protective order" and created negative publicity that would make it difficult to seat an unbiased jury in Montana.

Shelton Williams, a Missoula attorney representing the state, went so far as to ask Ostby to seal the entire court file from public view and order all parties in the case not to talk to any reporters until the case is completed.

The gag order will remain while Ostby and the parties debate what will be "the least restrictive means" to ensure the state gets a fair trial. The parties also are sorting out just what documents in the case must be kept from the public by the protective order.

Neufeld said in court documents that it's quite clear the deposition of McGrath (who is not a named defendant in the lawsuit) is not protected, and was properly released.

Regardless of what the court may rule on sealing certain documents in the Bromgard lawsuit, the McGrath deposition is in the public domain. In fact, the Chicago Tribune has posted the transcript of the deposition on its Web site at www.chicagotribune.com/news/local/chi-mcgrath,1,324227.htmlstory.