11:09 AM PST on Saturday, March 11, 2006

By WILLIAM McCALL, Associated Press Writer

An FBI mistake on fingerprint identification in the 2004 Madrid train bombings was a "watershed event" that has resulted in improvements but more needs to be done to make sure the mistake isn't repeated, according to a federal report.

(kgw.com graphic)

The 330-page Department of Justice report released Friday, the day before the second anniversary of the bombings, repeats the department's earlier conclusion that FBI fingerprint experts could have prevented the misidentification that led to the mistaken arrest of Portland attorney Brandon Mayfield.

"Among other things, the examiners applied circular reasoning, allowing details visible in Mayfield's known prints to suggest features in the murky or ambiguous details ... that were not really there," the report by the department's internal watchdog said.

FBI experts mistakenly matched fingerprints found on a bag of detonators in Madrid to Mayfield's after the March 11, 2004 train bombings that killed 191 people and wounded more than 1,500.

Mayfield, who was jailed for two weeks in 2004 on a material witness warrant, was released after the FBI acknowledged the fingerprint was not his. Mayfield is suing the federal government.

The Justice Department's Inspector General, Glenn Fine, conducted an investigation into Mayfield's arrest.

An executive summary of Fine's conclusions was released in January. In that report, Fine faulted the FBI for sloppy work but concluded the government did not misuse the Patriot Act during its investigation and did not target Mayfield because he is a Muslim convert.

The report released on Friday elaborates on those findings.

Whole blocks of some pages of the report have been blacked out -- especially details of the FBI's search of Mayfield's home and office, which are the subject of Mayfield's lawsuit.

The full report states that the misidentification of the fingerprint was a "watershed event for the FBI Laboratory, which has described latent fingerprint identification as the 'gold standard for forensic science."'

The report praises the FBI for undertaking "an ambitious research project" to improve fingerprint identification as a result of the mistake.

But Fine adds that "we found that some of the changes adopted by the (FBI) Laboratory were not fully responsive to the issues raised by the Mayfield misidentification."

The report outlined at least six recommendations, including alternate procedures for verifications; reviewing previous cases based on a single fingerprint identified through the FBI's Integrated Automated Fingerprint Identification System; and requiring documentation of observed features before comparisons are made to confirm the identification.

U.S. Attorney Karin Immergut, whose Portland office handled the investigation with the FBI, defended local agents and noted that a separate report by the Justice Department's Office of Professional Responsibility found there were no ethical or professional violations by her prosecutors in the case.

"The errors that occurred were in the FBI lab and the people in Portland were not at fault," Immergut said Friday.

FBI Special Agent Ann Todd, a spokeswoman for the lab, said the two FBI experts who first examined the print are not doing case work, "although both have been cleared to do so."

An FBI contractor who verified the identification, John Massey, has "not performed any work for the FBI laboratory subsequent to the Mayfield error," Todd said.

She said that in the 73-year history of the latent prints operation, the lab has made an erroneous identification once every 11 years, on average.

Gary Jones, a former FBI fingerprint expert and author of two books on forensic science, called the Mayfield misidentification an "aberration."

But Jones said that his own examination of the print, even in hindsight, indicated there was doubt about a match.

"I don't understand how the identification was made, much less verified in triplicate like it was," Jones said. "It's like a mathematician saying 2 and 2 is 5."

Jones said fingerprint experts are suffering the fallout from the mistake, including him, even though he had left the agency long before the Mayfield case.

"Even though I'm an independent examiner, I still get hit with this," Jones said. "If I testify in court the first thing they ask is, weren't you in the latent print section, and isn't this the same section that misidentified Mayfield?"