How a Dubious Forensic Science Spread Like a Virus

Leora Smith:

The prosecution’s star witness — a forensics specialist named Herbert MacDonell — set out an array of props before the jury: a medicine dropper, a mirror hastily yanked from the wall of the courthouse bathroom and a vial of his own blood, drawn that day at a nearby hospital.

It was a strange sight in the 1985 Texas courtroom, and the jurors, the judge and even the defense attorneys watched, rapt, as MacDonell laid the mirror flat and then climbed up on a chair, holding the vial and dropper.

MacDonell’s expertise lay in an obscure discipline known as bloodstain-pattern analysis. He claimed he could reconstruct the events of a crime by reading the bloodstains left behind.

Like a professor performing a classroom demonstration, he dipped the dropper’s tip into the blood and, with a practiced hand, released a single drop onto the mirror. It landed with a muted thud, forming a perfect crimson circle.

Blood landing on a flat surface should not spatter, MacDonell told the jurors with satisfaction. He let another drop fall onto the white shirt he was wearing. Blood lands differently on fabric, he showed them.

A defense attorney shot up from his chair in protest. This was a murder trial. There was no mirror at the crime scene. No medicine dropper. The demonstration was not reliable science, he argued. The judge disagreed.

MacDonell’s testimony would be pivotal to proving the Fort Bend County prosecutor’s theory that 21-year-old Reginald Lewis had murdered his family, shooting his mother and two brothers, and setting his father on fire. MacDonell had identified dozens of minuscule blood spots on Lewis’ clothing, and he said they placed Lewis at the scene during the crime