The Evolution of Forensic Firearms Identification

I’ve previously discussed forensic bullet and cartridge case comparison, but I want to include some history. First, “firearms identification,” not “ballistics,” is what crime labs do.

Firearms identification today is highly evolved. Labs have digital microscopy capabilities and can share images of evidence items with faraway labs via the “send” button.

It was not always high-tech. True microscopes appeared about 1590 to 1625, but three centuries passed before effective and expedient firearms evidence comparison devices appeared.

Rifled firearms were documented between 1493 and 1506, but any forensic usefulness of rifling impressions on bullets had to wait another four centuries to be appreciated. Early firearms evidence was painfully basic.

The first civilian firearms were bored in many diameters for round balls and came with a bullet mold to cast the correct size ball. Diameters became part of crime investigation. People testified to a bullet’s compatibility with a suspect’s firearm, and although an exact match was impossible, size could absolve a suspect if a fatal ball was incompatible with his firearm(s).

In 1835 Londoner Henry Goddard advanced the discipline when he saw small marks on a recovered homicide bullet. He realized they were not from a firearm but from machining marks in the mold that cast the evidence bullet. He matched the marks to a mold found in the suspect’s possession.

In an 1852 homicide trial in Oregon Territory, the court allowed a sheriff’s testimony regarding his testing to determine whether a hole in the victim’s shirt was a tear or the result of a bullet. Test shots he fired into the victim’s shirt led the sheriff to conclude the defect was indeed a bullet hole.

During the American Civil War, surgeons removed a roughly .69-caliber round lead ball from CSA Gen. “Stonewall” Jackson. Union forces he engaged that day had only .58-caliber rifled muskets firing conical Minié bullets, but some of Jackson’s own troops carried .69-caliber smoothbore muskets. Oops.

Rifled arms became nearly universal after the Civil War, and researchers noticed striations that rifling had imparted to fired bullets. An 1879 Minnesota court admitted a gunsmith’s testimony that two different suspect revolvers produced different rifling patterns on bullets, eliminating one revolver. A 1900 article by Dr. A.L. Hall detailed the nature of rifling impressions on bullets and measurements thereof.

---

Examinations of rifling impressions on bullets were constrained to microscope photography, comparing conventional photos—not the actual bullets—to each other. In about 1912, the French researcher V. Baltha-zard developed better photo methods that recorded all the impressed surfaces of a bullet in a sequence of images. It was still a slow and indirect method.

A deputy sheriff from Miami, C.A. Peterson, improved on Balthazard’s methods with a “geared camera” that passed photographic film across a slit lens that faced a bullet on a mount synched by gears to the film drive. Bullets rotated at the same rate that the film moved across the slit lens, producing one strip image of the bullet’s entire bearing surface.

This was a seriously complicated tool, requiring more photography and engineering tweaks than firearms expertise. Overly tedious for any lab with a heavy case load, it also failed with deformed bullets. Improving production and eliminating complicated handling while working with real-world evidence became critical.

A “comparison” microscope from Germany was previewed in a 1913 issue of Popular Mechanics. It was a transmitted-light scope used for translucent materials—textiles, crystalline chemicals, etc. It had no provision for reflected-light exams or mounting bullet-shaped objects. However, it had the basics: two identical yet independent scopes sharing a common eyepiece bridge that showed objects under each scope side by side. Its time would come.

Chicago’s organized crime problem showed the need for live-view comparisons. The 1929 St. Valentine’s Day massacre left over a hundred bullets and cartridge cases; bullets alone would have required months to process with a geared camera. Because of visionary research, the solution was in place four years before the massacre.

In April 1925, C.E. Waite, Calvin Goddard, Phillip Gravelle, and John Fisher formed the Bureau of Forensic Ballistics (BFS) in New York City to assist U.S. law enforcement agencies, few of which had their own facilities. That was seven years before the FBI had any laboratory!

The most significant landmark for all modern firearms identification came from the BFS through Gravelle. He developed the means to adapt the older transmitted-light comparison microscope to allow proper live-view side-by-side examination of bullets and cartridge cases. Firearms identification today stands on that foundation. From that point on, most changes in how firearms examination is done has been evolutionary, not revolutionary.

The classic American Optical comparison scope I used for 16 years in Dallas was little removed from the converted models Gravelle and his associates used so skillfully. It was in superb condition and likely dated to the early 1950s. Compact and comfortable, it was good for hours of use with minimal fatigue. I let the other guys have the modern and more cumbersome scope we bought. Modern scopes are commonly tall and wide. If you were less than about 70 inches, many could be uncomfortable. I felt like I was riding most of them, and I’m a six-footer! My sense of history sitting at that old AO scope was undeniable.

However, science can be sidetracked by edict. An Argentine firearms examiner visited our lab in about 1983. In passing, he mentioned a geared camera, definitely stopping the conversation. I asked the translator, one of our lab staff, to verify. He said they made all determinations with modern scopes and techniques and testified to their opinions, but national court procedures required them to present strip photos of matching bullets to the judge. The judge made the final pronouncement of “match” or “no match.” I hope that has changed.