Analyzing fingerprints was the first method that identified a unique personal signature for each individual, the details of which remain the same from childhood to old age.  Whenever we touch something, we leave a residue of oil, grease, dust, or sweat that forms from the imprint of our fingers, and that can place us at the scene of a crime.  (Unless, of course, some wily criminal has lifted your print from something you touched and placed it at the scene.)

Sirhan Sirhan (AP)

It was a fingerprint that revealed the identity of the man who had assassinated Senator Robert F. Kennedy.  Sirhan Sirhan would not give his name, but his prints gave him away.  Many other cases have been solved with fingerprints, and fingerprint identification has become one of the primary tools for solving crimes.

The first person to classify fingerprints was Prussian professor Johannes Purkinje.  In 1823, he described nine basic types.  Then in 1858, William Herschel recognized the individuality of fingerprints and used them for contracts for illiterate workers, while physician Henry Faulds soon discovered that prints could be made visible with powders.  He used this in a criminal case to free an innocent man.

Other scientists found that fingerprints were unchanging over time, and Sir Francis Galton proposed that all prints had three primary features: loops, whorls, and arches.  From these he could devise 60,000 classes.  Edward Henry added two more features---tented arches and loops that were either radial or ulnar---and developed a classification based on these five types that is still in place.  One had to first establish to which of these classes a print belonged and then to sub-classify it by distinct deltas (the formation of ridges).

In 1894, the British examined fingerprint classification as a way to replace the slow and impractical method of "bertillonage," which identified criminals via eleven body measurements.  In one case in the U. S., the measurements failed to distinguish between two criminals who were identical twins, but their fingerprints were dissimilar.  This case brought fingerprinting into its own as the leading tool for identification.

In 1910, Thomas Jennings was the first person in the United States to be convicted with fingerprint evidence.  When he broke into a home and shot the homeowner, he left four clear prints in wet paint.  The conviction was appealed, but the appeals court was satisfied that fingerprinting had a solid scientific basis.

In order to make a comparison match, ink pads were used to take a suspect's prints.  The prison systems quickly adopted this in order to keep on file fingerprints of known criminals.  By 1924, Congress had established a national depository of fingerprint records at the FBI, and today there are several hundred million sets of prints there.

How it works

Print chart and powder in crime lab (AP)

The surface of the skin on the palms of the hands and soles of the feet differs from the skin on the rest of the body, in that it is covered with ridges with clearly defined patterns.  The way these ridges lie gives us our individual marks.

Fingerprints can be latent, visible, or molded (plastic).

• Visible prints are left in some medium, like blood, that reveals them to the naked eye.
• Plastic prints are indentations left in pliable objects such as soft wax that take the impression.
• Latent prints are formed from the sweat from sebaceous glands on the body. They can't be seen, except under certain lighting, but can be raised to visibility with various methods. 

If left on a hard surface that remains undisturbed, prints can be permanent.  Often prints are only partial, but may still show a sufficient amount of the fingertip surface for point-by-point identification.

The crime scene investigator's job is to locate latent prints, develop them, and preserve them.  Developing latent prints so they can be seen involves powders or fuming with  compounds like ninhydrin or Super Glue.  Then the prints are preserved on a card for identification, and marked for the time they were lifted and the location in which they were found.

This involves using a powder appropriate to a particular surface, usually in a color that contrasts with the color of the surface.  The powder adheres to the print, so when a fine brush is used to remove excess powder, the print shows a clear pattern. It's then photographed, and if it's on a portable object, the whole object is taken to the lab.  Otherwise, the print is lifted with transparent tape and placed on the index card.  This method is still in use, although other methods have been developed as well---especially for surfaces that don't respond to powders.

FBI expert examining fingerprints (CORBIS)

To compare them, the print technician must first make sure that prints are taken of everyone who was at the scene or might have been there (if possible).  That includes taking them off any dead bodies.  To take a print at a scene, an ink roller is run over the fingertips and the tips are then pressed against a card.  At the police station, the person dips his fingers into printer's ink and then presses them onto a card.  Then each handprint is inked and preserved in the same way.  That provides the database to which they're sent a set of ten distinct prints, along with the size of the hand and the shape of the palm.

Since 1972, fingerprints have been compared and retrieved via computer.  By 1989, they could be sent back and forth online.  State and local agencies built up automated fingerprint identification systems (AFIS), and the FBI opened the National Crime Information Center (NCIC), which expedited the exchange of information among law enforcement agencies.  They introduced a standard system of fingerprint classification (FPC), so that information could be uniformly transmitted from one AFIS computer to another.

The computer scans and digitally encodes prints into a geometric pattern according to their ridge endings and the branching of two ridges. In less than a second, the computer can compare a set of ten prints against a half million.  At the end of the process, it comes up with a list of prints that closely match the exemplars (the originals).  Then the technicians make the final determination, which involves a point-by-point comparison. (Before computerization, searching manually through print files was an arduous, time-consuming task.)

The patterns for fingerprints, formed by the ridges, are classified into four basic groups:

1. Arches: formed by ridges running from one side to the other and curving up in the middle.  Tented arches have a spike effect.  About five percent of people have this print.
2. Whorls: Thirty percent of prints form a complete oval, often in a spiral pattern around a central point.
3. Loops:  These have a stronger curve than arches, and the ends exit and enter the print on the same side.  Radial loops slant toward the thumb and ulnar loops toward the other side (which makes it important to know from which hand the print came).  This comprises about 60 percent of fingerprints.
4. "Composites" intermix two of the other patterns, while "accidentals" form an irregular pattern.

Once prints are classified and sub-typed, they get stored in the database for future comparisons.

Other types of physical evidence have classification systems and databases as well, but evidence alone does not make a case.  Whatever the technicians find must support a theory about the crime.  Let's look at how that's derived.