The Science of Sherlock Holmes: From Baskerville Hall to the Valley of Fear, the Real Forensics Behind the Great Detective's Greatest Cases

by E. J. Wagner

  Across the channel, the French had followed the Tichborne case with interest. Since the formation of the Sûreté, the identification system in Paris had depended largely on the extraordinary memory of Eugène François Vidocq. The huge number of files that he had developed were of limited use without him. Dossiers were listed under the name of the criminal, and a criminal typically used many names. Vidocq was able to recall enormous numbers of aliases and could tap into his proficient system of police spies to keep abreast of name changes, but his less adroit successors were lost.

  The advantages afforded by photography were quickly appreciated, and “mug” shots by daguerreotype, the complex photographic process first introduced by Louis Jacques Daguerre in 1839, were used by police in Belgium as early as 1843. But daguerreotypes were extremely expensive and required skill and considerable time to produce. As a result, they were not often used by police.

  Even by the middle of the nineteenth century, by which time photographic methods were simplified, the process was still cumbersome. Slow lenses required identification photographs to be made in sunlight, and the exposure time could be as long as twenty minutes. The unhappy subject was often strapped in a chair to limit movement.

  As lenses became faster and photography cheaper, police made greater use of the technology. In Paris, Gustave Macé, head of the detective division, decided to require photographs of all criminals. The photographic files grew, filling cabinets and boxes and lining corridors but providing little useful information. Not only was there no clear method of organizing them, there was no standardized method of taking them.

  All the photographs were taken full face and at whatever distance seemed convenient at the time. Light came from whichever side the photographer fancied at the moment. Hair was allowed to cover ears, and facial hair often obscured features. The files and the pictures in them continued to be classified by the criminal’s name.

  And then, in 1882, it all changed, thanks to a twenty-six-yearold neurasthenic clerk in the Paris Police named Alphonse Bertillon. It is possible that Bertillon possessed some social graces, but if so, he was amazingly discreet about them. He rarely spoke, and when he did, his voice held no expression. He was bad-tempered and avoided people. He suffered from an intricate variety of digestive complaints, constant headaches, and frequent nosebleeds. He was narrow-minded and obsessive.

  Although he was the son of the famous physician and anthropologist Louis Adolphe Bertillon and had been raised in a highly intellectual atmosphere appreciative of science, he had managed to be thrown out of a number of excellent schools for poor grades. He had been unable to keep a job. His employment at the police department was due entirely to his father’s influence. But this misanthropic soul managed to accomplish what no one else had: he invented a workable system of identification.

  Sherlock Holmes remarks in The Hound of the Baskervilles, “The world is full of obvious things which nobody by any chance ever observes.” It was Bertillon who first observed the obvious need for a scientific method of identifying criminals. He recalled discussions in his father’s house about the theory of the Belgian statistician Lambert Adolphe Jacques Quetelet, who in 1840 had suggested that there were no two people in the world who were exactly the same size in all their measurements.

  Bertillon reasoned from this that if multiple measurements were taken of individuals and classified according to type, locating records would be easier and the chance of confusing subjects unlikely. In 1879, after a few months as a clerk, during which he had done some preliminary research, he wrote a report to offer his ideas to his superiors. The report managed to be both pedantic and complex and was dismissed as a presumptuous jest. The senior Bertillon, who realized that his difficult son’s theory was a brilliant application of science to criminal investigation, tried to intervene, but the police administration, demonstrating an impressive flair for avoiding creativity, refused to even consider the matter. There it rested until 1882 when the administration, intensely pressured by influential friends of Dr. Bertillon, gave in. The persistent clerk was given two assistants and some funds to further research his system.

  After considerable study, he designed a procedure he named anthropometry. It required at least eleven bodily measurements that were believed to be unchangeable after the age of twenty. These included: the total length of the arms when outstretched; height both sitting and standing; the length of the head; the width of the head; the width of the cheek; and the lengths of the right ear, left foot, left little finger, left middle finger, and each arm from elbow to outstretched middle finger. Each measurement was to be made three times and the mean chosen for the record.

  Alphonse suggested that all photographs be taken from the same angle and lit the same way, and that profile views be included. He also invented the portrait parlé, or “speaking picture,” to accompany the photographs. This required the recording of eye color, hair color, and complexion, shape of the head, stoutness, posture, voice, accent, scars or marks, and usual style of dress. The complete version of the portrait parlé required hundreds of exact bits of information, and detectives resisted its demands. Eventually, it was simplified into the “Wanted” descriptions familiar to us today.

  It was not a perfect system. It was time-consuming. It required training to implement and a severe eye for detail if measurements were to be made with complete accuracy. But it was a vast improvement. Bertillon had created some order out of chaos, and his new system became an established part of police procedure in France. Reading about it in Austria, Hans Gross at once began to adapt the system in his country.

  It became the general standard in much of the industrialized world, although some countries used different numbers of required measurements. Even England, never comfortable with French leadership, was considering bertillonage.

  The sour, introverted Bertillon was a success. He was given the title of director of the Police Identification Service. He was awarded a staff and a new office. He became world famous.

  By 1893, Conan Doyle, in his story “The Naval Treaty,” had Watson say, in describing traveling with Sherlock Holmes, “His conversation, I remember, was about the Bertillon system of measurements, and he expressed his enthusiastic admiration of the French savant.” Sherlock Holmes admired Bertillon! What greater accolade could be hoped for?

  But a tiny worm was wiggling within Bertillon’s apple. In 1880, while Bertillon was impatiently waiting for a chance to prove his theories, a Scottish medical missionary in Japan named Henry Faulds wrote a letter to the journal Nature. It was published on October 28 of that year and was titled “On the Skinfurrows of the Hand.” It was a very moderate statement of a concept that would cause a revolution in criminal investigation. Faulds’s letter began:

  In looking over some specimens of “prehistoric” pottery found in Japan I was led, about a year ago, to give some attention to the character of certain finger-marks which had been made on them while the clay was still soft. Unfortunately all of those which happened to come into my possession were too vague and ill-defined to be of much use, but a comparison of such fingertip impressions made in recent pottery led me to observe the characters of the skin-furrows in human fingers generally. From these I passed to the study of the fingertips of monkeys, and found at once that they presented very close analogies to those of human beings.

  “An ordinary botanical lens is of great service in bringing out these minor peculiarities,” Faulds wrote (reminding us inevitably of Sherlock Holmes and his magnifying lens). He continued:

  Where the loops occur the innermost lines may simply break off and end abruptly; they may end in self-returning loops, or, again, they may go on without breaks after turning round upon themselves. Some lines also join or branch like junctions in a railway map. All these varieties, however, may be compatible with the general impression of symmetry that the two hands give us when printed from.

  Faulds went on to explain how he took copies of these intriguing lines:

  A common
slate or smooth board of any kind, or a sheet of tin, spread over very thinly and evenly with printer’s ink, is all that is required. The parts of which impressions are desired are pressed down steadily and softly, and then are transferred to slightly damp paper. I have succeeded in making very delicate impressions on glass. They are somewhat faint indeed, but would be useful for demonstrations, as details are very well shown, even down to the minute pores. By using different colors of ink useful comparisons could be made of two patterns by superposition. These might be shown by magic lantern… . A little hot water and soap remove the ink… .

  I am sanguine that the careful study of these patterns may be useful in several ways.

  Faulds then discussed other possible uses of prints—for instance, in historical or anthropological research. And then he made this great leap into forensic history, writing:

  When bloody finger-marks or impressions on clay, glass, &c. exist, they may lead to the scientific identification of criminals. Already I have had experience in two such cases, and found useful evidence from these marks. In one case greasy finger-marks revealed who had been drinking some rectified spirit. The pattern was unique and fortunately I had previously obtained a copy of it. They agreed with microscopic fidelity… . Other cases might occur in medicolegal investigations, as when the hands only of some mutilated victim were found. If previously known they would be much more precise in value than the standard mole of the penny novelists. If unknown previously, heredity might enable an expert to determine the relatives with considerable probability in many cases, and with absolute precision in some.

  And then, making it clear that he knew the Tichborne case well, Faulds wrote, “Such a case as that of the Claimant even might not be beyond the range of this principle. There might be a recognizable Tichborne type.” Faulds was basing this statement on his observation that there are similarities in print patterns in some families, although hardly enough to provide definite identification. (However, if Lady Tichborne had preserved an unwashed object she was certain her son had handled from which clear prints could have been obtained, it could have served as an exemplar. If no prints on the object matched the Claimant’s, perhaps it might have influenced her opinion.) Faulds continued:

  I have heard, since coming to these general conclusions by original and patient experiment, that the Chinese criminals from early times have been made to give the impressions of their fingers, just as we make ours yield their photographs. I have not yet, however, succeeded in getting any precise or authenticated facts on that point… . There can be no doubt as to the advantage of having, besides their photographs, a nature-copy of the for-everunchangeable finger-furrows of important criminals.

  Henry Faulds was not the first person to have noticed ridge patterns on human fingers. As he mentioned, finger marks had been used in Asia. And in the West, as early as 1686, Marcello Malpighi, an anatomist at the University of Bologna, had mentioned the marks. In 1823, Professor John Evangelist Purkinje, a pathologist and physiologist at the University at Breslau, had presented a paper on finger impressions, listing nine types and suggesting a method of classifying them, but Faulds was the first to express the idea that finger marks could aid criminal investigation.

  Only a month after Faulds’s letter was published in Nature, Sir William Herschel, a British civil servant stationed in India, wrote to Nature that he had been using fingerprints for identification since 1860. There is no evidence that he foresaw the forensic use of prints, however.

  Sir Francis Galton, a cousin of Charles Darwin and a wellknown anthropologist, had never been convinced that bertillonage was the perfect solution to the problem of identification. Reading the reports in Nature, he quickly saw the advantage of fingerprints and devoted himself to studying the subject. He corresponded frequently and at length on the subject with Herschel (although he ignored Faulds’s contribution, infuriating that gentleman). Galton satisfied himself that just as no leaf is exactly duplicated and no two snowflakes are precisely alike, each fingerprint is unique. Further, he determined that an individual’s prints never change. He wrote:

  There appear to be no bodily characteristics other than deep scars and tattoo marks comparable in their persistence to these markings; at the same time they are out of all proportion more numerous than any other measurable features. The dimensions of the limbs and body alter in the course of growth and decay; the color, quantity and quality of the hair, the tint and quality of the skin, the number and set of the teeth, the expression of the features, the gestures, the handwriting, even the eye color change after many years. There seems no persistence in the visible parts of the body except in these minute and hitherto disregarded ridges.

  This, of course, was a direct contradiction of Bertillon’s concept, which held that most bodily measurements did not change.

  Galton was championing fingerprinting as the identification process that should be used in Britain. The one great remaining problem, as he saw it, was exactly how it was established for a presentation at court that two prints matched. Merely announcing that they looked alike wouldn’t do—hanging a man on the basis of such a vague statement would be feckless even for the enthusiastically punitive judiciary of the time.

  In his book Finger Prints, published in 1892, Galton presented the results of his research as well as a system he had devised for classifying the patterns of whorls, loops, and triangles found on fingertips. This method was to eventually form the basis for the identification process adopted by Scotland Yard.

  The same year, in Argentina, Juan Vucetich, the head of the Statistical Bureau of the La Plata police, was hard at work establishing his own method of print classification. It was put to the test quickly. Two children were found on their blood-soaked bed battered to death by a heavy object. Their distraught twenty-sixyear-old mother, Francisca Rojas, accused a neighbor of the crime. He had threatened to kill what she most loved if she would not succumb to his romantic overtures, she told the police. But the neighbor had an unshakable alibi.

  The investigators heard rumors that Francisca had a lover who had often been heard to say he would gladly marry her if only she weren’t saddled by her two children. The detective in charge, Alvarez, remembering Vucetich’s system, carefully combed every inch of the crime scene and finally came upon a small brownish spot on the door. It appeared to be dried blood, which held the print of a thumb. The piece of the door was taken to the station, as was Francisca. Her thumbprint was taken with the aid of a stamp pad. It was compared to the bloody print through a magnifying glass. It was clear they matched.

  Confronted with the evidence, the loving mother confessed that in the interest of romance she felt compelled to crush her children’s heads with a rock. She had then disposed of the weapon in a well. It was the first murder known to be solved by the use of fingerprints.

  (A hauntingly similar clue would appear years later in Conan Doyle’s story “The Adventure of the Norwood Builder” when Watson says, “With dramatic suddenness he struck a match, and by its light exposed a stain of blood upon the whitewashed wall. As he held the match nearer, I saw that it was more than a stain. It was the well-marked print of a thumb.”)

  In France, Bertillon cast a baleful eye on the new identification procedure. He saw it as a threat to his method and his prestige. He was resentful of Vucetich and Galton. He would tolerate the science of fingerprints, or dactyloscopy, as it was commonly called, only as an adjunct to bertillonage. Within this limit, he grudgingly allowed the Identification Bureau to record and file the fingerprints along with the anthropometric measurements of criminals. Perhaps it was Bertillon’s resistance to fingerprinting that kept him from establishing an efficient method of classifying fingerprints, and that led to an excruciating embarrassment in August 1911: the French nation awoke to the hideously momentous news that the Mona Lisa had been abducted from the Louvre.

  It had happened on a Monday, when the museum was closed to the public. The glass case that contained the wooden panel on which the
picture was painted stood open. Examination by the police located a fingerprint on the glass. If it appeared in the great Bertillon’s files, the culprit would be identified. France and the world held their collective breaths, but there was no match to be found. It was assumed that the daring thief had no record.

  Two years passed, and then an approach was made to an art dealer in Italy from someone who claimed to have the Mona Lisa. He turned out to be a house painter named Vincenzo Perugia, who had been keeping the most famous missing painting in the world under his bed since he had casually hidden it under his smock and taken it out of the Louvre. Perugia had an arrest record in France. His prints had to have been among Bertillon’s files. Why had they not been found?

  Bertillon was forced to explain that he had classified prints of the right thumb only, and Perugia, who seemed to have had a particularly poor sense of occasion, had left behind only a clear print of his left. The case left the stubborn Bertillon with even more resentment toward the fingerprinting process, but in spite of him, outside of France the new science had been growing in strength.

  Since 1893, Edward Richard Henry, the chief of police in Bengal, India, having read Galton’s book, had been using fingerprints in addition to Bertillon’s method of anthropometry. He grew to favor the use of prints and eventually, collaborating with Galton, developed the Galton-Henry system of fingerprint classification still used today in much of the English-speaking world.