The story broke in the papers on the very day of the murder, when the Evening Journal ran half a column on its second page. KILLED BY POISON SENT AS GIFT, read the headline. The article went on to relate the principal features of the case: Cornish’s receipt of the Tiffany box containing a bottle labeled “bromo-seltzer” along with a small “silver stand” Mrs. Adams’s agonizing death upon “taking a dose from the phial” Dr. Hitchcock’s fruitless attempt to revive her “by the Silvester artificial respiration process” Cornish’s own terrible illness after sampling the remedy. Hitchcock was quoted as saying that, in his opinion, “the powder in the phial was cyanide of potassium.” The article concluded by noting that “A peculiarity of the address on the package sent to Mr. Cornish was that the street name Forty-fifth Street was spelled ‘Fourty fifth Street.’”1
Shocking as it was, there was little in this story to differentiate it from a dozen other recent poison-murders. Indeed, just the day before, the Journal had printed a page-one article about a still-unknown culprit who had poisoned the pancakes at a Paterson, New Jersey, rooming house, “putting the lives of the entire household in jeopardy” in order to kill his intended victim, a boarder named George Naef.2 Certainly, the public had no reason to suspect that they were about to be treated to one of the most rabidly covered stories in the annals of American crime.
All of that changed on the following day, when, having sicced their newshounds on the story, the masters of the yellow press plastered their front pages with competing accounts of the crime. For William Randolph Hearst, the murder of Katherine Adams was a sensation-monger’s dream come true: New York City’s very own equivalent of the Cordelia Botkin case. The Journal’s headline on the morning of December 29 played up the similarity by comparing Mrs. Adams to Mrs. Botkin’s intended target: MRS. KATE ADAMS KILLED LIKE MRS. DUNNING, BY POISON SENT THROUGH THE MAILS.3
Pulitzer’s World, meanwhile, managed to best its bitter rival with a sensational scoop: DOUBLE MURDER MYSTERY IN A NEW YORK CLUB, screamed the headline. Crowing that it had “unearthed a chain of circumstances that may greatly assist the police in unraveling one of the most mysterious crimes of recent occurrence in New York City,” the paper detailed the “strangely similar circumstances” surrounding the death of Henry Barnet, another member of the Knickerbocker Athletic Club who had recently “received a bottle containing what appeared to be a harmless remedy.”4
There were significant discrepancies between the Hearst and Pulitzer accounts. According to the Journal, for example, Assistant DA John McIntyre believed that “the party suspected of sending the package to Mr. Cornish is at present out of…state.” The World,on the other hand, quoted McIntyre as stating that “clues…point to certain persons who are not far away.”5 Both papers agreed, however, that there was only one “direct clue” that might lead to a break in the case: “the handwriting on the package that…was received by Mr. Cornish through the mail at the Knickerbocker Athletic Club.”
Nowadays (at least if TV crime shows are to be believed), such a significant piece of evidence would be scanned with an electron microscope, studied with a fiber-optic fluorometer, tested for DNA, and checked against a national biometrics database, leading to the swift and sure identification of the culprit. In the late nineteenth century, however, forensic science was still in its infancy.
The world’s first criminological lab had been established only fifteen years earlier by an employee of the Parisian police department named Alphonse Bertillon, who devised a way of keeping track of criminals by collecting detailed data about their distinctive physical traits, from the span of their arms to the length of their left middle fingers. Bertillon’s “anthropometric” method, however, was good for only one purpose—identifying repeat offenders—and required little more in the way of equipment than a camera and tape measure.
Even a technique as basic today as fingerprinting (which would eventually render Bertillon’s cumbersome system obsolete) had not yet been widely adopted in the late 1800s, though its essential principles had been discovered decades earlier. In 1858, a British civil servant in India named William Herschel began insisting that the native businessmen he dealt with “sign” their contracts with inked handprints. Over the next two decades, as he collected thousands of these “hand-mark” specimens, he realized to his amazement that the ridged patterns on the ends of the fingertips always differed from person to person and could therefore be used (as he wrote in a letter) as an “infallible method of identification.”6
Simultaneously—and independently of Herschel—a Scottish physician and amateur archaeologist named Henry Faulds, while working in Tokyo, was struck by the distinctive “finger-tip impressions” left in pieces of ancient Japanese pottery. Taking up the study of the different “skin-furrows in human fingers,” Faulds quickly concluded that these markings were not only unique to each individual but remained unchanged throughout a person’s lifetime. In 1879, when a neighbor’s home was burglarized, Faulds was able to identify the culprit by comparing the suspect’s fingerprints to a sooty smudge left on a whitewashed wall at the crime scene. Recognizing the forensic importance of his discovery, Faulds published a letter in the British science journal Nature in which he declared that “bloody finger-marks or impressions on clay, glass etc.,…may lead to the scientific identification of criminals.”7
When Herschel learned of Faulds’s piece, he wrote his own letter to Nature, asserting his claim to priority. In the meantime—while the two men launched a bitter dispute that would last for decades—the famed British anthropologist Sir Francis Galton turned his own attention to the subject and, in 1892, published a landmark book that set forth the first practical system for classifying fingerprints.
It was a police official in Argentina named Juan Vucetich, however, who first demonstrated the full potential of fingerprinting for solving serious crimes. The Croatian-born Vucetich (né Ivan Vucetic) immigrated to Buenos Aires in 1884 at the age of twenty-six. Within a year, he had gone to work for the La Plata police department and quickly rose through the ranks. Assigned the task of setting up a criminal identification bureau in 1891, he became intrigued by the technique of fingerprinting after reading an article about Galton’s work in a French science magazine.
Armed with an ink pad and roller, Vucetich began collecting all the specimens he could, fingerprinting everyone from newly arrested prisoners to the mummies at the city museum. Before long, he had worked out a classification method that made small but significant improvements to Galton’s system, affording police an efficient way of matching crime scene prints with archived identification records. He called his technique dactyloscopy (Latin for “finger description”) and began to vigorously promote it in self-published pamphlets.
His superiors, however, remained skeptical—until fate supplied a startling vindication. In July 1892, a ghastly crime occurred in the coastal town of Necochea. A young boy and girl, the illegitimate children of a twenty-five-year-old woman named Francisca Rojas, were found murdered in their blood-drenched bed, their heads caved in. The hysterical mother blamed a man named Velazquez, a middle-aged, somewhat simpleminded laborer on a nearby ranch.
According to the story Rojas told the police, Velazquez was madly in love with her and had begged her to marry him. Rojas, however, was in love with someone else—a younger, far handsomer man. When she informed Velazquez of her feelings, he had flown into a rage and threatened to exact a terrible vengeance. A few days later, as she returned home from the marketplace, she saw to her surprise that her front door was open. As she drew closer, Velazquez burst from the house and fled. Inside, she found her two slaughtered babes.
Velazquez was immediately arrested and subjected to a merciless interrogation. When he refused to confess, he was bound with ropes, placed on the bed beside the mangled corpses of his presumed victims, and forced to remain there overnight. Even this torture, however, failed to produce an admission of guilt. Stymied, the Necochea police called in help from La Plata, and an inspector named Alvarez was dispatched to the village.
As it happened, Alvarez had taken a keen interest in the experiments that his colleague Vucetich had recently done with fingerprints. Examining the crime scene, the inspector discovered a bloody smudge on the children’s bedroom door and recognized it at once as a thumbprint. Sawing off the blood-marked piece of wood, he carried it to the local police station. He then brought in Francisca Rojas, took impressions of her thumbs with an ink pad, and examined them with a magnifying glass. Even to his untrained eye, it was clear that the markings of her right thumb matched the bloodstained print from the door.
Confronted with this evidence, Rojas immediately broke down. Her handsome young lover—so she tearfully confessed—had agreed to marry her only if she “got rid of the brats.” She had done the job with a large stone, then—after disposing of the murder weapon in a well—set about framing Velazquez.8
The 1892 trial of Francisca Rojas for the double murder of her children resulted in a legal landmark: the world’s first criminal conviction obtained from fingerprint evidence. Four years later, Argentina adopted Vucetich’s system as the country’s sole method of criminal identification. Other South American nations soon did the same.
Largely for reasons of cultural chauvinism, however, police agencies in Europe and North America were slow to follow suit. Dactyloscopy would not be fully accepted in the United States until 1911.9 This was unfortunate. In all likelihood, fingerprint evidence would have led to a speedy resolution of the Adams case. As it was, the New York City authorities would be forced to rely on what was, at that time, state-of-the-art forensic science in the United States: handwriting analysis.
Handwriting analysis had first played a prominent role in an American courtroom during the so-called Howland will case of 1867. Two years earlier, a fifty-nine-year-old spinster, Sylvia Ann Howland of New Bedford, Massachusetts, had died and left more than two million dollars to various beneficiaries. Hungry to get her hands on the entire fortune, her thirty-year-old niece, Hetty Robinson—a legendary miser, later nicknamed “the Witch of Wall Street”—contested the will. In support of her suit, she produced a document, supposedly signed by her aunt, which left the entire estate to Hetty.
The trial, one of the most dramatic of its time, hinged on the authenticity of that single signature. Among the many experts called to testify were engravers, bank tellers, penmanship teachers, and various eminent scientists, including Oliver Wendell Holmes, Sr., and—most critically—the Harvard mathematician Benjamin Peirce, who used statistical methods to show that the disputed signature was, in fact, a forgery.10 None of these men was a professional handwriting examiner, for the simple reason that such a profession did not yet exist in the United States. The next twenty years, however, witnessed the rise of handwriting analysis as a distinct, supposedly scientific, discipline with a growing number of practitioners offering their services as forensic specialists.
Within twenty-four hours of Mrs. Adams’s death, William Randolph Hearst had already commissioned an opinion from one of these experts. Prominently featured on page one of the December 29 Evening Journal was a box containing the conclusions of a noted graphologist named David S. Carvalho, who deduced that the sender was a man “past the age of thirty years” who had failed in his efforts to disguise his usual handwriting. Carvalho expressed the utmost confidence that “no difficulty will be encountered in locating the identity of this writing when brought into juxtaposition with known writing.”11
Carvalho’s profile of the sender was, in fact, more accurate than those of other self-professed “experts” who had weighed in on the matter and who confidently declared that the writer was a woman.12 Still, as Carvalho himself suggested, it was impossible to identify the sender from this sample alone, not without another piece of “known writing” to compare it to.
For that, Hearst enlisted the assistance of the public. Alongside Carvalho’s analysis, and running down almost the entire length of the front page, he printed a facsimile of the handwritten address from the poison package. WHO KNOWS THIS WRITING? blared the headline.
As it happened, someone did.