He spent time figuring out the precise routes that cyanide took in the body, at one point using four dogs to compare the effects of inhaling the poison versus ingesting it. No one would describe these experiments as pretty, but he justified them as pure necessity. Two dogs received measured doses of potassium cyanide through a stomach tube; the others were forced to inhale hydrogen cyanide. The latter two animals were strapped to an operating table with their jaws taped shut. A cone-shaped mask would be placed over a dog’s nose and mouth, taped in place, and sealed with Vaseline to make it airtight. Once the mask was fixed, hydrogen cyanide gas would be piped into the cone until the dog died.
The results provided some of the first measured evidence of how quickly cyanide kills, dispelling the myth that victims neatly drop dead on the spot. When Gettler gave a dog 50 milligrams of cyanide (a little less than 2 drops), the animal died in 21 minutes. When he cut the dose to 20 milligrams, the dog lived for 2 hours and 35 minutes. When inhaled, less of the poison was needed to kill, and it worked more quickly, but again not instantly. One gassed dog, for instance, breathed in about 10 milligrams of cyanide; he was dead in 15 minutes.
The animal studies—later confirmed in analyses of human cyanide victims—showed that the poison is absorbed differently depending on how it is taken. When it is inhaled, it blows through the body, concentrates in the lungs, and swims through the bloodstream into the brain, the heart, and the liver. When swallowed, the poison is absorbed much more slowly. Necropsies (or animal autopsies) of the two dogs that were fed cyanide found that much of the poison, between 38 and 83 percent, was still in the stomach when they died, which helped explain why they died more slowly than the dogs that had breathed in the poison.
Remembering the Jackson case, Gettler made a point of exploring what happened to cyanide in a decaying body. He took slices of livers and brains and lungs from the bodies of cyanide victims, noting the poison content of each. He placed those pieces of tissue into containers and then left them to rot on a shelf in his laboratory. He checked the results at one week, two weeks, three weeks, and four, looking far beyond the time that Fremont Jackson had been buried so that he could be absolutely sure of the conclusion.
“During this time putrefaction developed to a high degree,” he wrote with serious understatement concerning those month-old organ slices. He analyzed the decaying tissues for cyanide and compared the amount to the levels he’d measured when they were still fresh. He found that decomposition altered the poison readings by the barest amount. Even after four weeks, 90 percent of the original poison content could still be detected, once again validating his testimony in the 1922 case.
Gettler investigated another contention made in the Bradicich case, that even if cyanide was present in the old man’s body, it was meaningless. The defense experts had insisted that the body naturally produced its own cyanide as apart of decomposition. He was determined to set that straight as well. So he took samples from eight different organs, all taken from bodies of people who had died natural deaths, and sealed them in glass flasks. Every week for the following two months, he removed one for analysis.
During the first week, decay produced trace amounts of cyanide—about 0.03 milligrams per 100 grams of tissue—but after that the poison seemed to disintegrate. By the end of two months, cyanide could not be detected at all. At its strongest, though, it was a mere whisper, a fading breath in the test tube, nothing close to the levels he’d found in Fremont Jackson. “Putrefaction therefore should in no way interfere with deciding a cyanide poisoning case,” he concluded.
It had taken him years in a laboratory, silent with the emptiness of night hours, to get his answer to the Jackson case. If Gettler could have carried his findings back in time to that courtroom, he would have done so and perhaps changed the outcome. He found satisfaction instead in building a better science out of an unhappy episode. Next time, he promised himself, such legal sabotage would not succeed.
Still, it might not have particularly bothered Gettler that decades later, in 1980, the Hotel Margaret, long abandoned, would undergo renovations during which it accidentally burned to the ground. No one in Charles Norris’s office held fond memories of that shining ornament of a building. Those coppery roofs and elegant balconies, those devious employees and that seeping poisonous gas; all of it reminded them only of mistakes that they intended not to make again.
FOUR
ARSENIC (As) 1922—1923
THE WEATHER in that summer of 1922 held steady at what the newspapers like to call “fair,” the skies a gas-flame blue, the temperatures hovering near 80 degrees. On the last day of July, as Lillian Goetz’s mother would forever recall, the morning was another warm one. She offered to make seventeen-year-old Lillian a box lunch, but the girl refused. It was too hot to eat much; she’d just grab a quick sandwich at a lunch counter, she said.
Lillian worked as a stenographer in a dress goods firm occupying a small set of offices in the Townsend Building, at the bustling corner of 25th and Broadway. There were plenty of quick eateries nearby, tucked among the offices and shops and small hotels. Lillian, like many of her co-workers, often stepped over to the Shelbourne Restaurant and Bakery, just half a block south on Broadway.
The Shelbourne catered to the office trade, opening in the morning, closing in the early afternoon. Stenographers and secretaries in their bright summer hats and stylish short skirts, businessmen and office managers in their dark tailored suits, crowded daily along its wooden counters and small square tables, hurrying through a meal of hot soup with fresh-baked rolls, a sandwich, coffee, and a slice of the bakery’s renowned peach cake or berry pie.
According to police reports, on July 31 Lillian ordered a tongue sandwich, coffee, and a slice of huckleberry pie. It was the pie that killed her.
BY EARLY AFTERNOON sixty people had been rushed to nearby hospitals after eating lunch at the Shelbourne, and by the end of the day, six of them, including Lillian Goetz, were dead. The scream of ambulances on lower Broadway was so constant that a number of people called the police in a panic, fearing that the whole city had caught fire.
The Townsend Building, where Goetz worked, was an 1896 neoclassical structure that normally conveyed a stately limestone sense of calm. Now it served as backdrop to a scene of hysteria. Office workers collapsed on every one of the twelve floors, convulsing, vomiting, gasping in misery. Doctors armed with stomach pumps—at least ten pumps were put to use throughout the building—hurried from floor to floor, crisis to crisis. In the excited words of the city newspapers, “Panic prevailed on some floors in the Townsend building as one employee after another turned pale, and then blue and began to complain of intense pain.”
Gradually the doctors began comparing symptoms, notes, and stories. It led them to realize two things: that every victim had lunched at the Shelbourne Restaurant, and that almost all had eaten either blackberry or huckleberry pie for dessert. The physicians called the health department and the medical examiner’s office to report their suspicions.
The following day Charles Norris and Frank Monaghan, the acting health commissioner, made a joint announcement. Arsenic had been found in the piecrusts and rolls served at the Shelbourne. Additional tests showed that none of the ingredients—flour, butter, salt—stored at the restaurant contained any poison. Therefore the investigators suspected that the arsenic had been added to the dough after it was mixed, perhaps into the covered dough bowl stored in the kitchen refrigerator.
In other words, Norris and Monaghan agreed, this was not a matter of a kitchen accident, a baker using flour from grain tainted by an arsenic pesticide. Someone had planned this: “the food had been poisoned with malicious intent.”
Knowing the poison is never the same as knowing the killer. The police wished it were. They had no answer as to who might have done this. No answer as to why anyone would wish to harm a seventeen-year-old stenographer, working to help out her family, whose mother repeatedly told police that she only, only wished she h
ad made that box lunch.
THE PREVIOUS October, in an unnervingly similar incident, two lunch patrons had been killed by arsenic at a restaurant down in the financial district. At that little eatery, near the old Liberty Street post office, health inspectors had at first suspected food poisoning. But then Alexander Gettler had isolated lethal amounts of poison in both men’s bodies.
The police had not identified a suspect in the killings at the Postal Lunch eatery, had never even come close. Maybe that was why people quit going to the eatery, which was now closed down. One of the first fears expressed by the police department was that the same poisoner had now moved up to Broadway; that this killer just enjoyed causing death, someone—detectives speculated—like the still-infamous Jean Crones.
Years earlier Crones had worked as an assistant chef at the exclusive University Club in downtown Chicago. In his spare time he belonged to the city’s thriving community of anarchists, an outspoken opponent of what he considered the government oppression ruining the country. But his political views never explained—neither to his kitchen colleagues nor to his political allies—why on February 10, 1916, while making sauces for a dinner honoring Catholic bishops, he added a liberal seasoning of arsenic to the meat stock.
Seventy-five of the three hundred people at the dinner became rapidly, horribly sick. In the ensuing maelstrom of shouting emergency workers and terrified diners, Crones simply walked out of the club and caught a train to the East Coast. He was never apprehended, although he briefly stopped in New York to mail police-baiting letters to the Manhattan newspapers. His mocking notes suggested that the incompetent Chicago police should take lessons in detection, perhaps enrolling in a correspondence school, although he doubted they would since “the city of Chicago officials are fools.”
Investigators came to believe that Crones had simply stirred in poison as an experiment, for the pure pleasure of seeing the results. They hoped that this wasn’t true of their New York poisoner, who seemed more adept than Crones. The Chicago chef had loaded far too much arsenic into the soup, causing many people to put down their spoons at the metallic taste, preventing them from falling ill. Those who kept eating became sick quickly, but doctors who rushed to the scene were able to administer rapid treatment. As a result, no one had died in Crones’s experiment.
At the Shelbourne, as at the Postal Lunch, the poisoner had calculated a much more effective dose, high enough to kill but low enough to fool the victims. The lunch patrons at the Broadway eatery had had time to return to work before they became ill. Their illness attracted medical attention but not quickly enough to save everyone. If the unidentified Shelbourne suspect and the never-caught Postal Lunch killer were the same person, then police were looking for a killer who knew all too well how to pick a poison, how to use it, and how to disappear.
PURE ARSENIC is a dark, grayish element, classed among the heavy metal poisons, often found in ores extracted from mines. It easily combines with other naturally occurring chemicals; heated with oxygen, for instance, it becomes a white, crumbly powder, the linking of two arsenic atoms with three of oxygen. In this form it is called arsenic trioxide (As2O3) or white arsenic.
White arsenic, the poison used at the Shelbourne, was a favorite of some of history’s most feared poisoners, ones who made Jean Crones look like the amateur he was. At the top of the list were Lucretia and Cesare Borgia, feared in fifteenth-century Italy for their ruthless mixture of politics and poison. The Borgias used white arsenic preferentially but experimented with different ways to make it more deadly. They would cook it into a more intense solution, mixing it with other poisons. They eventually created a poison they called la cantarella, which according to legend was so dangerous that the formula was destroyed after their deaths.
Basic arsenic is also deadly; the first recorded case of homicide with the pure element was reported in 1740, when a girl poisoned her father and three sisters by serving them a dish of dried pears that had been boiled in water containing rocky ore from a nearby mine. But for criminal uses, white arsenic is a better tool, slipped easily into food or drink. Its usefully murderous properties explained why, centuries after the Borgias, the poison earned another nickname: the inheritance powder. One of the best-known nineteenth-century American forensic scientists, Columbia University chemistry professor Rudolph Witthaus, coauthor of the massive 1896 tome Medical Jurisprudence, Forensic Medicine and Toxicology, had once tried to estimate arsenic’s popularity as a murder weapon. He selected 820 arsenic-caused deaths, recorded between 1752 and 1889—and found that almost half were homicides. (The rest were split fairly evenly between accidents and suicides.)
In Europe, by Witthaus’s analysis, arsenic accounted for the largest percentage of nineteenth-century criminal poisonings. In France, for example, between 1835 and 1880, arsenic was used in almost 40 percent of all poison murders. “In the United States, we are under the impression that arsenic still holds the first place in frequency of criminal administration,” Witthaus wrote. But at the time of his analysis, the United States lacked the statistical information available in France. As a rough measure, Witthaus interviewed New York State attorneys and determined that from 1879 to 1889, there had been thirty-one indictments for poison murder in twelve counties. Half of those were arsenic murders; in every case, white arsenic was specifically to blame.
A primary reason for arsenic’s popularity was that when mixed into food and drink, it is extremely difficult to taste. An over-the-top dose, the kind Jean Crones had used, was different. If arsenic was swallowed in undiluted form, that was different too. Witthaus and other scientists taste-tested small amounts of pure white arsenic and found it to be rather nasty. It was hot, tasters said; it was acrid, sweetish, metallic, and rough. But when the poison was added to soup, liquor, or a cup of hot coffee, the other flavorings easily masked it. Arsenic was “under the most favorable circumstances faint” in taste, Witthaus noted. In interviews with 822 people who had survived arsenic poisoning, he reported that only fifteen thought the food had tasted in any way strange. Six talked of a bitter taste, eight complained of a metallic feeling in the mouth, and one woman said she was aware of a “nauseous” taste. Witthaus doubted the latter, noting that as the unpleasant taste “escaped the notice of 14 other persons who ate of the same poisoned pudding, it was probably more imaginary than real.”
White arsenic mixes especially beautifully into alcoholic drinks, which tend to hide even a faint metallic sensation. One group of cheerful drinkers had shared a bottle of port wine and suffered from fairly severe arsenic poisoning, but “not the least taste was perceived by any of the parties.” Sometimes people complained of a sandy feel in their mouths; they seemed unusually sensitive to the rough texture of the powder, even mixed into food. In an 1860 New Jersey case, a man had murdered his wife by rubbing white arsenic into an apple. During the trial he acknowledged that “she said there was something gritty on it,” but she’d eaten anyway, thinking he just hadn’t washed it well. Mostly, though, ground extra-fine and mixed into baked goods, as at the Shelbourne, arsenic proved an almost undetectable ingredient.
In fact, handled with skill by a calculating murderer, the poison seemed to engender a homicidal overconfidence. In 1872, one notorious British murderer, Mary Ann Cotton, killed fifteen people, including all the children of her five husbands, and several neighbors who irritated her, before she was caught in 1872, tried, and hanged. “Arsenic has also been,” Witthaus wrote, “in almost every instance, the agent used by those who, having succeeded in a first attempt at secret poisoning, have seemed to develop a lust for murder and have continued to add to their list of victims until their very number has aroused suspicion and led to detection.”
AT THE TIME of the Shelbourne killings, scientists were still not sure how arsenic killed; the action of cyanide was far better understood. Not for decades would molecular biologists work out the method by which arsenic targets key enzymes, disrupting metabolism within cells throughout the body, breaking th
e system down cell by poisoned cell. Part of the problem for early toxicology researchers was that arsenic, as one complained, is a great mimic.
Physicians often mistook symptoms of arsenic poisoning for natural diseases, especially if the victim was dosed gradually. General practitioners and emergency room doctors had misdiagnosed arsenic deaths as influenza; as cholera, which also causes severe gastrointestinal distress; and as heart disease, which also causes gasping shortness of breath. Such errors were found only when a suspicious relative who was sure the victim had been healthy, and distrusted a husband or wife, demanded an autopsy.
From the poisoner’s perspective, arsenic became a risky choice when doctors looked directly at the body. With the stubborn, solid constitution of any metallic element, it breaks down extremely slowly compared to organic poisons and can be detected decades after death in a victim’s hair and fingernails. Even worse for those who hope to avoid detection, arsenic tends to slow down the natural decomposition of human tissue, often creating eerily well-preserved corpses. Toxicologists refer to this effect as arsenic “mummification.” Witthaus reported that one body, exhumed after fifty-four weeks in the ground, “did not differ from a living person” in appearance except for the patches of mold growing on his face. “The growth of molds is not interfered with by arsenic,” he added austerely.