Blood, Bullets, and Bones Read online

  At least in the public view, Spilsbury was no longer infallible. And the fact that two teams of pathologists disagreed on the cause of death showed that forensic science itself was fallible. Answers weren’t definite. They were up for interpretation and subject to the limitations of the pathologists’ knowledge and methods. And those pathologists were only human.

  Meanwhile, in America, a system of conducting scientific autopsies was just taking hold. The English coroner system had been brought to America in the 1600s—before Dr. Wakley’s campaign to require coroners be doctors, and so well into the twentieth century in America, most coroners were appointed by the mayor and had no scientific training. Coroner and jury would make a cursory examination of the body and haphazardly declare the cause of death. So if, for instance, someone died of a stomach ailment, the coroner, rather than testing the stomach for poison, might assume that the victim had fallen ill. Doctors were sometimes called in for further examination, but autopsies were rare. There was no protocol for when they should be conducted, and victims’ families were often against them, believing them to be desecration of the dead. Those autopsies that were performed were nowhere near as thorough as today’s, and much less helpful in solving crimes.

  By the early 1900s, coroners were falling out of favor in the nation’s largest city. The New York Times reported that coroners were accepting bribes to lie about official causes of death. That meant that if a loved one committed suicide, the family could pay the coroner to say it was an accident. Coroners were also bilking the city of money. They were paid by the body and so would report more deaths than there really were. In one case, several bodies were found on separate occasions floating in the East River. None were brought to the morgue for autopsy, and it turned out there was only one body, which the coroner was moving to new spots along the river. In this way, he collected $10,000.

  Beyond this shady behavior, most coroners simply didn’t know what they were doing. This was displayed in the 1900 murder trial of the lawyer Albert Patrick. He was accused of hiring a butler to kill his wealthy client, eighty-four-year-old William Marsh Rice. The coroner’s surgeon first testified that the death was natural, but later changed the cause of death to chloroform poisoning. Albert was convicted, but in a later chloroform murder case, the lungs of the victim were found to be completely different from William’s lungs at the time of his death. Five hundred physicians and other experts signed a letter urging a new trial. Albert was finally released in 1912.

  It was clear that the New York coroner system was ineffective. Massachusetts had developed a medical examiner’s office in 1877, but it hadn’t taken hold elsewhere. In 1915, New York officials decided to bring the Massachusetts system to New York. Needless to say, coroners were against the idea. It would mean they’d be out of jobs. They even made the outrageous claim that the medical examiner system was a ploy to funnel more organs into medical schools (because the city would be conducting more autopsies).

  Charles Norris, New York’s first chief medical examiner

  Nevertheless, Charles Norris, who had studied cutting-edge forensic pathology in Europe, became the city’s first chief medical examiner in 1918. Under the new system, all unexplained deaths were referred to the medical examiner’s office. Then, if an autopsy was needed, the body was taken to one of the five morgues, in Manhattan, Staten Island, Queens, Brooklyn, or the Bronx. The city’s autopsies did, in fact, increase—from about 8 to 20 percent of all bodies examined, for a total of 1,400 autopsies in 1918. And that meant that murders were detected that might otherwise have gone unnoticed.

  Many of these murders involved poison. As in Europe, poison deaths in America were rampant. It may be hard to believe today, when poisonings are so rare and shootings so common, but in 1922, New York had 997 poisonings, compared to 237 shootings. At the time, poison was ubiquitous. It was found in makeups and over-the-counter medicines. There were no warning labels or safety caps, so sometimes the poison was taken by mistake. But it was also used as a murder weapon.

  For this reason, Norris hired chemist Alexander Gettler to set up a toxicology lab—the first of its kind in America. In cases of possible poisoning, Gettler would examine the body and decide which tests to run. For instance, if the lips or skin appeared blue, it indicated possible cyanide poisoning, because that poison affects the body’s ability to process oxygen. Victims of cyanide poisoning also exude an odor of bitter almonds. That’s because the nuts, along with many other plants, contain cyanide—a defense mechanism that discourages insects from eating them. (Sweet almonds—the kind we eat as snacks—do not.) To find poison in the body, Gettler ground the stomach wall, distilled the sludge, and tested it for the given poison. Gettler didn’t test just bodies of possible victims for poisons; he also ran tests on other corpses to expand his knowledge. For instance, he tested bodies that hadn’t been poisoned to see how much poison they contained naturally; the human body contains trace amounts of poisons, and poisons can also be absorbed from the soil when the body is buried.

  Alexander Gettler, toxicologist

  Gettler became such an expert on poisons that it was difficult for defense attorneys to dispute his findings in court. Gone were the days of chemists being undermined as expert witnesses, as in the clam chowder case or the Hyde murder appeal. Not only was it getting harder to pass a poisoning off as a natural death, it was also harder for a guilty party to get an acquittal if the case went to trial.

  In a 1923 case, eighteen-year-old Charles Avery died of an apparent illness. But the police received a tip from neighbors to investigate Charles’s sister, Fannie Creighton. Charles had recently moved to Newark, New Jersey, to live with Fannie, her husband, and their child. The police exhumed the body, and the medical examiner’s office found it to have high levels of arsenic. The police learned that Fannie had taken out a life insurance policy on Charles without ever telling him. They also found in her home Fowler’s Solution, a cosmetic product that contained arsenic. She and her husband were brought to trial. Their defense attorney didn’t argue as to whether arsenic had been found in the body. But he said that Fannie’s Fowler’s Solution contained so little arsenic that it was unlikely to have killed anyone. It was more likely that Charles had gotten rat poison (which had a higher concentration of arsenic) from the store where he worked. Why? Using the old Styrian defense, the attorney suggested that Charles might have been an arsenic eater. Alternatively, the attorney said, Charles might have killed himself because he was depressed over an unrequited crush. Based on these possibilities, Fannie was acquitted.

  But neighbors had also tipped off the police about the death of Fannie’s in-laws, who had lived with her husband and her until their mysterious deaths. This was an Essex County case—out of Gettler’s jurisdiction. The prosecution’s experts found arsenic in the mother-in-law’s body, and Fannie was brought to trial for a second time. Gettler was asked by Fannie’s defense attorneys to conduct his own examination of the body. He, too, found arsenic. But he had a hunch that the young mother was innocent. What if, he wondered, the chemical he had found wasn’t pure arsenic? Fannie’s mother-in-law took several medications, one of which contained bismuth. Gettler knew that bismuth melted at a higher temperature than arsenic. When Gettler heated the arsenic to its melting point, only a small amount melted. The rest remained. That meant that what the prosecutors thought was arsenic was mostly bismuth. The amount of arsenic in the body was minute—not enough to have killed the woman. In other words, she hadn’t been poisoned, she’d simply taken a medicine that had traces of arsenic in it (common in medications at that time). Gettler convinced jurors that the mother-in-law couldn’t have been poisoned, and Fannie was found not guilty. Little did Gettler know that though Fannie was not guilty in this case, she was, in fact a murderer.

  In 1935, Ada Applegate died on Long Island. A physician ruled her death a heart attack. But police soon received an anonymous package containing newspaper clips from Fannie Creighton’s earlier murder trials. Ada and her hu
sband had recently moved in with the Creighton family in order to split their rent (a common practice during the Great Depression). Having no toxicologist of their own, the Long Island police sent the body to Gettler. He found it riddled with arsenic. Fannie became a suspect, and though she didn’t admit to this murder, she told police that she had killed her brother all those years ago. For weeks, she’d served her brother poisoned chocolate pudding in order to collect on his $1,000 life insurance policy. And she did indeed use her Fowler’s Solution to do the job. She’d already been found not guilty in the case, and a person acquitted of a crime cannot be prosecuted again, due to a constitutional clause known as double jeopardy. So she had safely gotten away with that murder.

  The police were forced to focus on the present case alone. They learned that Ada Applegate’s husband, Everett, had a romantic interest in Fannie’s fifteen-year-old daughter, Ruth, and that Fannie strangely had encouraged his pursuit. In fact, Fannie hoped that the two would marry so that she would no longer have to support her daughter. Prosecutors argued that Fannie had killed Ada so that the marriage could go forward. They charged Everett with the statutory rape of Ruth and the murder of his wife. Fannie and Everett stood trial, and this time Gettler testified against Fannie. He said that Ada’s body contained four times the lethal dose of arsenic and that it was the same type found in the rat poison Fannie had recently purchased. Again, the jury believed Gettler. Fannie and Everett were found guilty and sent to the electric chair.

  Fannie Creighton leaving court after being sentenced to death in the electric chair

  In other cases, Norris and Gettler were able to prove the innocence of suspects police thought were guilty. One November morning in 1926, a police officer saw a man kicking a bundle into New York Harbor. When the officer questioned the man, he gave a fake name and address. Later, a taxi driver pulled up, asking police why they were talking to Francesco Travia. So that was his real name! The cabdriver also knew his address: 56 Sackett Street—by the loading docks on the harbor. At Francesco’s home, police found a bloody scene: a woman’s torso with her severed head lying beside it. It belonged to Anna Fredericksen, a neighbor of Francesco’s.

  According to the memoirs of Dr. M. Edward Marten, the New York City medical examiner in Brooklyn at the time, Francesco readily confessed to the murder: “Yeah yeah! I killed her. I t’rew her arms and legs in de river. I was goin’ to t’row de rest of her in tonight.”9 Case closed. Or so police thought.

  Then Marten, along with Norris and his chauffeur, Charlie (last name unknown), arrived on the scene. Norris told police he bet they had no cause of death. Police said that of course they had a cause of death; the woman’s head had been cut off. Marten said it was Charlie who guessed at the real story. (During his time with Norris, he’d become an amateur sleuth.) He pointed out that the murdered woman’s face was cherry red, indicating carbon monoxide poisoning. Lab work indeed showed carbon monoxide—and not the gruesome dissection—to be the cause of death. Her brain also showed signs of intoxication.

  Francesco explained what had really happened. He and Anna had been drinking together. The night grew cold, and they shut the windows and lit the stove. At six a.m. he awoke to find Anna dead. Thinking he’d killed her in a state of drunkenness, he tried to hide the evidence by chopping up the body and throwing it in the river.

  Even with the medical examiner shedding light on the real cause of death, Travia still stood trial for murder. But his lawyer was able to prove that Anna did die from carbon monoxide poisoning. Some coffee had boiled over on the stove, putting out the flame and allowing gas to fill the apartment. Travia escaped the death penalty for murder but was jailed for mutilating a dead body.

  In spite of Norris’s keen mind and Gettler’s cutting-edge poison detection, some cases still went unsolved. It was one thing to detect poison in a body and another to find the poisoner, especially in a case where there was no clear motive. Seventeen-year-old Lillian Goetz lived at home and worked as a stenographer for a fabric company. On the morning of July 31, 1922, her mother offered to make her a lunch to take to work. “No, mother,” Lillian said. “It’s too hot to take a heavy lunch with me today. I’ll just step out and get a light sandwich and some pie.”10

  People often recall, after a tragedy, how a simple decision saved them from the fates of the victims. In this case, the opposite was true. Lillian did grab a sandwich and slice of pie on her lunch break—at the Shelburne Restaurant on Broadway between Twenty-Fifth and Twenty-Sixth Streets. On returning to work, she became sick to her stomach and was sent home in a taxi. She wasn’t the only one. Ambulances were arriving all over the neighborhood as more than fifty pie eaters fell ill.

  Earlier in the day, restaurant patrons had complained that the huckleberry and blackberry pies had burned their throats. Hearing this, the owner of the Shelburne, Samuel Drexler, asked his brother-in-law to try a piece. He did and assured Drexler that the pie was fine. Later, the brother-in-law suffered from intense stomach pain, and a physician pumped his stomach, saving his life. Once Drexler saw that the pies really were making people sick, he ordered that no more be sold. He also sent samples to a chemist, who said there was arsenic in the crust. It wasn’t just the huckleberry and blueberry pies that were poisoned, but all the pies and other pastries, too.

  Though many patrons were poisoned, few actually died. Norris, the medical examiner in the case, said that the high dose of poison in the pie had caused the pie eaters to violently vomit, purging the poison from their systems so that they survived. Lillian was not so lucky. Though she seemed to be recovering that evening, by four a.m. she had again become ill. A physician was unable to save her, and she died with her mother by her side, one of six murder victims.

  The investigation began. Accidental poisoning was ruled out. The owners didn’t keep rat poison on hand, and the pie ingredients were tested and found to be pure. While all employees had access to the refrigerator where the dough was kept, and even an outsider could have grabbed it, that wasn’t likely. The arsenic was found throughout the dough, suggesting that it had been mixed well, probably in the bakery.

  Louis Mandel was the head baker on Monday, his first day back to work since he’d left the restaurant several weeks earlier to start his own business. When that didn’t pan out, he returned to his old position. He told investigators that when he arrived in the morning, there were five pounds of dough in the refrigerator left over from Saturday. His assistant, Louis Freedman, mixed two more pounds in order to have seven pounds total for the day’s baking. The Monday dough was analyzed and found free of poison. It was the Saturday dough that was tainted.

  Saturday had been the last day of work for the previous baker, Charles Abramson. He had given his notice a few days earlier, fearing that he would soon be laid off when Louis M. was given his old job back. Charles said that Louis F.—his assistant, too—had mixed the dough, but Louis F. denied this. Either had the opportunity. But who had the motive?

  The impending layoff seemed a possible motive for Charles. He told investigators that on Saturday, he had been feeling bitter toward his boss. Then, when Charles returned to the restaurant the morning of the poisoned pie incident, Drexler said he’d never planned on firing Charles in the first place. So by Monday, Charles was no longer mad. Of course, the dough had been mixed on Saturday. But after questioning Charles, district attorney Joab Banton was convinced of his innocence. Things hadn’t turned out badly for Charles. He’d already found a new job that paid better than the Shelburne.

  Both Charles and Louis F. were suspects, but Banton repeatedly told reporters that there wasn’t enough evidence to arrest either man. Moreover, neither had a strong motive to poison the customers. In fact, no one did. It seemed likely that the killing was about bloodlust—the desire to kill people. In 1916, there had been a mass poisoning at a Chicago banquet for the archbishop. Several diners became violently ill, and the soup was found to be laced with arsenic. Police searched the apartment of Jean Crones, a cook
, and found poison there. But Crones himself was never found. In the poisoned pie case, police had no such clues that led to Charles, Louis F., or anyone else.

  Killings by strangers are difficult to solve because the police ordinarily rely on leads generated by interviewing the victims’ friends, family, and acquaintances. They also consider motive, narrowing the pool of suspects. If the killer didn’t know his or her victims and had no reason to kill them, then both trails are dead ends. As Banton told reporters: “This is a most baffling case, and may require a Sherlock Holmes to solve it, although on the other hand, it may have a simple explanation.”11 It turned out the first option was true: the case was never solved.

  Thankfully, the era of rampant poisonings was coming to a close. Authorities had clamped down on the sale of products containing poison. Medical examiners were better able to detect poisoning. And juries were finding expert testimony against poisoners to be more and more reliable. However, the decrease in poisonings didn’t mean there were fewer murders. As the population in New York grew, so did the number of bodies examined by medical examiners. By the 1930s, each borough morgue was handling about eight autopsies a day.

  One of those bodies came in piece by piece. In 1931, a salesman driving across a bridge saw what he thought was a ham lying in the road. He pulled closer to get a better look. It was a human thigh. He called over a policeman, who confirmed the gruesome find. The thigh was brought to the Brooklyn morgue, where Dr. Marten examined it. He found on it a clean cut through the hip and knee joints and theorized that it might have been severed by a doctor. Perhaps it was a cadaver limb placed on the bridge as a medical-student prank. But Gettler did a toxicology report and found neither the preservative used on cadavers nor any of the anesthetics used during operations.

  Marten found it extremely unlikely that in that day and age an amputation would be performed without anesthesia and that the amputated limb would somehow wind up on a bridge. It was also unlikely that someone would cut a thigh off a dead body after a natural death, and again, that the thigh would wind up on a bridge. It was far more likely that the thigh belonged to a murder victim, but who?