In late winter 1828 in the small town of Northumberland, Pennsylvania, William Logan fell ill and died. The local gossip that ensued raised suspicions about his wife’s virtue, the nature of her relationship with a neighboring gentleman, and her purchase of rat poison. Mrs. William Logan was arrested for poisoning her late husband and sent to prison on the basis of a report by four physicians. They conducted chemical examinations of the deceased’s stomach and its contents that seemed to confirm the presence of arsenic.

Ten days before Logan’s death he went out on a very cold day and returned home drunk. He soon came down with a fever and pain in his head, neck, and limbs. Later he developed a cold and cough, and his family physician, Samuel Jackson, was called to the house. Jackson treated him by bleeding, provided firewood to warm the house, and administered small doses of emetic tartar (a commonly used expectorant). A vein in one of Logan’s arms became inflamed, however, and Logan soon became delirious, fainted, and died.

Shortly thereafter Jackson’s colleague M. Aristide Rodrigue arrived to help Jackson examine Logan’s body to determine the cause of death. While Jackson comforted the widow, Rodrigue dissected out the inflamed vein. Jackson later wrote that Rodrigue’s dissection revealed “the most perfect specimen of intense inflammation we had ever seen.” In the doctors’ view, Logan had died of natural causes.

A few days after Logan’s burial, according to Jackson’s account, “popular clamour” intervened—suspicions arose because Logan’s wife had recently bought arsenic at an apothecary, ostensibly to keep rats and mice from her butter. The coroner deposed individuals whose testimony, in Jackson’s opinion, included hasty assumptions and “trifling or irrelevant” information that “when properly understood went rather to clear than to convict the woman.” Logan himself had talked about poisoning rats; moreover, Jackson saw no motive “for so hideous a crime.” He wrote: “There was no hope . . . of her being bettered by his death.” Nevertheless, Logan’s body was disinterred and examined by a board of four physicians. Despite Jackson’s age, experience, and relationship with Logan, he was excluded from the board’s examination. After only two days, the board reported that Logan had died of arsenic poisoning.

The Risk of Hubris

Jackson (1788–1869) was one of the first physicians in Northumberland, a town famous as the American home of British theologian and natural philosopher Joseph Priestley from 1794 to 1804. Jackson had arrived in 1813, a year after graduating from the Medical Department at the University of Pennsylvania, 160 miles away in Philadelphia. When the appointed board of physicians issued their report with the findings that led to Mrs. Logan’s arrest, Jackson reanalyzed the circumstances. His inquiry suggested that the community’s moral prejudgment of the accused had probably influenced the initial court-appointed scientific investigation. Jackson’s work eventually led to Mrs. Logan’s acquittal, and shortly thereafter, in 1829, Jackson published his findings in an article entitled “Case of Supposed Poisoning with Arsenic” in the American Journal of Medical Science. The article cataloged the numerous errors and weaknesses in the board of physicians’ report as well as the dangers of overconfidence and carelessness on the part of scientists involved in criminal proceedings. Such accusations of hubris in criminal investigations were to become a hallmark of scientific debate during this period.

After the board collected its findings, Jackson reviewed the minutes of the proceedings. The physicians described an inflamed stomach, a common indicator of arsenic poisoning. But Jackson immediately saw contradictions among the physicians’ respective descriptions. He later wrote: “As to the supposed inflammation, it appears to be a mere matter of opinion whether any existed.” Jackson thought the condition of Logan’s stomach could be accounted for by his drinking “country whiskey for many years.” This explanation was “more reasonable than to suppose an acute inflammation by arsenic, without puking or any mode of distress.” Jackson supported his argument by quoting British surgeon John Shaw, author of the influential Manual of Anatomy (1822). Shaw had written: “I have come to the conclusion that the appearance of the stomach . . . alone, in a question of poison, is not to be depended on.” Adding an ethical component to his critique, Jackson also quoted Shaw’s hope that “this degree of uncertainty will prevent the anatomist from being called on to decide a question which may involve the life of a fellow creature.”

In his article Jackson detailed the ambiguous results of the various liquid tests performed by the board of physicians, noting that the comparison of “colours is extremely difficult” and that “even experienced eyes may be deceived.” The board had used four tests to detect arsenic in Logan’s body. First, the board performed a “sulphus cupri” test, boiling fluid from Logan’s stomach with a few grains of the sub-carbonate of potash; the fluid turned light green when copper salt was added. In the second test, the physicians placed a sample of stomach fluid on white paper and drew a stick of silver nitrate over it, causing the spot of fluid to turn pale yellow. For their third test, the board placed a small quantity of the dried stomach between two plates of copper coated with black flux (a paste of potassium salts of boron and fluorine); the plates were then heated to produce a silvery white stain. In their fourth test, the physicians passed a stream of sulphuretted hydrogen through a sample of stomach fluid that had been boiled and cooled; this process caused the sample to turn yellow. While the board concluded that the four results revealed the presence of arsenic in Logan’s body, Jackson questioned the board’s application of the tests. He noted that the “sulphus cupri” test could detect arsenic but also “other matters which are sometimes found in the stomach”; that “the copper and silver tests were used in the most objectionable forms”; and that the sulphuretted hydrogen test might have allowed tartar emetic, which Jackson had administered to Logan before his death, to be mistaken for arsenic.

Jackson argued that the physicians’ failures in chemical experimentation meant that there was no proof of the presence of arsenic. “They leave us destitute of all positive proof, and greatly debilitate the circumstantial; therefore, since such strong suspicions arose in their minds, it is greatly to be regretted that they did not proceed further with the enquiry.” Moreover, Jackson continued, some “important leading tests were omitted”—no microscope was used, although “a very powerful one was within their reach,” and no drawings of the later destroyed stomach were made, although “some excellent delineators were at hand.” How could all these misjudgments be explained?

The answer, for Jackson, was hubris. He did not blame the physicians for jumping to conclusions or for performing half-hearted chemical experiments. Instead he blamed the hubris of John Ayrton Paris, a well-known Edinburgh physician who was the subject of harsh criticism from other scientists as well as Jackson. Paris’s pharmacological treatise, Medical Jurisprudence (London, 1823), coauthored with lawyer John Samuel Martin Fonblanque, was the main resource on which the board of Logan’s examiners had relied.

The physicians believed, as Paris promised, that the copper and silver tests were infallible. This belief, Jackson argued, was at the root of their folly.

For Jackson the colors from all the board’s tests provided “one degree of evidence only.” More conclusive evidence could have been produced by experimenting on a metal extracted from the precipate—if there was enough precipate to undergo metallization. Without the results of such an experiment, Jackson argued, the board could only assume the presence of arsenic in the deceased’s body. The hubris and ethical lassitude of Paris’s treatise had left the board of physicians in the Logan case “like mariners in an ocean to them unknown, the rocks and shoals of which were left unnoted in their only chart.” Following Jackson’s critical appraisal of the board of physicians’ report and notwithstanding the board’s sworn testimony that Logan died from arsenic poisoning (and no testimony in Mrs. Logan’s favor), the grand jury acquitted Mrs. Logan by a vote of 23 to 1.

Jackson’s concerns about incompetent “expertise” were not unique. Just two years earlier in England, in an 1826 trial in Sussex, Hannah Russel and a lodger in her home were accused of poisoning her husband. Evidence that she had purchased arsenic, together with the testimony of a local surgeon who said he found arsenic in the victim’s stomach, resulted in convictions. The lodger was hanged, but Russel’s execution was delayed. Gideon Mantell, a Sussex physician and geologist, took an interest in the story. Convinced that the deceased—who had experienced heart problems—had not been poisoned, Mantell criticized the surgeon’s tests and sought confirmation of his views from other physicians. When Russel was pardoned, the pattern of overconfident experts, later corrected by those with better procedures or credentials, was firmly established. Such disagreements, in the Logan and Russel cases, prefigured what would come to be known as the “arsenic wars.”

The Arsenic Wars

The term arsenic wars usually refers to toxicological debates between Parisian academics in the first half of the 19th century. This was a time when arsenic detection techniques progressed from a collection of relatively unreliable precipitation and reduction tests to the more substantial methods devised in 1836 by the British chemist James Marsh and in 1841 by both Swedish Jöns Jacob Berzelius and German Hugo Reinsch.

By the 1830s French toxicologist Mateu Orfila (1787– 1853) was a medical celebrity in Paris and abroad, the dean of the Paris Faculty of Medicine, as well as a frequent forensic expert. He was actively involved in the highly publicized 1840 trial of Marie Lafarge, who was accused of poisoning her husband with arsenic. Orfila’s authoritative detection of arsenic in the corpse, and his disagreement with other experts who found no arsenic, sent Madame Lafarge to prison and set off a debate around Europe about arsenic detection.

The Lafarge affair offers a picture of numerous tests and practices for arsenic detection and disagreements over their conclusiveness, prior to gradual acceptance of the Marsh test. Local physicians in the case had first relied on autopsies and symptoms, while Orfila encouraged chemical analysis and was critical of “smell” tests (e.g., attributing a “garlicky” odor to arsenic). Orfila, however, was also critical of the initial chemical tests, which led investigators to confirm the presence of arsenic on the basis of ambiguous colors. As a result of the growing popularity of the new Marsh test, three pharmacists employed it several times and found no arsenic. But doubts remained, and Orfila was called on to reapply the Marsh test—he confirmed that arsenic was present and that it did not come from the chemicals used in his analysis or the earth from which the victim’s body was exhumed. These qualifications were important because critics had pointed out that zinc used in the early Marsh test could contain arsenical impurities and that cemetery soils could contain arsenic.

Orfila was accused of hubris in his tests by François-Vincent Raspail (1794–1878), a frequent challenger of Orfila’s methods in trials and in scientific publications. Raspail raised concerns about the presence of arsenic in copper vessels used to boil cadavers and in the paint on the wood used to transport them. He viewed Orfila as too theoretical and experimental for the practical life-and-death concerns of the courtroom. In addition Raspail reflected on the difficulty of opposing Orfila in the courtroom, given Orfila’s fame and ability to decide university appointments and dismissals. Orfila’s role as a skeptic and as a critic of the overconfidence of the initial experts in the Lafarge trial was, in Raspail’s view, eclipsed by Orfila’s own overconfidence and blindness to the limitations of his toxicological experimental methods.

The Lafarge trial was a sensation in Europe, and Orfila’s celebrity caused international interest in the scientific procedures. Orfila’s reception in Britain was generally positive, but his hubris was a concern. Despite Orfila’s obvious brilliance, he seemed to overreach himself, to be too enthusiastic. Orfila’s British critics saw him as sacrificing the restraint required of a courtroom expert. In his textbooks the famed British toxicologist Alfred Swaine Taylor (1806–1880) referred to Orfila as an example of immodesty—of confidence even when the results of his arsenic-detection techniques were ambiguous.

Taylor himself, however, was soon accused of overconfidence in his work with his Guy’s Hospital colleague George Owen Rees on the trial of William Palmer in 1856. Palmer was a country physician suspected of poisoning his wife, his brother, and his gambling partner, John Parsons Cook. Taylor testified on the basis of clinical evidence that Cook had been poisoned by strychnine, even though the poison was not detected. Defense witnesses argued that a skilled analyst would have detected any strychnine present, and criticism continued in the popular press. Taylor defended himself in part by recalling Orfila’s sins of excess in the Lafarge case, claiming that his own critics, like Orfila, had inflated the capacity of chemical analysis. Nevertheless, despite Taylor’s successful efforts to reframe the Palmer case in his textbooks, his public image suffered, and he was condemned by some for his scientific pretentiousness. Three years after the Palmer trial Taylor testified in the arsenic-poisoning case of Thomas Smethurst; on the basis of a single result using the Reinsch test he declared the presence of arsenic and Smethurst was convicted. (In Reinsch’s test arsenic forms a metallic coat on a copper leaf treated with nitric acid when it is placed in a heated solution of hydrochloric acid and arsenic.) However, William Herapath, a Bristol toxicologist, later showed that Taylor’s use of the test was faulty and Smethurst was pardoned.

A Striking Parallel

In January 2007 a jury in San Diego convicted Cynthia Sommer of murdering her husband, a marine sergeant, with arsenic. It was reported that she contacted an Internet dating service before her husband’s death, and that soon afterward she threw wild parties, had breast augmentation surgery, and began a new romantic relationship. These potentially irrelevant circumstances might have interfered with scientific investigation. On the basis of laboratory tests and expert opinions that seemed to prove that her husband died of arsenic poisoning, Sommer spent over two years in the Las Colinas jail. But when the verdict was challenged, sufficient questions were raised concerning the laboratory procedures and results to lead the prosecutor to send newly discovered tissue samples (preserved in paraffin) to the Quebec Toxicology Center. No arsenic was found in any of the samples, and tests determined that the earlier reported tissue distributions of arsenic were “physiologically improbable.” The murder charge was dropped.

The defense’s criticism of the prosecution’s experts and evidence repeated a pattern begun in the 19th century when arsenic-detection technologies were rapidly evolving: a toxicological expert was overly confident, the defendant was convicted, a more rigorous expert reanalyzed the evidence, and the defendant was acquitted. In the Sommer case it became clear that the initial specimens had been improperly transported and managed and that the government laboratory where they were tested lacked standard operating procedures and was not accredited by the American Board of Forensic Toxicology. The inexperienced analyst lacked the necessary knowledge to test for arsenic, and he rejected the possibility of sample contamination on the basis of an unscientific belief that a contaminating source can be monitored. In addition the lab did not follow a peer-reviewed arsenic speciation method and lacked appropriate chain-of-custody documentation for the specimens. Finally, the finding of 100% dimethylarsenic acid (DMA) was not consistent with other case studies involving death by inorganic arsenic, and Sommer’s husband did not display the appropriate symptoms before his death. As in the Logan case, the only explanation for such carelessness is in the unjustified overconfidence of the prosecution’s experts. In Jackson’s words, arsenic detection in the service of criminal law is a “life and death . . . business which the most experienced ought to approach with fear and trembling, with terror and dismay.”

Since the early 19th century, arsenic-poisoning trials have provided fodder for speculation on the moral character and salacious misdeeds of the accused—and of the scientists who accuse them. In the United States, France, Britain, and elsewhere accusations of hubris and moral carelessness have been as important in the advancement of science as in the advancement of justice. And, as the Sommer case makes clear, the arsenic wars have not ended. The dynamic changes in arsenic-detection techniques in the first half of the 19th century are similar to rapid developments in forensic science in the last several decades. The modesty and moral circumspection of some of the experts in chemistry just before and during the arsenic wars provide a model for the appropriate attitude toward forensic science in contemporary courts.