In 18th century Europe, arsenic was close to a perfect murder weapon: odorless, tasteless, easily dissolved in tea or soup, and a death that looked just like cholera. The French called it inheritance powder. Early chemists could coax a garlic smell or a yellow precipitate from a sample, but nothing that would last long enough to show a jury. This episode tells how that changed, through a frustrated chemist, a gloating murderer, and a sensational French trial that gave forensic toxicology its legitimacy.

In 1832 James Marsh, a chemist at the Royal Arsenal in Woolwich, tested the coffee John Bodle was accused of poisoning to kill his grandfather. The yellow proof faded before trial, the jury acquitted, and Bodle later confessed. Marsh spent years on a fix and in 1836 published a glass apparatus that turned arsenic into gas, burned it, and left a permanent silvery black mirror on a cold ceramic bowl. In 1840 the test met its first great public trial in the case of Marie Lafarge.

• Carl Wilhelm Scheele found in 1775 that arsenic treated with nitric acid and zinc gives off arsine, a gas that smells strongly of garlic.
• Zinc and sulfuric acid in the Marsh apparatus generate hydrogen that strips oxygen from the arsenic, and the darkness of the stain shows how much poison was present.
• To rule out antimony, Marsh applied sodium hypochlorite: arsenic stains dissolve, antimony stains stay put.
• Local chemists botched the test on Charles Lafarge’s exhumed body and found nothing, until the toxicologist Orfila ran it correctly and Marie Lafarge got life in prison.
• After the trial, Parisian salons hired chemists to perform the test as entertainment, and arsenic murders fell into steep decline.