The Science of Mysteries: Instructions for A Deadly Dinner


by Deborah Blum

When people ask why I would choose to write a book about poisons (The Poisoner’s Handbook) I usually start with my brief stint as a chemistry major, my continuing affection for using poisons as a way to think about our chemical planet. But I always end up admitting that – and, yes, this will make me sound a little twisted –  I’ve been thinking about poison murders since I was in high school.

That was when I started reading my way through my mother’s collection of early 20th century murder mysteries – Sayers, Agatha Christie,  Georgette Heyer, Mignon Eberhart, Patricia Wentworth – women who spun the most intricate plots around the most evil chemistry. Of these, only Christie is really famous today, more for her brilliant plotting and  quirky detectives like Hercule Poirot or Miss Marple than for her savvy toxicology. But she knew her chemistry; she’d worked as a nurse and in a hospital pharmacy.

In her debut novel, The Mysterious Affair at Styles (1920), the killers deftly use bromine to precipitate strychnine into the bottom of  victim’s tonic bottle, carefully timing that last lethal dose. It was this that caught my attention and imagination – the elegant use of a lethal substances, the way the peculiarities of a poison could literally carry a plot. A much later Christie novel, The Pale Horse (1961), uses the unnerving symptoms of  poisoning by thalliumto produce both a puzzle and a ominous sense of disaster.

But none of these writers, I think, did more justice to that most famous of homicidal poisons, arsenic, than did Sayers in Strong Poison. The title comes from the lyrics of a 17th century ballad, The Poisoned Man: “O that was strong poison, my handsome young man/O yes, I am poisoned mother; make my bed soon/For I’m sick to the heart, and I fain wad lie down.”

But the chemistry is absolutely up-to-date for 1930, the year the book was published. In fact, Sayers for all her literary background (she was an Oxford University educated scholar of classical languages and considered her translation of Dante’s Divine Comedy to be her best work) performs as an outstandingly good science writer in the course of the story. Consider this description:

And presently, definitely, magically, a thin silver stain began to form in the tube where the flame impinged on it. Second by second it spread and darkened to a deep brownish-black ring with a shining metallic centre.

“Holy blank,” I said to myself. “She’s talking about the Marsh test.” I was at the moment rereading Strong Poison, and I was exactly 29 pages from the end but I put the book down so that I could run downstairs and inform my husband that Sayers really knew her arsenic.

I don’t know why I bother to tell him these things. He always looks so hunted.

Unknown artist, Mr James Marsh (1794–1846), Royal Arsenal Ordnance Chemist, c. 1840

But about the Marsh test. It was developed by a British chemist, James Marsh, during the 1830s. At the time, there was no good test for detecting arsenic (or really any poison) in a corpse. Marsh himself had been involved in the prosecution of one accused murderer who was found not guilty partly because the science was so inconclusive. After the verdict, the man admitted that he had indeed killed his grandfather with arsenic. Infuriated, Marsh dedicated his spare hours to looking for a better test. The technique he developed would eventually revolutionize toxicology – it worked, it led to convictions, and it was credited for a reduction in 19th century arsenic murders, as would-be poisoners for first time worried about scientific evidence.

The Marsh test used a mixture of zinc, acid, and heat applied to suspect tissue to generate a fine vapor. If the tissue contained arsenic, the vapor would include arsine gas which, when cooled, formed a dark silvery “arsenic mirror”. In fact, that shining metallic formation so beautifully described by Sayers above.

That revealing Marsh test, as you might imagine, appears fairly late in the story. Sayers begins the book at the murder trial of a mystery writer, Harriet Vane, accused of killing her former lover with arsenic. Vane had purchased arsenic shortly before the death although she claimed to have done so has part of research for a novel.  The detective who starred in most of Sayer’s mysteries, Lord Peter Wimsey, attends the trial and becomes convinced of Vane’s innocence. He’s also smitten by Vane’s intelligence and fierce independence – setting the stage for a courtship than runs through several following novels.

But the judge in the trial isn’t smitten at all. As he instructs the jury, he details Vane’s promiscuous behavior with the dead man (she lived with him) and dwells on her unnerving knowledge of arsenic, noting that Vane “evidently gave considerable thought to the subject, for there were a number of books on her shelves dealing with forensic medicine and toxicology, and also the reports of several famous poison trials..”

I had a moment in which I realized – is this why my husband occasionally looks so hunted? – that this fictional character might be describing my literal library. The rows of books about  forensic medicine and toxicology books, the specific case studies of murderous events. Like Sayers’  Harriet Vane, I even possess a book on “the Armstrong case.”

But in our mutual defense, this was a famous murder; Herbert Rowse Armstrong, executed in 1922,  poisoned his wife with so much arsenic that even after death it was found oozing out of her hair. Armstrong is still famed as the only solicitor in the history of the United Kingdom to have been hanged for murder.

Bob Shand: Poison Bottles, 2014 (CC)

In Sayer’s story, Vane is saved by a hung jury (the holdout is a friend of Wimsey’s) and before the new trial is set, the detective is able to discover the real killer. There’s never a doubt that Vane’s lover, Philip Boyes, was killed by arsenic. His death is a catalog of classic arsenic symptoms – from severe nausea and vomiting, to tremors and chilling of his hands and feet. Arsenic’s effects are well known, thanks to its long homicidal – and industrial –  history and the fact that it’s a naturally occurring element, known to contaminate water supplies around the planet.

So the problem in solving Boyes’ murder is mostly figuring out how he ingested that last lethal dose of the poison. The evening that the victim died, he’d shared a dinner with his cousin in which they both ate from the same dishes. The cousin, as the judge noted, remained perfectly healthy. Boyes had then gone to visit Vane who might, or might not, have served him a poisoned cup of coffee. But if not at her hands, then whose?

“You’ve got to have some plausible pretext for giving a bloke arsenic,” complains Wimsey to an associate. “You can’t just catching him standing on a doorstep and say ‘Here, have a drink of this,’ can you?”.

Eventually, Wimsey discovers that someone besides Vane (and no, I won’t tell you who) has a secret supply of packets of a mysterious white powder. At this point it’s clear that Sayers has also been reading up on Herbert Rowse Armstrong, who put arsenic weedkiller into paper packets and kept them in his pockets.

Armstrong claimed that he needed the poison on him in case he stumbled upon an unexpected weed.He had, in fact, been tipping it into his wife’s tea. One of the reasons that poisoners liked arsenic so much – before scientists like Marsh made it so detectable – is that it tends to be bland in taste, difficult for a victim to detect in food or drink. In some of my favorite 19th century tests by forensic chemists, they mixed it into everything from pudding to wine to prove that point. (In case you wondered, that fact also comes from my personal library.)

Anyway, it’s that neatly discovered white powder that Wimsey puts so successfully through the Marsh test. Once he’s found the poison, he is able to reconstruct the crime and to trick the murderer into a confession. I won’t give it all away here but it involves the idea of building up an arsenic immunity before serving a particularly deadly dinner.

When I arrived at the solution to the mystery, I thought to myself, “Holy blank, she’s talking about the arsenic eaters of Styria.” These were 19th century Austrian peasants who supposedly became resistant to arsenic toxicity by eating it in small regular doses. One report had them smearing a poisonous paste on their morning toast. At Sayer’s time, arsenic-eating was prominently featured in toxicology books and her description is, again, textbook perfect.

Or at least 1930s textbook perfect. It was a hot theory in her time but today it appears less convincing,  based on mythology as much as science. Modern science tells us  that chronic exposure to arsenic is likely to bring on a host of health problems, ranging from skin discoloration to malignancy, that Sayer’s villain would not have enjoyed unblemished good health.

But her ending holds up perfectly in terms of the science of 80 years ago. And the other details – the shimmering results of a Marsh test, for instance? They’re as good today as they were when Sayers plotted out her story of arsenic and murder by a very strong poison.

About the Author

Publication Rights

Piece originally published at PLOS Blogs. Republished here under a CCBY Creative Commons Attribution License 4.0. Updates were made in 2022.