First published: UK, Hodder & Stoughton, 1938; US, Doubleday, 1938, as A Puzzle in Poison
Mr. John Waterhouse died of arsenical poisoning. There was no one in the village of Anneypenney who bore him ill-will, yet he had undoubtedly been murdered. The story is told through the mouth of a close friend of the dead man who knew all the facts and who, though no detective, finally solved the riddle. There are glimpses of English village life, and of two unusual women; and it may perhaps be hinted that even the Nazis were not uninterested in the death of this English squire.
There will always be something slightly sinister about the third of September to Douglas Sewell, the narrator of this story, for on that date he saw his friend John Waterhouse dying in agony, and that is a thing one does not forget. According to the medical certificate Waterhouse died of gastric ulcers, but his estranged brother Cyril demanded that the body be exhumed because he suspected murder. It was only then that Douglas Sewell remembered the small dinner party several weeks before, which in time became a matter of great interest to the police.
Arsenic (for it was death by arsenic poisoning) could have been administered in various ways at various times, but the “how” or “when” could not solve this murder. The answer lay in the “why”. Through motive alone and through knowledge of the characters involved Sewell finally arrived at the startling answer.
A Puzzle in Poison is another of those top-flight stories of deduction which have made Anthony Berkeley one of the greatest names in present-day mystery fiction. As an absorbing and mystifying puzzle the book is a worthy successor to the author’s Trial and Error.
Here is a story of a murder committed with extreme care and amazing cleverness, yet it went awry. Despite all his cares and precautions the murderer was unable to allow for human emotions and human relationships.
One of Berkeley’s most straightforward and least experimental tales—deceptively so, perhaps? For, while the setting is the classic small English village and the murder the even more classic one of death by arsenical poisoning, Berkeley’s gifts for characterisation, psychology and dialogue are at their most impressive, showing how Cox’s gifts would work with the detective story proper and suggesting the direction in which they would have developed had he not softly and suddenly vanished away in 1939. There is little detection in the material sense. The first half is largely conversation, revealing character and psychology (detection from the inside); the clues, without detection, are given as evidence at the coroner’s inquest, recalling the similar approach adopted in The Poisoned Chocolates Case. Yet this novel approach, the detective story without detection, works extremely well. The red herrings of Military Intelligence and Nazism are neither too unconvincing nor distracting, and the solution is both psychologically and physically convincing, demonstrating how a sympathetic character cold-bloodedly commits an unnecessary (and undesired) murder.
Nicholas Blake reused the end for Head of a Traveller.
Times Literary Supplement (Maurice Percy Ashley, 9th July 1938):
Mr. Berkeley’s Not to Be Taken is very nearly as good of its kind. But here the element of pure mystery predominates. The murder is by arsenic; and although the number of suspects is strictly limited the construction is so ingenious that to attain the correct solution of the problem requires all the reader’s concentration; to skip is fatal. The murderee is a retired electrical engineer whose hobby is building, and he appears not to have an enemy in the world; yet he is painfully killed after the doctor, as is rather the way with doctors in such cases, has given a certificate. Before the end the author poses a challenge to the reader to explain how and why the murder is committed and what is the dominant clue. It requires real skill to answer all the questions directly, but there is perhaps room for a difference of opinion upon the dominant clue. In any case, it is a first-class detective story, and readers of the author’s superb Trial and Error will not be disappointed.
Observer (Torquemada, 17th July 1938):
Anthony Berkeley has now denied us Roger Sheringham twice running; but we have every reason to be grateful for what he has given us instead. First there was Trial and Error, a long and becoming vestment of satire draped upon a plot of paradox, and now we have Not To Be Taken, a compact and fascinating essay in the art of home detection. When John Waterhouse dies of arsenical poisoning, we are concerned more with the subdued suppositions of the few pleasant, village gentlefolk than with the more usual condemnations of the ale house and back fence. But the suppositions of the deceased’s priggish brother, who descends upon the community from London, are by no means subdued, and these lead to an inquest at which things look distinctly dark for the widow, until a letter, ruling out both murder and suicide, is received by the coroner from the victim himself. On page 227, the author issues a challenge defying the reader to tell and substantiate the truth. This amounted in my case to a challenge to justify a purely atmospheric certainty, and I could not do so. Plenty of sound and original clues were available, but most of them had dropped as lightly as petals upon the stream of the narrative, and I had let them float past me.
Books (Will Cuppy, 11th September 1938):
Douglas Sewell, narrator of this admirable tale, does not regard it as a detective story, chiefly because it contains “no detective as the central figure whose investigations, discoveries and suspicions the reader is allowed to share.” There are detectives in it, but as Sewell says: “In the end it was I and not any of the official detectives who eventually solved the mystery of John’s death.” As you see, Sewell is a bit of a hair-splitter, or else he’s never heard of amateur sleuths; but he turns in a solution of which Scotland Yard might be proud in the matter of the arsenic poisoning of John Waterhouse, a retired electrical engineer, of Oswald’s Gables, Anneypenny, Dorset, near Torminster. As for reader participation, there is an author’s note challenging you to answer the following questions, and otherwise share in the fun: “Who (or what) was responsible for John Waterhouse’s death? How did the arsenic find its way into John Waterhouse’s body, and why? Give a concise outline of the story behind his death. List as many deductions as you can draw from Douglas Sewell’s narrative, and the clues to them. Do you think there is a Dominant Clue in this story? If so, what is it?” Anyway, it sounds like a detective story, doesn’t it?
Well along in the story it looks bad for Angela Waterhouse, the widow, described by Dr. Brougham as “an almost pure leptosome type” and “a bit of a schizophrenic too,” as well as a “self-constructive egocentric”—she plays at being an invalid. You may also watch Philip Strangman, Angela’s sweetheart; Mitzi Bergmann, a Nazi companion–secretary; and the small circle of friends on the scene. Mr. Berkeley’s method is leisure in the extrme, maybe to match the narrator, who calls himself “a pretty dull, stolid fruit-farmer.” There’s a will, insurance amounting to ₤100,000, an exciting inquest and a conclusion of the “different” sort. The murder motive may strike you as a trifle tenuous; but that’s the way with some motives. This is a must item, of course—a sterling mystery on several counts.
Sat R of Lit (10th September 1938, 40w):
As neat a bit of bafflement as you’ll meet in months with surprise ending that may leave reader gasping.
NY Times (Kay Irvin, 11th September 1938, 180w):
The puzzle in poison—John Waterhouse’s death by arsenic—never was solved by the pursuit of tangible clues alone. The most important thread through the labyrinth was knowledge of character. But to tell you that much, reader, is not to make it easy for you to find the murderer. Anthony Berkeley’s new book is smoothly written, logical and surprising. This expert raconteur in mystery has once more written a suavely mystifying tale.
It is quietly told, and at first it seems disarmingly simple. Popular John Waterhouse and his selfish hypochondriac wife, the careless but clever young doctor with his mistaken diagnosis, his cleverer sister who can be depended on for quick action if something has gone wrong, the German girl hovering anxiously in the background, the story’s narrator and his charming and competent wife—they are all pleasant village neighbours. Other characters slip into the story later, and new complications appear. But all the pieces of the puzzle are placed fairly in the reader’s hands.
The question at the story’s end is merely ethical.
New Yorker (17th September 1938, 30w):
A grand book.