First published: UK, Hodder & Stoughton, 1941; US, Little Brown, 1941
It looked like an obvious suicide pact. The minister of a minor Protestant denomination and one of his parishioners are found dead in the bracken, shot dead just as they were about to elope.
This is my second assault on this late Connington. The first time, three years ago, I gave up, defeated by the drab setting (lower-middle suburbia and chapel-going corpses), paragraphs in Cockney, police investigation, and a lot of tedious stuff about the position of bodies, tracks in grass, and fingerprints, all told in soporific, expository corpse. “Not much need to go further for the present,” Insp. Rufford tells himself after enduring a few dry paragraphs about fingerprint whorls, deltas, and loops. I agreed, and gave up a quarter of the way through.
The 21 Clues isn’t bad – but it’s not particularly good, either. It’s a British orthodox detective story, and like far, far too many books in the genre, underwhelming and unsatisfactory. The investigation is solid and workmanlike enough, enlivened by a perceptive maid, an engaging journalist, and Connington’s dry wit. The solution is long-expected but clever enough; the hidden relationship is the same that one of my blogging comrades has encountered this week.
But it’s strangely empty fiction. There is almost no characterization. Only the first chapter offers an intimate look at some of the suspects; the other 16 are seen exclusively from the perspective of the police / journalist / Sir Clinton Driffield. For them, the mystery is an intellectual challenge, a puzzle to be solved, rather than a matter of life and death. Even when the inevitable third murder occurs, Driffield leisurely banters and debates, and speculates on the destination of the victim’s soul. Is he boring the choir of the immortal dead, or arguing with Charon?
“No need to be so unfeeling about it,” Wendover commented in a tone of disapproval.
“I don’t pretend to like white slavers [And how! – Ed.], blackmailers, murderers, or the writers of anonymous letters,” retorted Sir Clinton cheerfully.
Nobody has any stake in clearing up the mystery. One of John Dickson Carr’s achievements (building on A.E.W. Mason) was to make the viewpoint character’s happiness depend on solving the crime (either because he was suspected, or because the girl he loved was). From the 1930s on, Agatha Christie told much of the story from the perspective of one of the suspects, giving the reader both an inside look at the murder and (importantly) someone to care about. (Hell, even Crofts did this.) Christianna Brand used this approach brilliantly. She and Nicholas Blake (and the maligned Ngaio Marsh) showed the effect of murder on a small group of people. Margery Allingham used the genre to illuminate a little world (publishing, dance, high society), while making Albert Campion a character in the novels – in love with his suspect’s wife, for instance – rather than an impartial observer. Dorothy L. Sayers gave the detective story thematic depth and richness; as Chesterton had done, the murder was an allegory for a social or moral problem. Ellery Queen gave his namesake sleuth introspection (ANGST!) and responding to the ills of the 20th century. And all these authors strove to bring the detective story closer to the novel while maintaining – even heightening –the fair play puzzle plot. And they succeeded brilliantly.
But I increasingly find myself indifferent to the orthodox detective story that lacks characterization, style, and atmosphere, however well constructed. I’ve enjoyed very few of the detective stories I’ve read this year – except for Michael Innes‘s wonderful Stop Press.
Let’s go back about 20 years, to me aet. 16 or so, addicted to detective stories in the same way my schoolmates were to cut-price hashish, the pleasures of the palm, or evangelical proselytism. I was irked when I read Julian Symons’s Bloody Murder to see him dismiss Connington as a Humdrum, a constructor of mere puzzles.
I was lucky, you see, only to have read Connington’s best books: The Case with Nine Solutions (1928), with lots of juicy murders and clues; Murder in the Maze (1927), fast-paced enough to read in a single sitting at the National Library; the rather dull Nemesis at Raynham Parva (1929), more interesting in conception than … execution; and Jack-in-the-Box (1944), boasting a sinister mystic and a wealth of scientific slayings.
Three of those are, quite simply, FUN. That, after all, is why we read detective fiction; and if they’re not fun, why bother? (The epicurean aesthetic!) They are, however, the outliers. Many of Connington’s later books aren’t much fun. They’re praised for methodical detection, solid construction, and immaculate logic. Not for zest, humour, atmosphere, or action!
Many of his 1930s books feature police detection, murders in suburbia, and flat prose. Even the better books from this period – The Castleford Conundrum (1932) and The Ha-ha Case (1934) – are absorbing , rather than exciting.
By 1941, Connington had been in decline for some time. In Whose Dim Shadow (1935), Truth Comes Limping (1938), and For Murder Will Speak (1938) were the work of a tired, ill man. His last two books, The Counsellor (1939) and The Four Defences (1940), were failed attempts to rekindle his inspiration by writing about a new detective – a witty, wealthy radio host.
In fact, these days I often find myself agreeing with Symons about the Humdrums. Yes, the best of their work should be praised. But the best detective stories were written by artists, not artisans.
The Constable was content to call it a suicide pact. All the obvious facts were in favour of this solution. The bodies of John Barrett and Mrs. Callis were discovered in a lovers’ nook among some bracken. Beside them was a pistol with Barrett’s finger-prints on it, and torn up letters in the handwriting of Barrett and Mrs. Callis were scattered about on the turf. Inspector Rufford unearthed some puzzling problems. Arrangements for the elopement of the couple had apparently been complete. Why had their plans fallen through? Why had they turned their backs on the railway station, with tickets to London in their pockets? Why was a small sum of money missing from the bag in the wrecked car? It was not until Sir Clinton Driffield returned from his holiday that a satisfactory solution was found, broad enough to include in its scope all the twenty-one clues, as well as the reappearance of the Jubilee double-florin and the fate of the shooter of cats. Mr. Connington’s readers will expect to find all the clues honestly laid before them as usual, and they will not be disappointed.
The bodies of a worthy prelate and one of his wealthy lady parishioners are found in a bracken patch somewhat famed as a lover’s rendezvous. Torn-up love letters are scattered about. A pistol with the clergyman’s finger-prints lies beside the body…
If Mr. Connington says there are twenty-one clues, you can depend on it that they are offered to you as they turn up and that they will add up to the solution of the case. Mr. Connington has been aptly termed “the modern master of pure detection”. Can you fit together the pieces of his latest fascinating mosaic of death? Can you beat Sir Clinton Driffield to the solution?
Times Literary Supplement (Maurice Willson Disher, 3 May 1941):
What looks like a suicide pact soon looks too much like it. There are more clues than are wanted. Too many bullets have been fired, too many tracks disturb the bracken, too many proofs are provided of an old passion. This plainly is the kind of fake for Mr. Connington to peel layer by layer like an onion. The Counsellor, whom he established in his recent novels as an amateur detective with a promising future, does not appear. Regular police methods make a thorough job of the investigation, a newspaper reporter helps and the Chief Constable puts the bits together until the whole series of events becomes as clear as though it had been filmed. Calculating patience is Mr. Connington’s chief characteristic. At times he dawdles so calmly over routine detail that he exasperates, but this assurance compels admiration directly you are even dimly aware where his purpose lies.
Sat R of Lit (10th May 1941, 40w):
Methodical and painstaking investigation of ingeniously-planned crime should please those who care more for deduction than action. Well thought-out.
Books (Will Cuppy, 11th May 1941, 170w):
Without stating in so many words that this is a fiction based upon the notorious Hall–Mills murder case, we venture to hint that the wind is in that direction. Mr. Connington has veiled, adorned, elaborated and otherwise adapted the story with a goodly bag of tricks, including a most adroit substitution of identification. He has also provided a fine brand of detection and an answer that exactly fits his version.
NY Times (Isaac Anderson, 11th May 1941, 180w)
Observer (Maurice Richardson, 11 May 1941):
J.J. Connington makes a come-back to his best form in The Twenty-One Clues, a double provincial murder. Brilliant solution by Sir Clinton Driffield after thorough detailed investigation. The best Connington for several years.
Spectator (John Fairfield, 16th May 1941, 20w)
New Yorker (17th May 1941, 80w)
Manchester Guardian (E.R. Punshon, 27 May 1941):
Off with the new love and on with the old. Mr. J. J. Connington in The Twenty-one Clues deserts his broadcasting detective to return to Sir Clinton Driffield, who solves an apparent case of double suicide which both he and the reader soon suspect was really a case of double murder. The working out of the solution shows all Mr. Connington’s remarkable and careful ingenuity, a quality in which he remains unsurpassed.
Time (9th June 1941, 40w):
Clever, and a satisfying puzzle.
New Statesman (Ralph Partridge, 16 August 1941):
The Twenty-one Clues is old-fashioned detection (and quite adequate, but no more) solved by Sir Clinton Driffield in his prosiest vein. Instead of probing the bewildering mountain of clues, many readers will take a short cut by glancing at the characters of the persons involved. Sir Clinton can be trusted only to convict cads.
Booklist (September 1941)
Bookmark (November 1941)
A Catalogue of Crime (Jacques Barzun & Wendell Hertig Taylor, 1989):
A poor repetition of the situation in The Case with Nine Solutions. The 21 clues are in themselves ingenious and attractive, but the show gives itself away very early. The doings of Driffield, Wendover, the inevitable “young reporter”, and a few yokel policemen are Connington at his worst and wearisomest.
W.H.T. dissents: By no means bad. An apparent double suicide gives Insp. Rufford a chance to collect evidence but Sir C.D. has to come back from his holiday to put the 21 clues in order. The plot is elaborate, but sound—there is even a rumour that it had its origin in the Hall-Mills case, q.v.