It is my understanding that more than one collection of Roy Vickers’ inverted mystery stories have been put out under the title The Department of Dead Ends, but also that this The Department of Dead Ends (1949) is the first time it was done, with ten stories telling of ingenious murderers and the miniscule oversights that eventually caught them, thanks to the elephantine memory of that eponymous division.

I actually bought this ebook collection — a Haycraft/Queen Cornerstone title, no less — several years ago, and gave up on it after the first few stories because the style of mystery was not, at the time, to my liking, preferring as I did the genius detective whose rapier-like insight cut swiftly to the heart of the most baffling conundrums. Vickers’ intention is set out in the opening story to be very much not that, and so perhaps had to work against my prejudice rather harder than it otherwise might:

It was the function of the Department to connect persons and things that had no logical connexion. In short, it stood for the antithesis of scientific detection. It played always for a lucky fluke — to offset the lucky fluke by which the criminal so often eludes the police. Often it muddled one crime with another and arrived at the correct answer by wrong reasoning.

However, the intervening years have seen me drink deeply at the waters of the procedural and inverted mysteries of R. Austin Freeman and Freeman Wills Crofts, among others, and I return to this collection a better-educated and more tolerant man, ready to see the value in that which I summarily dismissed in my callower days.

So, shall we…?

The first-ever DDE story, ‘The Rubber Trumpet’ (1934), sets out the store perfectly, with the Department “hitting the bull’s-eye by mistake”, in frank contravention of what detective fiction had become by this stage in the Golden Age. While it’s not strictly true that the eponymous toy instrument “had nothing logically to do with George Muncey, or the woman he murdered”, there’s a pleasing circularity in the absurdity of how the events conspire against our murderer by sheer happenstance.

Vickers does well to invest the lives of his killer and his victim with no small amount of pathos, tracing the details of their otherwise-dull lives in a way that both makes much of the inconsequential and the quietly moving — c.f. the newly-wed Mrs. Muncey (well, Mrs. Prince) only resorting to the trumpet for entertainment “after a honeymoon of neglect and misery”. It amuses me, too, that Vickers originally seeks to undo the whole ‘psychological moment’ trope by having George “read[ing] very little and…not know[ing] that murderers were popularly supposed to be haunted by their crime”. It’s only a shame, then, that this must be overturned almost immediately to make the latter stages of the story work.

It was the inclusion of ‘The Lady Who Laughed’ (1948) in the British Library collection Final Acts [ss] (2022) which inspired this reread of Vickers’ work, so let’s add a few further thoughts here. Vickers really does excel at the psychology of his characters, taking time to set up their personalities, frustrations (“Other men, of course, could make their wives laugh like that.”), and ideals in a way that rarely feeds back into the detection — the details here very much out of step with the minutiae that would betoken the puzzle plots which the genre had taken to heart previously, and hinting at the general shift towards crime fiction at the time this was published.

The story of how Lucien Spengrave, great clown actor of his generation, kills his wife is told in an engaging manner, but the way in which he is caught by the Department — essentially he writes his plan down on a bit of paper and leaves it lying around for someone to find — feels like a lazy parody of detection which is beneath the point Vickers was making. Or maybe I’m wrong, and that’s exactly what Vickers was doing. We have eight more stories in which to find out…

Now, ‘The Man Who Murdered in Public’, a.k.a. ‘The Clue-Proof Murders’ (1934) — in which a man drowns four women, is strongly suspected by the police and knows the police suspect him — is a bit more like it. Not only does it get into the mind of our killer a little more…

What fools, we imagine him reasoning, are murderers to be caught! To mess about with poison and guns and knives, which always leave clues! Whereas, if you have an accident which lots of people can witness it does not matter if you contradict yourself a bit. You are expected to be flurried.

…it also highlights the gentle comedy of Vicker’s tone (a comely woman using the “instrumentality” of a frock to get what she wants) and shows the genius of the same offhanded relation of events when veering into the less salubrious side of events (“After the thumping she was extra docile.”). Equally, George Macartney’s uppance cometh via the Department in a way that “[n]o purely logical detective” would devise, showing the value of this sort of anti-detection far better than the two stories preceding it here.

Diverging slightly from the established pattern, ‘The Snob’s Murder’ (1946) leaves us in some doubt how its murder — that of a music hall singer who has a claim of marriage on a hereditary peer — was accomplished, and by whom. That Nelly Hyde is strangled in quite brutal fashion is undoubted, and that Lord Brendon has the capability to achieve such ends following his military training is admitted by the gentleman himself…the difficulty is whether he could accomplish it in 30 seconds with his uncle standing mere metres away. And a claim in public that he could sees the accused resorting to the courts for satisfaction.

As it happens, we’re never told whether he did it: the Department apparently coming into this on the back of the theft of the family jewels — careful — and instead drawing a conclusion about their pilfering from evidence which, I have to be honest, is so subtle that I can’t work out how it implies what Vickers seems to think. And even if I have missed something, how does guilt of the theft prove the murder? I don’t mind being treated as intelligent by my fiction, but I cannot shake the feeling that the parts are not explicitly joined up here simply because they do not join.

Another slight variation on the expected order of things occurs in ‘The Cowboy of Oxford Street’, a.k.a. ‘The Case of the Merry Andrew’ (1936), with Andrew Amersham killing not Constance Amersham, the wife who lied her way into his affections (“…she would have married almost anyone who could afford to support her…”), but rather William Harries, the ex-admirer of Constance. With Harries establishing the three of them in a household and then slowly edging the diffident, unsuspicious Andrew out of his own marriage, the amazing thing is how long it takes murder to result in the first place!

The means of being caught by the Department, however, again doesn’t ring true. I can believe the deduction which finds something suspicious in Harries’ death, that’s just good detective work, but surely, despite their outward physical similarities, (rot13 for minor spoilers) n ynqvrf jngpu jbhyqa’g tb bire n zna’f jevfg. A lovely moral touch rounds this out, with again solid character work done by Vickers in letting us know his killers, but this is another disappointing return on the promise of this core conceit.

Also, Inspector Horlicks. Best character name ever.

We’re back on firmer ground with ‘The Clue of the Red Carnations’ (1949). The plot sees solicitor Hugh Wakering encounter Jeannie Ruthen, the woman he had loved over a decade before, as she leaves prison. Learning of her descent into a degenerate life, Wakering seeks vengeance on the man he blames: Cuthbert Bristowe, Wakering’s romantic rival for Jeannie’s affections back in the day.

The detection is good, but leans far more into the Humdrum approach of a simple police procedural than the sort of non-logical, accidental detection Vickers has stated as the point of these stories. Indeed, following up the eponymous clue is simply good policing, and would occur to anyone doing this sort of job ‘properly’. So a return to form, but at the expense of the wider point it feels Vickers wanted to make elsewhere.

Read in isolation, I might enjoy ‘The Yellow Jumper’ (1946) more. Certainly there’s an interesting point here being made about the spontaneous murder by someone temporarily insane when thrown in the face of the meticulously-planned schemes which multiplied throughout the Golden Age. And there’s something very acute in the calm acceptance of the murderer of her deed once the murder is done: “she took it for granted that her own life was, in effect, at an end, and this gave her an immense freedom”.

But the detection here, while relying in part on small and easily-overlooked ideas — “[She] was convicted for no other reason that that she had put on the dead girl’s dress and worn her perfume.” — requires one witness to hold an Olympic medal in unobservance and for all men to get in a helplessly blustering fluster whenever wimmin’s cloves are mentioned. Interesting use of (very, very off-handed) forensic science, though. Wonder how possible most of it was…

Again, I get the impression from ‘The Case of the Social Climber’ (1947) that what Vickers really wanted to be was a writer of Humdrum detection. The tone here is again delightful, effortlessly bringing the snobbery of the English fully into view…

The more insular types will assure you that there are only six public schools in England, two in Scotland, and none at all anywhere else in the world.

…but, once more, in order for Dead Ends to solve this what inspector Rason actually does is solid detection rather than the sort of accidental, non-logical linkage that would be a) huge fun, though b) far harder to conceive. There’s some magnificent psychology in here (“He blamed [the victim] for the murder as a man will blame a too attractive wanton for his own debauchery.”) and a savage takedown of the importance placed on position over happiness, and in isolation it would be delightful. But that’s not what I was promised at the start, and so, once again, Vickers, for me, frustratingly fails to fulfil his own brief.

Again, we have a superbly set up story of unplanned murder in ‘The Henpecked Murderer’, a.k.a. ‘The Eight Pieces of Tortoiseshell’ (1947), and again we have Vickers not quite meeting his own conditions in how the crime is detected. By deliberately drawing parallels with the Dr. Crippen case, Vickers cleverly draws your eye in slightly the wrong direction, and then drops a lovely development which really rounds out to the experience. It’s good, solid fun, insofar as someone strangling someone else and burying their body clandestinely qualifies as ‘fun’.

And, again, Vickers’ use of tone is magnificent — c.f. his muted prurience when it comes to talking about sexual enjoyment — and he raises the possible reasons a man may be blackmailed for £1,000 in a manner that is never explicit yet seems to admonish the reader’s tantalising willingness to delve into smut. How he does this should really be studied, because the tightrope walked in these sentences is startlingly close to art. But, really, his detection is once more the finest Humdrummery, and as such I come away from this frustrated once more.

Finally, ‘Blind Man’s Bluff’ (1948) — a popular title in crime fiction circles — in which lawyer Robert Swilbey (or maybe Silbey, the text goes back and forth) is blinded by a discontented witness and so switches to the career of playwright…only for the attentions an impresario is paying his wife to come to his attention and, due to a mental quirk of this blind man in denial which borders on a Jim Thompson-esque realisation of schizophrenia, murder seeming the only way out of the difficulty.

Vickers’ eye for the subtleties of phrasing are on keen display here (“There is no gentlemanly method of committing murder.”), and in the closing stages this perhaps ends up affecting the plotting, with the reader required to do a certain amount of joining the dots in order to see how the conviction of murder comes to Rason…but it feels more like the sort of intelligent subtlety I’d expect from this sort of enterprise. I don’t mind being treated intelligently, as I’ve said, but this feels like a little bit of a reach, where a soupçon more effort could have tidied the thing up for good…though I can, this time, at least believe it possible.

Can I do a top five? Let’s try and do a top five:

1. ‘The Man Who Murdered in Public’, a.k.a. ‘The Clue-Proof Murders’ (1934)
2. ‘Blind Man’s Bluff’ (1948)
3. ‘The Case of the Social Clumber’ (1947)
4. ‘The Clue of the Red Carnations’ (1949)
5. ‘The Yellow Jumper’ (1946)

Most of these, read in a collection, surrounded by the works of other authors, would be highly enjoyable tales, and it’s worth bearing that in mind when I say that this compendium fails to live up to its promise. I’m a massive fan of Humdrumming, but Vickers’ stated intention was to do something rather more in the lines of anti-detection, playing long odds and stumbling onto the answer by mistake. Doubtless he achieves this elsewhere in these DDE stories, but it’s a frustrated ambition here, and on those ground I can’t all this anthology a success.

Vickers uses tone superbly, however, telling you his various tales of murder in an entertaining way that rounds out the psychology of all involved well, playing well into the boundaries of the crime novel that was taking over from the pure detection of the Golden Age. His slightly pedantic narrative voice recalls a man who is determined to get each of the minor details of his laborious story correct even though half of them don’t matter, and if you’re able to get on board with that you’ll have a great time. I, for one, thoroughly enjoyed it.

I can’t honestly say I’ll seek another anthology of Vickers’ DDE stories, but if they turn up in multi-author collections I’d read anything in this series — including a revisitation of the tales above — with great pleasure, especially now that my expectations are suitably adjusted.