Agatha Christie's Secret Notebooks

Mystery of the Baghdad Chest’ (January 1932) were published in 1923/24. All the Poirot short stories after 1932 feature Poirot alone. No notes for any of those early stories survive and where Christie refers to them in the Notebooks, it is only as a reminder to herself of the possibility of expanding or reusing them. In many ways ‘The Incident of the Dog’s Ball’ is similar in style, setting and tone to many others dating from the early 1920s and published in Poirot Investigates and Poirot’s Early Cases. But if it was written early in Christie’s career, this in turn raises the question of why it would have lain for almost 20 years without appearing in print. It does not appear in her agent’s records of work received by them and offered for sale. I hope to show that it dates from later in her career.

In Notebook 30 it is included in a list (illustrated on the jacket of this book) which may help us to establish more accurately its date of composition:

Ideas

A. Dog’s Ball

B. Death on the Nile

C. Strychnine absorbed through skin?

D. Double Alibi e.g. A and B murder C but—A is accused of trying to murder B at same time.

E. Figurehead woman. Man back from Africa.

F. Second Gong elaborated

G. Mescaline

H. Illegitimate daughter—apomorphine idea?

Ideas to be incorporated

Brownie camera idea

Brooch with AO or OA on it AM. MA

If we apply some of Poirot’s own methods we may be able to arrive at a timetable.

This page from Notebook 30 is part-transcribed on the opposite page. Note the three intertwined fish logo in the top right-hand corner.

Clue No. 1

There are a number of immediately recognisable stories here: Dumb Witness or ‘The Incident of the Dog’s Ball’ (A), Death on the Nile (B), They Do it with Mirrors (D) and Sad Cypress (H). ‘The Second Gong’ (F) was first published in June 1932 and both Dumb Witness and Death on the Nile in 1937, the former in July and the latter in November of that year. So it is reasonable to assume that the list was written between those dates, i.e. after ‘The Second Gong’ in June 1932 and before Dumb Witness in July 1937.

Clue No. 2

Unlike Item F on the list, ‘Second Gong elaborated’, there is no mention of elaboration in connection with ‘Dog’s Ball’, lending support to the theory that it did not then exist as a short story.

Clue No. 3

In the Christie Archive there are two letters from her agent Edmund Cork. One, dated 26 June 1936, acknowledges receipt of a revised version of Dumb Witness; another, dated 29 April 1936, expresses delight at her news that Death on the Nile was finished. We now have two new limits—later than June 1932 and before April 1936. Can we do better? I think so.

Clue No. 4

It is not unreasonable to suppose that the writing of Death on the Nile and Dumb Witness, both of them among her longest books, took over a year, which would bring our latest date back to April 1935. Our new dates are now June 1932 and April 1935. And if we add two items of conjecture to the equation…

Clue No. 5

In the change from ‘The Incident of the Dog’s Ball’ to Dumb Witness, the setting moves from Little Hemel, in the county of Kent, to Market Basing, Berkshire:

General Plan P. receives letter—he and H—he writes—then he tears it up—No, we will go—Market Basing —The Lamb

Market Basing is commonly assumed to bear more than a passing resemblance to Wallingford where Agatha Christie lived. She bought her house there in 1934 and this may account for the change of setting for the novel. There is evidence for this in the reference to The Lamb, a Wallingford pub, in Notebook 63. This is, admittedly, conjecture but as Poirot would say, ‘It gives one furiously to think, does it not?’

Clue No. 6

Miss Matilda Wheeler writes to Poirot on 12 April, a Wednesday, according to Poirot’s exposition: ‘Consider the dates, Hastings’ (section v). The 12th of April fell on a Wednesday in 1933.

Conclusion?

So we may conclude that ‘The Incident of the Dog’s Ball’ was

written, in all likelihood, in 1933.

Why was it never published?

Agatha Christie was now a household name. By the mid-1930s and Three Act Tragedy she was selling 10,000 hardbacks in the first year of a new title; she was one of the first writers to appear in paperback; her books had been dramatised and filmed. Why would any magazine not jump at the chance to publish a little gem of a new Poirot story, with its guarantee of increased sales? If it was offered to them…Again, we are in the realms of speculation, but I think the reason it never appeared in print is disappointingly mundane: it was never published because she never offered it to her agent. Because, in turn, she decided to turn it into a novel. Consider the evidence:

Clue No. 1

Her production of short stories had decreased from the multiple appearances of earlier years—27 in 1923 and 34 in 1924—to a mere half-dozen in 1933 and seven the following year. As she said when she refused to contribute to the Detection Club’s collaborative novels, ‘the energy to devise a series is much better employed in writing a couple of books’. She may well have thought the same about this short story and decided to turn it into a complete new Poirot book.

Clue No. 2

The Edmund Cork letter referred to above, dated 26 June 1936, acknowledges receipt of a revised version of Dumb Witness. This seems to refer to the first four chapters, a domestic English village setting, which were added to help ensure a US serialisation sale (it was serialised in the Saturday Evening Post in November/December 1936) and would lend support to the idea that it was an expansion of ‘The Incident of the Dog’s Ball’. In the novel Hastings begins his firstperson narration only at Chapter 5; up to then the story is told in the omniscient third person with the assurance from Hastings, when he begins his narrative, that he did not witness the earlier events personally but that he ‘has set them down accurately enough’. And the opening scenes of both the short story and Chapter 5 are, apart from the month of the year, identical.

Clue No. 3

This not-offered-for-sale theory may also account for the major oversight involving the dates within the story. In section i Poirot says ‘No, April the 12th is the date [on which the letter was written] assuredly’ but in section iv he refers to August as the month when Miss Wheeler wrote the letter. An agent and/or an editor would surely have noticed a mistake of this magnitude, and one so germane to the plot.

Conclusion

In conclusion, it is entirely possible that ‘The Incident of the Dog’s Ball’ was written in 1933 and never offered for publication but, instead, transformed, in 1935/36, into the novel Dumb