The Manuscript of
"A Warning to the Curious"

by Rosemary Pardoe

(from Ghosts & Scholars 32.)

For some further thoughts on the manuscript of this story, click here to read Steve Duffy's comments, from his introduction to A Pleasing Terror: The Complete Supernatural Writings of M.R. James (Ash-Tree Press, 2001). See also the textual annotations to "A Warning to the Curious" in A Pleasing Terror (pp.341-355).


The manuscript of M.R. James's "A Warning to the Curious" was included, shortly after his death, in Sotheby's sale of November 1936, where it was purchased by "Elkin Matthews". Michael Cox's notes on the tale, in Casting the Runes and Other Ghost Stories (1987), describe the whereabouts of the manuscript as "untraced", but thanks to the alertness of Christopher Roden this has now changed. Hope Mayo, in her 1990 introduction to the anthology of Morgan Library Ghost Stories (Fordham University Press), elaborated on MRJ's friendly working relationship with the collector J. Pierpont Morgan and his private librarian (and eventual first director of the Pierpont Morgan Library), Belle da Costa Greene, in New York. MRJ prepared the descriptions of medieval manuscripts in the Morgan collection ("James worked in Cambridge, where the manuscripts were sent carton by carton to him"), which eventually appeared as part of a lavish catalogue published in 1906-07. Later, Greene, with Morgan's son J.P. Morgan Jr, tried (unsuccessfully!) several times to induce MRJ to cross the Atlantic and visit as a guest of the Morgan Library. In 1933, Greene wrote begging MRJ for "a new ghost story from your hand. Won't you give us one soon?". The Morgan Library Ghost Stories introduction concludes: "[but] the Library has had to content itself with the manuscript of the late story 'A Warning to the Curious', sold after James's death and presented as a gift by J.P. Morgan in 1942."

In January 2001, Christopher and Barbara Roden visited the Morgan Library in New York, and were able not only to examine but to obtain a photocopy of this manuscript which is, indeed, exactly what Hope Mayo says it is. Written in MRJ's unmistakably awful hand, in pencil on twenty-two foolscap sheets, it contains the full text of the story with almost all the published wording, but also with many alterations and deletions throughout (plus clarifications in black ink in a different hand, presumably as an aid to the typesetter). Among other significant insights it provides is the rather astonishing possibility that the tale took its final form in a very early - and possibly even a first - draft.

This is especially evident from two paragraphs deleted in their entirety (far the longest deletion). They come close to the end and begin: "We got the old man at the martello tower to go for help & we stopped by him [Paxton] till they came with the stretcher"; going on to briefly describe the ensuing events and inquest, and concluding: "He was so totally without connections that all the inquiries that were made came up against a blank wall". All of this is crossed out, then followed by an expansion of the same content into the last five paragraphs of the published version.

Everywhere in the text there are profuse smaller deletions and amendments, some of considerable interest. As readers will recall, during Paxton's final day of life, the narrator and his friend Henry Long agreed to meet that unfortunate gentleman after a game of golf and lunch at the links. They "had lunch there rather early, so as not to be late back". In the manuscript, on the other hand, they play a second round after lunch, getting back to the hotel "soon after three". They then arranged to meet Paxton half an hour later, after changing and having baths, but in the manuscript the fact that the pair "overstayed our half-hour" by ten minutes is emphasised, thus explaining why Paxton was so easily hoodwinked by the spectre of William Ager into thinking they had gone for their walk without him.

That Seaburgh in "A Warning to the Curious" is Aldeburgh in Suffolk is obviously no revelation. Even were the topographical descriptions in the tale not as clear as they are, MRJ's specific identification of Seaburgh as Aldeburgh in his preface to the Collected Ghost Stories (1931) has left no possible doubt. Nevertheless, it is amusing to note that at one point in the manuscript he actually wrote Aldeburgh instead of Seaburgh, only to cross out the "Alde" and replace it with "Sea"! More intriguing is the question of Froston, the village "only about five or six miles" from Seaburgh, whose porch contains the carving of the shield with the three crowns of East Anglia. Michael Cox in Casting the Runes identifies it as "probably Theberton, where the three crowns are conspicuous". Theberton is about six miles north of Aldeburgh, while Friston - surely the inspiration for the name - is about four miles north-west. However, in the manuscript, MRJ first wrote "Thor", before crossing through those letters and putting "Froston". It is unquestionably "Thor" and not "Theb": was he thinking of Thorpeness, three or four miles directly up the coast from Aldeburgh (and in the same general direction as Theberton)? Sadly, Thorpeness itself, largely created from scratch as a golfing resort around 1900, contains no ancient church.

Nowhere in the published story is a precise date given for the events, although it's made quite clear that William Ager, when alive, had guarded the single remaining crown during the Great War of 1914-18 (his grandfather Nathaniel performed the same duty during the Franco-Prussian "war of 1870"; his father William during the "South African War" - the Second Boer War of 1899-1902). Despite the evidence, some commentators and adapters have persisted in locating the occurrences in the tale near the turn of the century. True, to begin with, MRJ himself must have been undecided on a date: in the first paragraph of the main narrative, after "the particular thing that happened on our last visit", he added and then (presumably immediately) crossed through, "as long ago as 1905 it was". Having then settled on the next decade and inserted dates accordingly, he later went over the rest of the manuscript and crossed them all out, except in the flyleaf inscription in the Ager prayer-book, where the undeleted date for William Ager (the younger) is "1910" (which must have been altered sometime in the editing process to the published "19--"). On Ager's gravestone, instead of the eventual "died...19--, aged 28", there is, "died...1916, aged 28", with the "16" crossed through. Similarly, near the start of the narrative in the published version (in fact, straight after the above-mentioned reference to what "happened on our last visit"), we are told that "It was in April 19--". In the manuscript, the main text simply says "It was in April", while "1917" has been inserted (in MRJ's hand) above, only to be crossed out and replaced alongside by "19++".

In common with most people who read "A Warning to the Curious" carefully, I have always assumed that it was set a year or two after the end of the Great War, and perhaps four or five years before it first saw print in the London Mercury in 1925. If, as seems certain, MRJ's intention while writing the tale was to set it during the War, this gives an added urgency and piquancy to the replacing of the crown, at a time when the very real threat of "them[ing] with their ships..." was by no means over. Why did he decide to alter it? In 1930, when judging a ghost story competition in The Spectator, he wrote: " feels that almost anything might have happened in the War. It is the wrong setting to choose for a ghost story: you cannot make it more terrifying in that way." Perhaps, on thinking things over, MRJ felt uncomfortable with a war-time date for "A Warning to the Curious".


ADDENDUM by Christopher and Barbara Roden


In Casting the Runes and Other Ghost Stories, Michael Cox notes that the manuscript for the "A Warning to the Curious" is untraced, but that it was sold to Elkin Matthews at the Sotheby's auction in 1936. Whether Elkin Matthews was acting as an agent, or making a speculative purchase, is unknown.

There is no documentation with the manuscript to assist with the task of tracing its history during the period between its sale in 1936 and its subsequent presentation to the Morgan Library in 1942. The Library makes the manuscript available for inspection in three separate folders:

Handwritten sheets forming layout for the title and contents pages of the book edition of A Warning to the Curious and Other Ghost Stories, clearly from the files of Edward Arnold & Co. These sheets are not in James's hand.

Tear sheets of the story from its appearance in The London Mercury XII, 70 (August 1925), pp.354-365. These are heavily marked in pencil, presumably by the editor preparing the story for book publication.

The actual manuscript of the story. The Morgan Library's catalogue describes it thus:

Original autograph manuscript written in pencil.
22pp. fo unbound in tan cl. box with bl. leather label
Gift of Mr. J. P. Morgan, May 1942.

The manuscript is a quire (24 sheets) of unruled paper folded to foolscap size, and therefore forming 48 sides. The first 22 sheets are handwritten by MRJ in pencil, on rectos only. The final page has "A Warning to the Curious" written (not by MRJ) on the verso. The manuscript itself is in good condition, though there seems to have been some fading to the pencil handwriting in places. Thankfully, with the conservation techniques available at the Morgan Library, we can rest in the knowledge that it is likely to survive in its existing condition for several future generations.

Copyright © 2001 Rosemary Pardoe. Addendum © 2001 Christopher and Barbara Roden.

back to top
back to Ghosts & Scholars Archive
back to Ghosts & Scholars Home Page

Bar by Syruss