Edgar Allan Poe's
"The Mystery of Marie Roget" --
Poe's Second Detective Story!

by Drew R. Thomas

When, as a child, I first attempted to read "The Mystery of Marie Roget," I found it long, tedious, and boring. I am not alone in this assessment: In Murder for Pleasure, Howard Haycraft writes, "This longest of Poe's three major excursions into detective literature is, unhappily, the least deserving of detailed attention. It might be better called an essay than a story. As an essay, it is an able if tedious exercise in reasoning. As a story, it scarcely exists. It has no life-blood. The characters neither move nor speak. They are present only through second-hand newspaper accounts.... Only a professional student of analytics or an inveterate devotee of criminology can read it with any degree of unfeigned interest."

I have since read several books about the murder of Mary Rogers (upon whom the story is based) and have become fascinated; nay, nearly obsessed with the story. (Two books are Poe the Detective: The Curious Circumstances Behind the Mystery of Marie Roget by John Walsh and Who Murdered Mary Rogers? by Raymond Paul, [Prentice-Hall, Englewood, NJ, 1971]). What Poe has accomplished with this story is nothing short of brilliant and has given rise to a whole school of detective fiction -- that of the arm-chair detective who solves cases from newspapers or second-hand accounts alone (in many cases) and who seldom, if ever, visits the scene of a crime in order to investigate. In addition, it is the first work of fiction that purports to solve a real crime -- the murder of Mary Cecilia Rogers of New York City.  (You can read about the actual case here.)

Poe had already established his detective character -- the Chevalier C. Auguste Dupin -- in "The Murders in the Rue Morgue," which was entirely fictional. Poe writes in the third paragraph of "The Mystery of Marie Roget,"  "When, in an article entitledThe Murders in the Rue Morgue, I endeavored about a year ago, to depict some very remarkable features in the mental character of my friend, the Chevalier C. Auguste Dupin, it did not occur to me that I should ever resume the subject" (Tales of Edgar Allan Poe).

Since he had already established that Dupin lived in Paris, it was convenient to change the name of Mary Rogers to Marie Roget and to transport the events which actually took place in New York to Paris. Then he quoted voluminously from actual newspaper accounts as published in the New York newspapers.

Although Poe changed the names of the newspapers he quoted from, footnotes indicate the actual names of the specific papers, as well as actual names of the people who appear in the narrative. (He was sticking close to the actual case, after all, although John Walsh indicates that Poe added a half-dozen or so newspaper "quotes" which he fabricated. One presumes he manufactured these to support his own conclusions and to fill in the gaps to make the story more cohesive. Most of the quotes are, after all, derived from the actual newspaper accounts.)

Although Dupin analyzed the newspaper accounts and never left his arm-chair to investigate, Poe actually visited the scene of the crime. Dupin's (and Poe's) analysis is superb and he actually employs science to dispel myths and falsehoods that some newspaper accounts promulgated, so I suppose we can add "scientific analysis" to our list of "firsts" for this story.

Here's an example. The newspaper account which Dupin quotes from says:

"All experience has shown that drowned bodies, or bodies thrown into the water immediately after death by violence, require from six to ten days for decomposition to take place to bring them to the top of the water. Even where a cannon is fired over a corpse, and it rises before at least five or six days' immersion, it sinks again, if let alone. Now, we ask, what was there in this cave to cause a departure from the ordinary course of nature?" . . .

Dupin's analysis is much more scientific. He tells his narrator:

"...we can easily test by it the assertions of L'Etoile. 'All experience shows,'says this paper, 'that drowned bodies, or bodies thrown into thewater immediately after death by violence, require from six to tendays for sufficient decomposition to take place to bring them to thetop of the water. Even when a cannon is fired over a corpse, and itrises before at least five or six days' immersion, it sinks again iflet alone.'

"The whole of this paragraph must now appear a tissue ofinconsequence and incoherence. All experience does not show that'drowned bodies' require from six to ten days for sufficientdecomposition to take place to bring them to the surface. Bothscience and experience show that the period of their rising is, andnecessarily must be, indeterminate. If, moreover, a body has risen tothe surface through firing of cannon, it will not 'sink again if letalone,' until decomposition has so far progressed as to permit theescape of the generated gas. But I wish to call your attention to thedistinction which is made between 'drowned bodies,' and 'bodiesthrown into the water immediately after death by violence.' Althoughthe writer admits the distinction, he yet includes them all in thesame category. I have shown how it is that the body of a drowning manbecomes specifically heavier than its bulk of water, and that hewould not sink at all, except for the struggles by which he elevateshis arms above the surface, and his gasps for breath while beneaththe surface - gasps which supply by water the place of the originalair in the lungs. But these struggles and these gasps would not occurin the body 'thrown into the water immediately after death byviolence.' Thus, in the latter instance, the body, as a general rule,would not sink at all - a fact of which L'Etoile is evidentlyignorant. When decomposition had proceeded to a very great extent -when the flesh had in a great measure left the bones - then, indeed,but not till then, should we lose sight of the corpse.

"And now what are we to make of the argument, that the body foundcould not be that of Marie Rogêt, because, three days only havingelapsed, this body was found floating? If drowned, being a woman, shemight never have sunk; or having sunk, might have reappeared intwenty-four hours, or less. But no one supposes her to have beendrowned; and, dying before being thrown into the river, she mighthave been found floating at any period afterwards whatever.

" 'But,' says L'Etoile, 'if the body had been kept in its mangledstate on shore until Tuesday night, some trace would be found onshore of the murderers.' Here it is at first difficult to perceivethe intention of the reasoner. He means to anticipate what heimagines would be an objection to his theory - viz: that the body waskept on shore two days, suffering rapid decomposition - more rapidthan if immersed in water. He supposes that, had this been the case,it might have appeared at the surface on the Wednesday, and thinksthat only under such circumstances it could so have appeared. He isaccordingly in haste to show that it was not kept on shore; for, ifso, 'some trace would be found on shore of the murderers.' I presumeyou smile at the sequitur. You cannot be made to see how the mereduration of the corpse on the shore could operate to multiply tracesof the assassins. Nor can I.

" 'And furthermore it is exceedingly improbable,' continues ourjournal, 'that any villains who had committed such a murder as ishere supposed, would have thrown the body in without weight to sinkit, when such a precaution could have so easily been taken.' Observe,here, the laughable confusion of thought! No one - not even L'Etoile- disputes the murder committed on the body found. The marks ofviolence are too obvious. It is our reasoner's object merely to showthat this body is not Marie's. He wishes to prove that Marie is notassassinated - not that the corpse was not. Yet his observationproves only the latter point. Here is a corpse without weightattached. Murderers, casting it in, would not have failed to attach aweight. Therefore it was not thrown in by murderers. This is allwhich is proved, if any thing is. The question of identity is noteven approached, and L'Etoile has been at great pains merely togainsay now what it has admitted only a moment before. 'We areperfectly convinced,' it says, 'that the body found was that of amurdered female.' "

Mr. Walsh reveals that Poe's "The Mystery of Marie Roget" was to be published in three installments. The first two installments were published, but then the police announced that they solved the case. It appears that Poe was on to something but that his conclusion was at variance with the solution the police arrived at.

Walsh tells us that Poe suppressed the final installment, which delayed its appearing in print for another month. But it was already set in type, and Poe evidently made a few minor changes to align his solution with that announced by the police. (Mr. Walsh includes Poe's entire story with notes and annotations indicating precisely where the textual changes were made.)

After Poe died (in 1845), the story was collected along with others and published in Tales of Edgar Allan Poe. A concluding footnote made the followng claim:

Upon the original publication of "Marie Roget," the foot-notes now appended were considered unnecessary; but the lapse of several years since the tragedy upon which the tale is based, renders it expedient to give them, and also to say a few words in explanation of the general design. A young girl, Mary Cecilia Rogers, was murdered in the vicinity of New York; and, although her death occasioned an intense and long-enduring excitement, the mystery attending it had remained unsolved at the period when the present paper was written and published (November, 1842). Herein, under pretence of relating the fate of a Parisian grisette, the author has followed in minute detail, the essential, while merely paralleling the inessential facts of the real murder of Mary Rogers. Thus all argument founded upon the fiction is applicable to the truth: and the investigation of the truth was the object. The "Mystery of Marie Roget" was composed at a distance from the scene of the atrocity, and with no other means of investigation than the newspapers afforded. Thus much escaped the writer of which he could have availed himself had he been upon the spot, and visited the localities. It may not be improper to record, nevertheless, that the confessions of two persons, (one of them the Madame Deluc of the narrative) made, at different periods, long subsequent to the publication, confirmed, in full, not only thegeneral conclusion, but absolutely all the chief hypothetical details by which that conclusion was attained.

Haycraft says the murder was never solved, contrary to popular misconception. But Walsh disputes this.

Haycraft categorizes "The Mystery of Marie Roget" as a "mental" detective story because there is no story per se, no development of plot, and no action. It is all pure analysis of the newspaper accounts.

It is, however, a steppingstone to the next stage in the development of the detective, crime, and murder mystery genre.

In itself, in spite of its lack of luster and tediousness it is, somewhat oxymoronically, brilliant. It has influenced writers from Arthur Conan Doyle  right up to television programs of today (such as Law and Order, which often relies on actual crimes in the news for various plots!). (Sherlock Holmes's brother Mycroft, who appeared in "The Greek Interpreter," and "The Bruce-Partington Plans," was an armchair detective who seldom left his chair in the Diogenes Club. When Mycroft did so, it was solely upon occasions of momentous import.) Baroness Orczy, creator of "The Scarlet Pimpernel," also created her armchair detective, known as "The Old Man in the Corner" -- you can read one of these ("The Millionaire in the Dock") for free in the Reading Room.

Where would we be today without Poe's contribution? Yes, I know that some of you are likely to say that sooner or later someone would do it. This is undoubtedly so; but it is, after all, Edgar Allan Poe who did explore this new direction. Not content to simply repeat his past success, he pushed forward into new territory. That he conceived this idea, in the first place, and that he didn't stop there but continued even further to create the perfectly balanced story (with "The Purloined Letter") is remarkable indeed.

Genesis: 1841 | Home

Click on the Slideshow below to Buy from Amazon: