This book has two parts. Part one is an analysis of a little known story by Edgar Allen Poe, 1CThe Mystery of Marie Roget. 1D Part two is Poe 19s story itself.

Poe, as the author argues in part one, invented the modern detective story because, while other writers had included detectives in their fiction, Poe was first to write stories that focused primarily on the detection of a crime with a detective as the main character. He created a detective named C. Auguste Dupin who appears in three stories, 1CThe Murders in the Rue Morgue, 1D 1CThe Mystery of Marie Roget, 1D and 1CThe Purloined Letter. 1D Of the three, 1CMarie Roget 1D is the least satisfactory, but that does not mean it is bad. (The other two are more often read for the reason that they are great, or at least near-great, literature.) The character of Dupin should be of particular interest to fans of Rex Stout 19s Nero Wolfe mystery series, which was written a century later. Like Wolfe, Dupin has an assistant who goes around finding clues, which he then presents to the detective, who solves the crime without ever leaving the comfort of home.

What drew John Walsh into his thorough study of this long short story was that several Poe scholars and biographers have drawn attention to the fact that 1CMarie Roget 1D was based on a real murder case. There has been no clear consensus, argues Walsh, about whether Poe got the facts of the case and its solution right, but some prominent writers claim that Poe, writing before the case was solved, actually pulled a Dupin himself by solving the case before the police did, just by reading newspaper accounts. This myth began with Poe himself, who claimed that he had solved the crime. Balderdash, says Walsh.

First of all, the actual 1841 murder was never completely solved. Lost police records hamper a modern investigation, but, going by newspaper accounts, it appears that suspects (or accessories) might have been arrested but were apparently never convicted. Some kind of deal may have been struck. In the record, then, the case rather peters out than comes to a satisfactory conclusion.

Secondly, the latest (and most nearly final) facts in the case came out before the final installment of 1CMarie Roget 1D was published, not long after, as Poe claimed. To understand this, it should be remembered that in the nineteenth century, before movies, radio and television, many people got their entertainment by reading serialized stories in magazines. Many of the books that are now considered classics began life as a series of separately published magazine installments. (A good example of this is the work of Charles Dickens, which now appears in the form of several long novels, many of which were originally published in magazines, about a chapter at a time. Dickens is an interesting case for two more reasons: He was a friend of Poe with whom he discussed mystery writing, and Dickens 19s novel 1CBleak House 1D is an example of a work contemporary to Poe 19s that includes a murder mystery but does not make it the primary focus.)

Walsh switches back and forth between the real murder of Mary C. Rogers of New York in July 1841 and Poe 19s relationship to the case as he wrote the thinly disguised mystery based on it, but set in Paris. The actual homicide became a top story in the New York newspapers during August and September of 1841. Poe, who had lived in New York in the late 1830s, but who lived in Philadelphia in 1841 (then six hours away by train 14it 19s half that today), not only followed the case, but made at least one trip to New York before the final installment of his fiction in February 1843. Walsh makes a case that Poe also visited New York in 1842, during a period when it is uncertain whether Poe, who still resided in Philadelphia, was at home or not. Part of the uncertainty about Walsh 19s theory is that some people told a vivid story of meeting with Poe when he visited New York in the early 1840s but they are fuzzy on the date. Walsh argues that he could have done some research on the actual case during this visit. Poe had connections among news reporters who had background information that they were not able to print.

As Walsh tells it, the crisis for Poe came when, in early November 1842, just after the first of three installments of 1CMarie Roget 1D had been published in 1CThe Ladies Companion, 1D one of the witnesses made an ambiguous but nonetheless revealing deathbed confession of more knowledge about the case than had been released to the police and the press. This revelation rather spoiled Poe 19s story which he had already finished. It was too late to prevent the publication of part two in the December issue of the magazine, but, while there is no record of Poe having asked the magazine to postpone publication of the final installment, it is a fact that the last part was not published in the January issue but in the February one.

By analyzing the final installment, Walsh argues that far from solving the case and anticipating the deathbed confession, Poe rewrote the last part of his tale after the confession. Although the rewrite still left a few holes in the story as a whole, it did seem to the casual reader that the fictional story had anticipated all of the facts available at the time of publication. Poe further muddied the waters when 1CThe Mystery of Marie Roget 1D was published as part of his collection of stories in book form, 1CTales, 1D in 1845. For this edition, Poe added a footnote claiming that he had anticipated facts that did not come out until after he had written the story, when, in fact, those revelations came out between the publication of the first and last installments, giving Poe enough time to cover his tracks if only he had the collusion of the publisher.

This book 19s account of everyday life in New York circa 1840 is, to me, reason alone to read it. I confess to being a sucker for accounts of everyday life in historical periods. Among the amusing things to note is the fact that newspapers in those days were not above printing hoaxes deliberately. This made intelligent readers skeptical of what they read in the press. This often seems to me a healthy attitude that modern media consumers ought to cultivate for their own protection.

Of course, New York, New York, was a different place in 1840 from what it is today. A thriving city, it nevertheless was not as built up as it would become more and more over each succeeding decade. The New Jersey Shore, across the Hudson River from New York, was regarded as a more idyllic place than it would be today. New Yorkers got out of the city on hot summer days by going to New Jersey. One of the shocking things about the Mary Rogers case was that her body was found on the New Jersey side of the river by New Yorkers enjoying their weekend at the shore. It was noted that New York 19s criminal element also took weekends on the New Jersey side and brought their lawless ways to the otherwise more or less bucolic environs of Hoboken and Weehawken. At first, some assumed that her body had floated across the river from New York to New Jersey, but, as the evidence mounted, it became widely recognized that crime, heretofore more closely associated with New York, was just as likely to be found in New Jersey, where, as it turned out, Mary Rogers had actually died.

Another point of historical significance has to do with the publication of Walsh 19s book in 1967. He could compare the publication of magazine fiction in the 1840s and 1960s with much less nostalgia than would be evoked by a similar comparison made today. The opportunity for a new writer in 2010 to get fiction published in a magazine is not what it was in 1967 let alone in 1840.

The second part of the book reprints "The Mystery of Marie Roget" but with notes by Walsh. His argument is amply supported by Poe's text. For example, on page 134, Poe says, "There might have been a wrong here, or, more possibly an accident at Madame Duluc's." The problem is that up to this point in Poe's story, there has been no support for there having been any accident at Madame Duluc's. Poe pulls this out of his behind and never supports it. What is more, Walsh says that one of Poe's editors noted that this is an additions to Poe's original text. Poe's story provides a few such places where he seems to defy logic and hold out the possibility of two different solutions for his mystery: "a wrong" (code for murder) or "an accident" (code for a botched abortion). Also, while the talky Dupin's analysis of the case is intellectually very interestingly reasoned, I am not sure that modern readers will be up for so much philosophy, as pithy as much of it is. ( )