Text: Clarence S. Brigham, “[Introduction],” Edgar Allan Poe’s Contributions to Alexander’s Weekly Messenger (1943), pp. 3-11 (This material is protected by copyright)


[page 3, unnumbered:]

Edgar Allan Poe’s Contributions to
Alexander’s Weekly Messenger


THE fame of Edgar Allan Poe rests in a considerable degree upon his activities in the field of ciphers and the solution of cryptograms. As one critic has said in a volume on Poe: “Doubtless nothing contributed to a greater extent than did Poe’s connection with cryptography to the growth of the legend which pictured him as a man at once below and above ordinary human nature; but the whole subject is still unfortunately wrapped in some obscurity, and it is impossible to be sure of the facts as distinguished from his own report of them.”(1)

Poe’s career as a solver of cryptograms began during his connection with a Philadelphia newspaper, Alexander’s Weekly Messenger, from December 1839 to May 1840, during which period he claimed to have solved without a failure all of the ciphers submitted to him. But the want of a file of this paper made it impossible to substantiate Poe’s boast. Finding Poe’s statement in Graham’s Magazine in 1841 that he had solved many ciphers the previous year in a Philadelphia paper called Alexander’s Weekly Messenger, Poe students such as Professor John M. Manly, Hervey Allen, and many others, searched in vain for such a newspaper. Colonel William F. Friedman of the United States War Department Signal Office, in an article entitled “Edgar Allan Poe, Cryptographer,”(2) said: “Unfortunately the records that [page 4:] remain of Alexander’s Weekly Messenger are exceedingly fragmentary. Despite painstaking research by numerous Poe experts, not a single issue containing any cipher solutions that Poe may have published as a result of his asserted challenge has ever been found, and there seems to be no way at the present moment of corroborating Poe’s statements.”

The recent acquistion by the American Antiquarian Society of a complete file of Alexander’s Weekly Messenger for 1840 now provides a means to test the truth of Poe’s assertions, and incidentally reveals other of his writings contributed to that paper.

Poe’s first statement on the subject was in an issue of the Messenger for December 18, 1839, the only known copy of which is in the Library of the Ohio State Archaeological and Historical Society. In this issue, under the heading of “Enigmatical and Conundrum-ical,” he alluded at length to the value of applying a rigorous and analytical method in the solution of enigmas. He asserted that if any reader submitted an example of secret writing in which arbitrary symbols were substituted for letters of the alphabet, no such cipher could be propounded which he would be unable to solve. In discoursing upon conundrums, he offered a collection of twenty-five examples, which “have at least the merit of originality,” and ingeniously gave as his last conundrum: “Why ought the author of the ‘Grotesque and Arabesque’ to be a good writer of verses ? Because he’s a poet to a t. Add t to Poe makes it Poet.”

In Graham’s Magazine for July 1841 Poe contributed a signed article entitled “A Few Words on Secret Writing.” Following a paragraph on the method of solving cryptographical puzzles, he said: “In the discussion of an analogous subject, in one of the weekly papers of this city, about eighteen months ago, the writer of this article had occasion to speak of the application of a rigorous method in all forms of [page 5:] thought — of its advantages — of the extension of its use even to what is considered the operation of pure fancy — and thus, subsequently, of the solution of cipher. He even ventured to assert that no cipher, of the character above specified, could be sent to the address of the paper, which he would not be able to resolve. This challenge excited, most unexpectedly, a very lively interest among the numerous readers of the journal. Letters were poured in upon the editor from all parts of the country; and many of the writers of these epistles were so convinced of the impenetrability of their mysteries, as to be at great pains to draw him into wagers on the subject. At the same time, they were not always scrupulous about sticking to the point. The cryptographs were, in numerous instances, altogether beyond the limits defined in the beginning. Foreign languages were employed. Words and sentences were run together without interval. Several alphabets were used in the same cipher. One gentleman, but moderately endowed with conscientiousness, inditing us a puzzle composed of pot-hooks and hangers to which the wildest typography of the office could afford nothing similar, went even so far as to jumble together no less than seven distinct alphabets, without intervals between the letters, or between the lines. Many of the cryptographs were dated in Philadelphia, and several of those which urged the subject of a bet were written by gentlemen of this city. Out of, perhaps, one hundred ciphers altogether received, there was only one which we did not immediately succeed in solving. This one we demonstrated to be an imposition — that is to say, we fully proved it a jargon of random characters, having no meaning whatever. In respect to the epistle of the seven alphabets, we had the pleasure of completely nonplus-ing its inditer by a prompt and satisfactory translation.”(3)

In Graham’s Magazine for December 1841, in another [page 6:] article on “Secret Writing,” Poe refers a correspondent “to the files of ‘Alexander’s Weekly Messenger’ for 1839 — where he will see that we read numerous ciphers of the class described, even when very ingenious additional difficulties were interposed.”(4)

Since all of the Poe articles in the Messenger from December 18, 1839, to May 6, 1840, are herewith reprinted, the reader can judge for himself whether Poe justified his claims. The solved puzzles appear in almost every issue from January 15 to April 29, under such headings as “Enigmatical,” “Another Poser,” “More of the Puzzles,” “Cyphers Again.” In all, thirty-six ciphers were recorded as received.

For a summary of the articles on ciphers and puzzles in the Messenger, I have been favored with a letter from Dr. W. K. Wimsatt, Jr., of Yale University, which with his permission is included in this introduction. Dr. Wimsatt has given much study to Poe’s interest in cryptograms, and an article by him entitled “What Poe Knew about Cryptography” is scheduled to appear in Publications of the Modern Language Association of America.

The cryptograms (or hieroglyphical writings, puzzles, posers, cyphers, as Poe calls them) in Alexander’s would appear to be all of a very simple sort, even simpler than the ones he later solved in Graham’s. My statistics are as follows: in the fifteen 1840 Alexander’s articles on ciphers Poe refers altogether to thirty-six ciphers which he has received in response to his challenge of December 18, 1839, or to renewals of it in the later articles. Poe prints the text and solutions of nine ciphers — all of these simple substitution ciphers, i.e., composed in accordance with the formula which he gave in the note at the end of the December 18, 1839, article, and which he repeats once or twice in the later articles. (A simple substitution cipher is one in which the same symbol, be it of whatever sort, invariably stands for the same letter of the alphabet in the concealed message.) Poe prints the solution, but not the text, of fourteen ciphers (and part of the solution of one other, February 19), all of which there is every reason to suppose were also simple substitution ciphers. Three times Poe simply makes the statement that he has solved a cipher, and again there is no reason to suppose anything beyond simple substitution [page 7:] (cipher of D.D. and that of J. H., February 26; that of Hamilton Brown which Poe says cannot be set in type — implying, I believe, that he has solved it — April 8). Six times Poe says that he has not solved a cipher — for one reason or another (Munger’s cipher, written in pencil and “defaced,” cipher of J. R. H., composed of 51 characters and hence not a simple substitution cipher, and Kulp’s “imposition,” all of February 26; imposition of “Incog,” March 25; second cipher of Colfax, which Poe had no time to look at, and cipher from Austinburg, which Poe has lost, April 29). In one case (Feburary 26) Poe prints text and solution of what is not a cipher at all, but a mere jumble of two sequences of phrases, one of which makes sense if the other is ignored. Poe italicizes the words to be read.

I have reserved two cases for the last, as being the only ones which stand out from the others in point of difficulty: 1). February 26, cipher from J. H., text and solution. As Poe complains, J. H. has used some symbols, to stand for more than one letter of the “plain text.” This, then, is a simple substitution cipher with certain indeterminacies. The key-phrase ciphers which Poe was later to solve in Graham’s involved the same difficulty in aggravated form. As Poe was to learn, and as Col. William F. Friedman has clearly explained, a sufficient number of inde-terminacies can make a cipher insoluble even to him who has the key. But this cipher from J. H. is relatively simple even with the indeterminate symbols (none stands for more than two letters). The ones which Poe was later to solve in Graham’s were difficult for an amateur, but, as Col. Friedman reports, were not difficult for students of elementary cryptography to whom they were submitted.

2).  February 19, the seven-alphabet cipher of which Poe speaks in the July 1841 Graham’s article. Col. Friedman has pointed out that such a cipher may be only seven simple-substitution ciphers strung together. Only if the alphabets changed with every letter (as in the Vigenere cyclic ciphers, the chiffre quarre and its various modifications) would there be a greatly increased difficulty. It is unfortunate that Poe did not print the cipher text, but from the nature of the “plain text” which he prints I am completely confident that he did not solve a true cyclic cipher. As Poe says, the “puzzle is nothing more than the well-known acrostic called ‘The Siege of Belgrade,’ beginning thus:

An Austrian army, awfully arrayed,

Boldly, by battery, besieged Belgrade,”

where each line is composed of words beginning with the same letter, and the order of lines is alphabetical. It looks very much as if the seven alphabets of the cipher were simply arranged one to a line in rotation. Very likely the alliteration gave the thing away to Poe. I do not deny that he was very quick at picking up clues of this sort. But the methodical, more abstruse, analysis of cyclic ciphers he seems not to have known anything about. [page 8:]

Here are some detached observations:

1.  Poe’s correspondents often made the task easy for him by choosing some well-known text, the Lord’s Prayer, the opening lines of “Twelfth Night,” the above acrostic, or a verse enigma where the pattern helped to betray the secret or one part quickly led to another. On the whole his correspondents were extremely naive and had little conception of how to make their puzzles difficult within the limits he had set down.

2.  Note that the first cipher of April 22 is constructed on the simple mnemonic scheme of numbering the letters of the alphabet, 1, 2, 3, etc., and that Poe has solved it with great carelessness.

3.  D. D.’s cipher of February 26, “the letters being formed upon a square with diagonal crosses,” was probably no more than a simple cipher. Some mnemonic system was used for forming a symbol alphabet — and this far from adding any difficulty may only have served to give the thing away immediately. The fact of simple substitution is not altered by any key scheme whatsoever.

4.  Three times (February 26, March II, April 29) Poe complains about letters being run together (not separated according to the words of the plain text). He stipulated not only for simple substitutions but for every aid in the arrangement of the symbols. (There is no reason why those who possess the key of a cipher should not always arrange the symbols in false groups, as is always done in cyclic ciphers nowadays.)

Poe contributed other articles and paragraphs to the Messenger during his four months’ connection with the paper, all of which were unsigned. We have reprinted all which bear a resemblance to Poe’s style or for some other reason might be attributed to him.(5) Poe’s connection with the paper apparently stopped with the issue of May 6, 1840, after which date there are no more ciphers, or articles in any way suggestive of Poe’s style. The issue of May 6 was the last to carry upper and lower case headlines for its editorials, and thereafter the typographical style of the paper suffered a decided change. It was in this issue, too, that the column on theatrical affairs, which was evidently conducted by [page 9 Burton and which had been carried for several months, made its last appearance. It was in the spring of 1840 that Poe broke with William E. Burton, and presumably at the same time with Charles Alexander, who printed Burton’s Gentleman’s Magazine.(6) Also it was on May 4, 1840, that Alexander established The Daily Chronicle, and thus became the publisher of a daily as well as a weekly paper.

Charles Alexander, the publisher of Alexander’s Weekly Messenger, was one of the most prolific of Philadelphia newspaper publishers. Graduating as a printer from Poulson’s Daily Advertiser, he joined with Samuel C. Atkinson in establishing the Saturday Evening Post, August 4, 1821. He remained a member of the firm of Atkinson & Alexander until March 1828. The same firm established a literary periodical called The Casket in January 1826, publishing the magazine in connection with the Saturday Evening Post and frequently using the same material for both. When Alexander dropped out of the Post in March 1828, he also left the Casket.

On April 7, 1828, Alexander established The Daily Chronicle, which in 1834 he sold to James Gordon Bennett, and it soon after expired. He also claimed that he “laid the foundations” of Godey’s Lady’s Book, and although this popular magazine did not carry his name in the imprint, it was at first printed from the office of The Daily Chronicle. On November 29, 1834, he issued a specimen number of The Gentleman’s Vade-Mecum, which sporting and dramatic journal began in January 1835 and ran until June 1836. On January 2,1836, he established a bi-weekly illustrated journal entitled The Salmagundi, and News of the Day. Another of his ventures was Everybody’s Album, a popular monthly magazine, started by him in July 1836 and lasting one year. [page 10:]

The Salmagundi appeared under that title until December 28, 1836, and is interesting in that it was directly succeeded by the American Weekly Messenger, with the first issue on January 4, 1837. This became Alexander’s longest journalistic venture, changing its title on January 3, 1838, to Alexander’s Weekly Messenger, early in 1842 to Alexander’s Express Messenger, and on January 6, 1847, to Alexander’s Pictorial Messenger. It was published by him until November 1848, when it was continued by Samuel D. Patterson as the Family Messenger and National Gleaner. Files of Alexander’s Messenger are scarce, with the Antiquarian Society having the most issues from 1837 to 1844, and the Wisconsin Historical Society from 1844 to 1848.

In July 1837 Alexander published for William E. Burton the Gentleman’s Magazine and so continued until January 1839, when Burton took it over as publisher as well as editor. Alexander could not long exist without publishing at least two papers at the same time. On May 4, 1840, he took in a partner, Andrew Scott, to publish a daily paper as well as a weekly, and called it The Daily Chronicle. An editorial announcement in the first issue states that “The Lady’s Book, The Casket, The Gentleman’s Magazine, the Saturday Evening Post, the Saturday Courier, and the Weekly Messenger — of all of which the foundations were laid by the senior member of our firm — are examples of the success which has attended our endeavors.” On January 1, 1845, the firm was dissolved and Alexander became sole publisher. The last known issues are in 1846, and the paper probably ceased in 1847. The Antiquarian Society has the only known file of the Daily Chronicle from 1840 to 1845. Charles Alexander’s name is listed in the Philadelphia Directories until 1850, and thereafter disappears. From 1835 to 1850 his residence was given as 141 North 10th Street.

To return to Poe. That the subject of cryptography greatly [page 11:] intrigued him, there is no question. He was an exponent of the process of exact thinking, talked about “ratiocination,” and defended the use of technical devices even in the most fanciful of literary efforts. He was also interested in various forms of mystification — mesmerism, reading of character by handwriting, and phrenology. His was a strange mind, able to produce stories based upon the wildest flights of imagination, poems which seemed to spring from dreams and fantasies, and yet a mind which pleaded for use of pure and unimaginative logic in literary composition. Whether his interest in cryptography was the result of an attraction for the supernatural and the mysterious, or whether it came from his desire to parade his learning, or whether he possessed the power of an extraordinary analytical mind,(7) is for each reader of Poe to decide for himself.

As a preliminary to writing the notes in connection with Poe’s contributions to the Messenger, I have read — or rather, re-read — all of Poe’s printed works. I have consulted the magazines, nearly all of which are in the Antiquarian Society’s collection, to which Poe contributed during his stay in Philadelphia. I have also examined rather thoroughly the following Philadelphia newspapers, owned by the Antiquarian Society, from December 1839 to May 1840: Public Ledger, Philadelphia Saturday Courier, National Gazette, and Saturday Evening Post. I have received aid from many Poe scholars — notably Professor Arthur H. Quinn, William K. Wimsatt, Jr., and Thomas Ollive Mabbott. Dr. Mabbott especially has shown an active interest in this article and has contributed a detailed and authoritative opinion regarding Poe’s contributions to the Messenger. Without the help of these scholars my notes would have been of far less value.


[The following footnotes appear at the bottom of page 3:]

1.  Joseph W. Krutch, Edgar Allan Poe, a Study in Genius, 1926, p. 103.

2.  American Literature, November 1936, vol. 8, p. 271.

[The following footnote appears at the bottom of page 5:]

3.  Graham’s Magazine, July 1841, vol. 19, p. 34.

[The following footnote appears at the bottom of page 6:]

4.  Graham’s Magazine, [[December 1841,]] vo1. 19, p. 308.

[The following footnote appears at the bottom of page 8:]

5.  The work of a writer like Poe, with known hobbies and a habit of self repetition, is not difficult to identify in most cases, and especially so in a newspaper with a small staff including none of his close followers. Dr. Mabbott and I searched the file independently at first, and found unanimity of opinion upon consultation. My careful re-reading of the paper has brought to light only three additional items. “Thomas Paine” was regarded by both searchers as doubtful, and is the only item which seems hopelessly indeterminate.

[The following footnote appears at the bottom of page 9:]

6.  The disappearance of the theatrical column early in May would indicate a change in Burton’s connection with Alexander. This could have preceded Poe’s break with Burton. There is no evidence that the two breaks were related.

[The following footnote appears at the bottom of page 11:]

7.  Henry B. Hirst, in his biography of Poe, in the Philadelphia Saturday Museum, of March 4, 1843, enlarges at length on Poe’s analytical mind and his ability in solving ciphers. In the same paper Thomas C. Clarke refers editorially to Poe’s perceptive mind and says that Poe, when in his company, solved immediately a difficult cipher which had been printed in the Baltimore Sun, and which he reprints, with the solution.



In the original printings, the date of the article in footnote 2 is given incorrectly as 1836. The journal American Literature was not even in print that early. Instead, this date is obviously an error for 1936, and has been corrected in the current presentation, with this note as documentation of this minor editorial change.


[S:0 - CSB43, 1943] - Edgar Allan Poe Society of Baltimore - Articles - Edgar Allan Poe’s Contributions to Alexander’s Weekly Messenger (C. S. Brigham) (Introduction)