Ciphers of the Revolutionary Period (1917, Burnett)

From Kook Science

Burnett, E. C. (Jan. 1917), "Ciphers of the Revolutionary Period", American Historical Review 22 (2): 329-334, 

Ciphers of the Revolutionary Period

During the Revolutionary period cipher was employed extensively not only in public correspondence where secrecy was especially important but in the private correspondence of public men as well. It is true that most of the letters written in cipher that have come down to us are accompanied by some form of translation, oftenest an interlinear decipherment by the recipient; yet the quantity of writing that has remained undeciphered is in the aggregate considerable. There are, for instance, numerous undeciphered passages in the published writings of Jefferson, Madison, and Monroe, as well as in letters of theirs that have not been printed.[1]

Of ciphers used in the period under consideration five different types have been encountered: first, a mere transposition of the alphabet; second, a dictionary or book cipher; third, a sentence or longer passage used as a key, the letters being numbered in the order in which they occur; fourth, a columnar table of alphabets built upon a key-word; and, fifth, a collection of words, syllables, and letters more or less arbitrarily arranged and numbered. The transposition of the alphabet, a mere substitution of one letter for another, was not apparently much used.[2]

The book cipher was employed rather extensively, particularly in the earlier part of the period.[3] In its simplest form its use required only the possession, by each of the correspondents, of the same edition of a dictionary. A notable instance of a dictionary cipher is found in the correspondence of Arthur, Richard Henry, and William Lee in the years 1777-1779. As used by them an arabic numeral designates the page, an a or a & the column, and a roman numeral the line. The book was probably Entick's New Spelling Dictionary, edition of 1777.[4] As no copy of the edition used could be found the process of solution by the present writer consisted in, first, the identification of certain ciphers from the context,[5] next, locating these words in the edition of 1782 (the nearest to 1777 obtainable), and then, by a process of approximation, determining other words.[6]

Of the third sort of cipher mentioned the principal example come upon is that used by C. W. F. Dumas. It is evidently this code that is found in the Franklin Papers in the American Philosophical Society.[7] The key is a long passage in French running to 682 letters, numbered consecutively. In such a cipher each letter has several numbers corresponding to it, which may be used indifferently.[8]

The fourth form of cipher is made by taking a key-word and constructing columns of alphabets beginning with the letters of the word in the order of their occurrence and numbered from 1 to 26 or 27 (when 27 letters are used the character & follows Z). In writing in this cipher the letters are sought in the columns successively and the corresponding numbers are used. This form of cipher seems to have been introduced by James Lovell when he was a member of the Committee of Foreign Affairs, and was used by him in letters to John Adams, Mrs. Abigail Adams, and others.[9] Livingston, when he became secretary for foreign affairs, used the same cipher in his first correspondence with Adams,[10] and Jefferson used such a cipher in some earlier correspondence with William Short.[11]

The most noteworthy series of letters in a cipher of this sort were written by Madison to Edmund Randolph in the summer and autumn of 1782.[12] These letters have never hitherto been deciphered. In fact Randolph himself was never able to decipher them, owing partly, no doubt, to certain errors which Madison made in writing the cipher. Although these errors occasioned some difficulties the solution was accomplished through the successful guessing of the cipher for "commission", from which the alphabetical table was reconstructed.[13]

It was the fifth type of cipher that came to be most generally employed. Such a cipher might consist of only a comparatively few numbers for persons, places, etc., or it might run to hundreds of items. While some ciphers of this type were sent abroad by Charles Thomson and Robert Morris in 1780 and 1781, not many examples of its use have been found prior to the autumn of 1781, when Robert R. Livingston became secretary for foreign affairs, after which individuals of the type were rapidly multiplied both for public and for private use. Livingston had some forms printed, having on one side of the sheet the numbers from 1 to 1700, on the other the alphabetical list of words, syllables, etc. These forms were a convenient basis on which correspondents could prepare their identical codes.[14]

The earliest of these numerical codes which the writer has come upon is that used by the Virginia delegates to Congress in 1782 in their official correspondence with the governor. It is not on a printed form and runs to only 846 numbers.[15] Madison and Randolph used this code to a considerable extent in their private correspondence also in that year.[16] The several ciphers used by Jefferson, Madison, and Monroe from 1783 on require some explication. In January and February, 1783, Jefferson and Madison used a dictionary cipher.[17] From April, 1783, to May, 1785, they used a numerical cipher, the key to which has not been found. Of some of these letters however there is a decipherment or translation (they are chiefly Madison's), and from these a code has been reconstructed by which the others have been deciphered.[18] In the same period Monroe used in a few letters to Madison a numerical code of very limited extent, of which most of the ciphers used are interpreted in the texts of the letters. From May, 1785, to May, 1786, however, Madison and Monroe used a cipher for which no code has been found. In this case also it has been possible, by means of such decipherments as exist, to decipher (with the possible exception of occasional words) those letters for which there is no translation. The correspondence of Jefferson and Monroe from May, 1784, to March, 1785, offers some difficulties, but as the codes by which these letters were written are in existence[19] a careful attention to the several explanations of the writers enables one to overcome these difficulties. In the spring of 1785 Jefferson prepared a new and more extensive code on the printed forms referred to above, which was thenceforward used in his correspondence with both Madison and Monroe and in theirs with him.[20]

One incidental discovery, although somewhat afield from this particular investigation, deserves nevertheless to be recorded here. A short while ago a professor in a western university sent to the Department of Historical Research a body of letters from President Jackson to a diplomatic agent, of the year 1832, written in cipher, and asked whether some means might not be found of deciphering them. The department happened to have a cipher code, constructed on one of the printed forms heretofore referred to, found among the Monroe Papers in the New York Public Library, without date, and merely endorsed: "Mr. Monroe's cypher". It was found upon test that the Jackson letters were written in this code. It was further discovered that the same code was used by James A. Bayard when he was one of the commissioners for negotiating the Treaty of Ghent. It has since been learned that Monroe used this code in 1805 when he was minister to England. It was evidently therefore an official cipher.

Edmund C. Burnett.


  1. This note is not to be understood as in any sense a complete exposition of the use of cipher in the Revolutionary period, some of the examples referred to being indeed merely incidents of an investigation which has had, naturally, considerable ramifications.
  2. See however the Letters of William Lee (ed. Ford), II. 417, 666; also the letter of Jay to Morris mentioned in note 6 below.
  3. In a letter of June 3, 1776, to the committee of secret correspondence, Arthur Lee proposed the use of a dictionary for cipher correspondence. (See Force,Am. Arch., 4th sen, VI. 686, and Wharton, Rev. Dipl. Corr., II. 95.) The original of the paragraph concerning the cipher, written upon the fly-leaf of a small book and bearing an endorsement by James Lovell, is preserved in the Bureau of Rolls and Library (Pap. Cont. Cong., no. 53, I. 21).
  4. Arthur Lee sent a copy of the book to his brother, Richard Henry Lee, in the autumn of 1777. See his letter of November 25, 1777,Life of Arthur Lee, II. 117. In this work theb of these ciphers is usually printed as6. Concerning this and other ciphers used by William Lee, see an editorial note in the Letters of William Lee, II. 417, and passim.
  5. There could be no question, for instance, that 115 b xxxviii stood for "Deane".
  6. Instances of a different method of employing a dictionary cipher are found in letters of William Carmichael and John Jay in 1780 and 1781, and in a few letters of Jefferson and Madison in January and February, 1783. In a letter to Robert Morris, November 19, 1780 (N. Y. Hist. Soc,Collections, Revolutionary Papers, I. 451), Jay suggests the use of Entick's dictionary paged backwards, to be supplemented by the use of a transposed alphabet.
  7. Franklin Papers, L (i), 24. There is another cipher code in the same collection of Franklin Papers (LXI. 1), which consists merely of a collection of words alphabetically arranged and numbered.
  8. A cipher similar to that of Dumas was proposed to Franklin by Barbe Dubourg, June 10. 1776. See Force, Am. Arch., 4th ser., VI. 782.
  9. Letters in which it is used are found in the Adams Manuscripts, June to December, 1781. The key, as suggested by Lovell, was "the first sixth part of that family name where you and I spent our last Evening with your Lady before we sat out on our Journey hither." The key turns out to be "C R". The name was probably Cranch. A letter from Lovell to General Gates, March 1, 1779 (N. Y. Hist. Soc, Gates Papers), uses the key-word "James".
  10. It seems however that Adams did not quite understand the cipher. See Wharton, Rev. Dipl. Corr., V. 73, 192, 459.
  11. See Southern Bivouac, new series, II. 425. The alphabetical table used by Jefferson, with the key-word "Nicholas", is given, ibid., II. 427.
  12. Some of Madison's letters to Randolph at this time were written in the type of cipher next described, others partly in the one and partly in the other. Randolph himself had suggested that they use "the cypher which we were taught by Mr. Lovell. Let the keyword be the name of the negro boy who used to wait on our common friend Mr. Jas. Madison." In a foot-note to the letter Madison says: "probably Cupid". Randolph to Madison, July 5, 1782, Library of Congress, Ac. 1081.
  13. Madison's foot-note giving the key (see note 12, above) was found subsequently.
  14. A good many ciphered passages in the diplomatic correspondence of the period remain undeciphered. In particular may be mentioned the letters of Livingston to Jay from November 1, 1781, to April 16, 1782. See Wharton, Rev. Dipl. Corr., IV. 716, 814; V. 29, 44, 144, 149, 150, 374, 404. Livingston's letter of April 16, almost wholly in cipher, has not been printed. It is in the Bureau of Rolls and Library, Pap. Cont. Cong., no. 79, vol. I.
  15. The code is found among the Executive Papers in Richmond.
  16. See ante, note 12. Beginning with Madison's letter of December 30, 1782, they adopted a new code of the same kind.
  17. Only a limited effort has been made to identify the dictionary used, but by the process of approximation used in the case of the Lee correspondence the undeciphered ciphers may, with at least a high degree of probability, be solved. For instance, in Jefferson's Writings (ed. Ford), III. 310, a word is "lacking". The cipher is 369.9 and doubtless means "frivolous".
  18. In the Writings of Jefferson (ed. Ford) some attempts toward decipherment have been made, but with indifferent success. Not to speak of erroneous renderings of ciphers, some mistaken editorial interpretations call for correction. A foot-note to Jefferson's letter to Madison. March 18, 1785 (Writings, IV. 35), suggests that the paragraph relates to Patrick Henry. Jefferson is actually speaking of Lafayette. In his letter of August n, 1793 (VI. 367), he says: "Just as I had finished so far, 812.15 called on me." A foot-note says: "Edmund Randolph". The cipher means, "the President", that is. Washington. In the letter of April 25, 1784 (III. 470) several wrong renderings give quite erroneous suggestions.
  19. In the Jefferson Papers, 2d ser., vol. LVIL, fol. 17a, are three codes, one marked "1st cypher", another "2d cypher", and a third endorsed: "Cypher sent in Col. Monroe's lre of April 12, 1785". In fact, the latter is a copy of a cipher sent to Jefferson by Monroe July 20, 1784. At the same place is found the alphabetical part of the "2d cipher".
  20. One copy of this code is in the Jefferson Papers. 5th ser., vol. XI., fol. 35, and another in the Monroe Papers, vol. XXII., fol. 2926. See Jefferson to Monroe, March 18, 1785, and to Madison, May 11. Jefferson was still using this cipher with Madison in 1793.

Bibliography of Cryptography: a Catalog of Books Pertaining to the Science of Codes And Cyphers. (Cincinnati, 1938)