Benjamin Franklin's Codes and Ciphers

Dumas' Passage Cipher (WE041)

When the Revolutionary War broke out in Lexington and Concord in April 1775, Benjamin Franklin, almost 70, had already achieved an international fame as a scientist. When the Committee of Secret Correspondence was set up in November to seek foreign connections, Franklin was its most prominent member. To begin with, Franklin wrote to his correspondent Charles William Frederick Dumas [Charles Guillaumes Frédéric Dumas], a scholar residing in the Netherlands and an ardent supporter of the American cause, and asked him to serve as an American agent.

Dumas accepted the offer and sent a cipher to Franklin in his letter dated 30 April 1776. Dumas' cipher was a passage cipher in which a 682-letter passage taken from Dumas' own writing in French served as the key. Each letter of the passage was assigned numbers 1-682. Thus, a letter can be enciphered in as many ways as the letter occurs in the passage. For details, see here.

Dumas' early use of this cipher includes:

Dumas to Committee of Secret Correspondence, 14 May 1776 (PCC Roll 121, Page 36)
Dumas to Unknown, 30 June 1776 (PCC Rol 121, Page 44)
Dumas to Unknown, Sept. 1-Oct.10, 1776 (PCC Roll 121, Page 52)

Soon after the declaration of independence, Franklin was named as one of the three commissioners for obtaining French aid. Leaving Philadelphia on 26 October, he reached France on 4 December. Franklin had told Dumas to continue to report to the Committee of Secret Correspondence (24 October). The Committee, now consisting of Robert Morris and three others, told Dumas that they took over a cipher from Franklin.

We pray you to have frequently your news; and if you make use of cipher, the Doctor [Franklin] has conveyed the key to one of our members.
(Committee of Secret Correspondence to Dumas, 24 October 1776; back translated from French by the author)

Thus, Dumas continued to write to America, while he corresponded with Franklin in Paris. In the mean time, the Committee of Secret Correspondence was renamed the Committee of Foreign Affairs as of 17 April 1777. Dumas used the cipher in the following letters.

Dumas to Committee of Secret Correspondence, 12 April 1777 (PCC Roll 121, Page 62)
Dumas to the Courier of Bas-Rhin (extract), July 1777 (PCC Roll 121, Page 138)
Dumas to Committee of Foreign Affairs, 2 August 1777 (PCC Roll 121, Page 138)
Dumas to Franklin, 14 August 1777 (PCC Roll 121, Page 100)
Dumas to Committee of Foreign Affairs, 24 February 1778 (PCC Roll 121, Page 158)
Dumas to Committee of Foreign Affairs, 28 April 1778 (PCC Roll 121, Page 172)
Dumas to Committee of Foreign Affairs, May 1779 (PCC Roll 121, Page 276)
Dumas to Committee of Foreign Affairs, 21 June 1779 (PCC Roll 121, Page 288) (PCC Roll 121, Page 272)

This cipher was also used to explain Lovell's cipher and Franklin would repeatedly express his preference of this cipher, as discussed below. It is noted that Franklin used this cipher as late as in 1781:

Franklin to Dumas, 16 August 1781 (PCC Roll 127, Page 206)

Barbeu-Dubourg's Passage Cipher (WE072)

Franklin's fame was such that a collection of his works went into a fourth edition in 1769, which was translated into French by Jacques Barbeu-Dubourg in 1773.

Barbeu-Dubourg, in his despatch written during 10 June to 2 July 1776, sent Franklin a cipher, which was probably a passage cipher similar to Dumas'.

I attach here a model cipher alphabet for use hereafter between us, if you judge it appropriate. Each principal letter is represented by several different cipher numerals, of which you use one at one time and some other at another time for defying the curious. The words will be distinguished by intervention of a Greek character without significance. Two such characters should have the value of a comma, three a full stop. You will have to eliminate all these Greek characters to read the letter without pain or confusion. Let us try it immediately.
Barbeu-Dubourg to Franklin, 10 June 1776 (the author's translation from French)

The above quotation is followed by an exemplary deciphering as seen in the image below.

Inspection of this example reveals the following correspondence:

1 2 3 4 5 6   10   12   16 17   19   23 24   28....
s a m u e l    d    e    m  i    f    m  l    o....

The first six letters suggest that Barbeu-Dubourg's cipher was also a passage cipher like Dumas'.

When Franklin reached the French shore on 4 December 1776, he addressed his first letter to Barbeu-Dubourg: "My dear friend will be quite surprised to receive a letter from me dated in France ...." (Franklin to Barbeu-Dubourg, 4 December 1776, back-translated from French by the author)

One-Part Code (WE003, THE=840)

Dumas, Franklin, and John Paul Jones, the naval commander, used a one-part code WE003, which was designed in Europe, possibly by Dumas (Weber, p. 79). Judging from the timing, this seems to have replaced Dumas' passage cipher WE041, which Dumas apparently ceased to use after June 1779. Jones may have referred to this when he wrote to Franklin on 25 July, "Before I depart I will send you a Cypher for a private Correspondence".

This is a code list assigning numerical codes of 1(Aberdeen)-928(Your) to words. Basically, the words are arranged alphabetically ("one-part code"), though some blocks are swapped.

Since WE003 did not have codes assigned for single letters or syllables, a substitution alphabet was attached for use in enciphering words or names not in the list.

The news of the victory of Saratoga allowed Franklin to bring France into open alliance with the Americans in 1778. In 1779, Jones was given command of the Bonhomme Richard, a converted merchantman renamed in honor of Franklin after his old publication. Jones set sail on 14 August 1779 from Lorient and won a victory over the British Serapis on 23 September. Jones lost the Bonhomme Richard and transferred his flag to the Alliance and arrived at a neutral port in Holland on 3 October.

A letter written by Dumas probably shortly after this used WE003 (see here).

One-Part Code (WE004, THE=873)

Apparently, a new cipher was introduced between Dumas and Franklin in early 1780. Dumas' letter to Franklin on 8 February 1780 includes a passage partly in cipher, as shown below.

873<The>. 337<France[French]>. 64<ambassador>. 833<still>. 879<think[s]>. 470<it> 419.<impossible>, 30 & 16 would 168<cheat>. But 2. 879<think[s]>. 470<it>. 601<not>. 419<impossible>. till 8. 48<adopt[s]>. 935<what>. 395<Holland>. 757<resolve[s]>.
This is but to try our cypher. Somebody tells me, that the Baron de La H-- of whom I wrote you formerly, is no more to be trusted than 10. 12. & their 373<great>. 339<friend>.
Dumas to Franklin, 8 February 1780

This scheme (WE004) was almost identical with WE003 but codes were offset by 33. Thus, instead of 1(Aberdeen)-928(Your), it ranged from 34(Aberdeen) to 961(Your).

Apparently, the codes 1-33 were assigned to some names and grammatical endings. The following may be extracted from the existing letters:

7<the United States>


In subsequent correspondences between Franklin and Dumas, some key words were occasionally enciphered with this scheme. Names and words not listed in the table were enciphered with a substitution table (e.g., "Brunswick", "Monsieur", "between", "state").

Sometimes, a code was followed by an ending enciphered with the substitution table (e.g., "598<north>PBX<ern>", "657<pention>LBI<ary>" in 2 March 1780).

While word endings might be indicated with specific codes (e.g., "26", "27", "28" in 2 March 1780) or enciphered with the substitution table, they were often just implied. At times, they were simply spelled in plaintext (e.g., 3 August 1780). Further, adjectival forms were often represented by the code for the noun. For example, the word "French" was not in the list and code 337 for "France" was substituted. Similarly, 783<Russia> was used to represent "Russian".

Dumas usually wrote in French but the vocabulary of WE004 was in English. While codes for names as well as codes such as 196<Congress> and 78<arrive> could also be used in a French sentence (e.g., 29 May 1780), some words had to be enciphered letter by letter (e.g., 11 April 1780).

Dumas often switched to English when he used cipher in the overall French letter (e.g., 18 February 1780). At times, Dumas used English for only part of a sentence (e.g., 25 February 1780).

See here for examples of use of this cipher.

Capture of Laurens

The letters between Franklin and Dumas at this period often referred to Henry Laurens, who was to sail from his hometown Charleston on a mission to Holland. However, in 1780, Charleston was attacked by the British and fell in May. As late as 26 July, Franklin thought Laurens sailed as planned. Actually, however, Laurens could not leave Philadelphia till 13 August 1780 . Moreover, his ship was captured by the British on 3 September off New Foundland and Laurens was taken prisoner.

Franklin's letter of 9 October still shows that he did not know of the fact. Dumas told it to Franklin on 10 October. "You will have already learned of the misfortune that befell Mr. Laurens of being taken with his secretary and taken to England. I have written to him, by a sure hand, a letter that offers him the services which he might consider I am capable of doing him." (the author's translation from French)

From this letter on, Dumas stopped using WE004. (According to David R. Chesnutt and C. James Taylor, The Papers of Henry Laurens, pp. 526-528, Henry Laurens' use of cipher is only extant in letters to James Bourdieu. This is different from WE004, assigning "45" to "peace commission".) Apparently, there was no definite agreement for the disuse and Franklin used WE004 at least on two subsequent occasions (3 December 1780 and 18-20 January 1781).

James Lovell's Polyalphabetic Cipher

CHARDON cipher

Back in 1777, James Lovell, who had joined the Committee of Secret Correspondence after Franklin left for France, sent Franklin in Paris a polyalphabetic cipher he gained "by accident". (The text of the undated letter is here. Franklin acknowledged receipt of it on 17 October 1777.)

Lovell described a 27x27 alphabet square as follows. (Lovell included the ampersand (&) in his regular alphabet.)


He gave an example of enciphering with this table.

chardo n char
Powder & Ball are already sent to the amt. ordered,
    doncha                 rdo
and Cannon as directed will be shipped by May. are sent to the amt. ordered,
and as directed will be shipped by 23.25.11.

Some elaboration would be desired to supplement his typically terse explanation: "each letter of the Alphabet will be referable to every figure of the 27".

The text indicates a nonce keyword "chardon". Each of the letters in this keyword is used in turn to encipher respective letters of the plaintext "Powder & Ball" "Cannon", and "May".

First, to encipher the letter "P" with the keyletter "C", one would look at the column headed by the letter "C", i.e., the third column from the left. Then, looking down the column, the plaintext letter "P" is located. The figure on the left side of the table, 14, is used for "P" in the cipher.

Next, to encipher the letter "O", another column of the alphabet is consulted (thus, "polyalphabetic"). Now, in the column headed by "H" (i.e., the second letter of the keyword CHARDON), the plaintext letter is found on row 8. Thus, the "O" is enciphered as "8".

The remaining letters would be similarly enciphered. It is noted there is one error in Lovell's enciphering (or at least in the text the author used). While BALL is enciphered as, the second "L", being enciphered with the keyletter "R", should be "22" instead of "2".

Lovell advised Franklin to use a new keyword if the letter arrived "under dubious circumstances". The full 27x27 alphabetic square allows enciphering with any keyword.

The table for this particular CHARDON cipher would be as follows.

 1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
 C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z  &  A  B
 H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z  &  A  B  C  D  E  F  G
 A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z  &
 R  S  T  U  V  W  X  Y  Z  &  A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q
 D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z  &  A  B  C
 O  P  Q  R  S  T  U  V  W  X  Y  Z  &  A  B  C  D  E  F  G  H  I  J  K  L  M  N
 N  O  P  Q  R  S  T  U  V  W  X  Y  Z  &  A  B  C  D  E  F  G  H  I  J  K  L  M

The correspondence between the cipher and the plaintext is as follows.

14. 8.23.14. 2. 2[22]. are sent to the amt. ordered,
 P  O  W  D  E  R  &  B  A  L  L

and 27.14. 1.12. 8.14 as directed will be shipped by 23.25.11.
     C  A  N  N  O  N                                 M  A  Y

COR cipher (WE043)

Lovell used the polyalphabetic system with his various correspondents and the scheme was often called Lovell's cipher. Lovell took care to use different keywords with different correspondents. Franklin was given keyword COR. Other keywords (WE042 to WE052) include CR for John Adams (see here (in Japanese)), YO for Henry Laurens (as noted above, actual use of this is not known), BY for John Jay (see here), and UNT for William Palfrey.

The table for the COR cipher would be as follows.

 1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
 C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z  &  A  B
 O  P  Q  R  S  T  U  V  W  X  Y  Z  &  A  B  C  D  E  F  G  H  I  J  K  L  M  N
 R  S  T  U  V  W  X  Y  Z  &  A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q

PCC, Roll 72, Page 142 explains the Lovell scheme and permutation of Franklin's keyword "COR" in Dumas' passage cipher (WE041).

Use three columns of regular alphabet placing j & v after their vovvells, "&" being necessary to make up tvventy seven. Several higher answer for baulcs, if especially your other parts for the same purpose are vvrong scor'd. In your columns my figures have a relation perpetually alternate.
If you please to transpose your columns (for which purpose mine are on different keys), you may do it correspondently with my exemplar, only marking your epistle in the margin with one of its figures, original

Franklin Puzzled by Lovell's Cipher

About the same time Dumas started to use WE004 to Franklin, James Lovell urged Franklin to use cipher in his correspondence. Lovell, in his part, was prompted by Luzerne, French minister at Philadelphia.

The Chevalr. De la Luzerne expressed to me an Anxiety because we do not correspond by Cypher. I early communicated to you from Baltimore a very good one, tho' a little tedious like that of Mr. Dumas. I inclose you a Sample at this Time.
Lovell to Franklin, 24 February 1780

From the first, Franklin just ignored Lovell's cipher. But Lovell sent him a letter in his cipher.

Our Affairs at the Southward are to be judged of by the Gazettes. We 11. 14. 8. 12. 1. 3. 27. 13. 11. 17. 6. We have a very good Prospect that the late War between 3. 6. 18. 23. 3. 4. 13. 6. 14. 24. 18. 13. 16. 26. 4. 23. 3. 4 is the last that will spring up between those Tribes....
Lovell to Franklin, 4 May 1780; Franklin Papers, Letters of Delegates

Franklin tried to decipher this on the reverse side of the address sheet.

11 14  8 12  1  3 27 13 11 17  6
 m  a  y  n  o  t  b  &  a  s  t
       j  z  c  e  n  o  m  d  h
 y     v  t  r  q  q  i  y  g  w
 3  6 18 23  3  4 13  6 14 24 18 13 16 26  4 23  3  4
 t  h  e  y  e  t  o  h  p  z  t  o  r  a  f  y     f
             q  r  &  t  a  k  e  &  c  m  r     q  r 
    i  t  j  t  n  e  w  d  n  h  c  f  p  u  m     u 
    w  h  m

The two phrases should read "may not boast" and "the merchant & farmer" but Franklin could not make it out. (In the above, highlighted letters are correct plaintext letters.)

In the first phrase, Lovell made an error when enciphering "o" in "boast". Following "b" enciphered with the C-row (first line) in the COR cipher table above, he should have used the O-row (second line) but he used the C-row again. With hindsight, Franklin correctly deciphered all but one letter but Franklin was at a loss and wrote other candidates below.

For the second phrase, Franklin got right the first three letters 3(t) 6(h) 18(e) but somehow misapplied his table and put "y" for 23. After this, he switched his way of deciphering and just wrote down the candidate letters below each cipher letter.

After all his effort, Franklin could not decipher this cipher.

The Cypher you have communicated, either from some Defect in your Explanation or in my Comprehension, is not yet of use to me; for I cannot understand by it the little Specimen you have wrote in it. If you have that of Mr Dumas, which I left with Mr Morris, we may correspond by it when a few Sentences only are required to be writ in Cypher, but it is too tedious for a whole Letter.
Franklin to Lovell, 10 August 1780

Here, Franklin expressed his preference of the cipher he left with Robert Morris when he left the Committee of Secret Correspondence (see above), that is, Dumas' passage cipher WE041.

Franklin, Dana, and Adams, All at a Loss

While the above letter did not reach Lovell, Franklin communicated his trouble to Francis Dana, to whom he sent an excerpt of Lovell's letter and the instructions to the COR cipher in Lovell's hand. Francis Dana had been sent to Europe in 1779 as a secretary to John Adams and stayed in Paris during December 1780 to April 1781, before his new mission to the court of St Petersburg.

Inclos'd I send you M. Lovell's Cypher which you desir'd. The following is a Copy of a Paragraph of his Letter in which he has made Use of it. If you can find the Key & decypher it, I shall be glad, having myself try'd in vain. "Our Affairs at the Southward are to be judged of by the Gazettes. We We have a very good Prospect that the late War between is the last that will spring up between those Tribes. They have convinced each other by every Skirmish that they ought to be in perpetual Amity on the Ground of reciprocal Benefits."
Franklin to Dana, 2 March 1781

Actually, Dana had been troubled with Lovell's cipher himself. Dana had just received a letter dated 6 January 1781 from Lovell but could not read it because he could not recall until some days afterwards the name used for a key (BRA) Lovell hinted in his letter. He may have talked about Lovell's enigmatic cipher over dinner with Franklin and offered to make a try.

Dana reported Franklin's letter to John Adams and promised to send a copy of the ciphers by a private opportunity. Dana wondered if he could read the ciphers and only said "I will make the attempt."

The copy in Adams Papers has plaintext letters written over cipher numbers but it does not mean deciphering was successful, for every plaintext letters for the second phrase is placed on a wrong number! (Weber p.32, 43; Helen Jones p.460)

 m  e  r  c  h  a  n  t  &  f  a  r  m  e  r  s
36.18.23. 3. 4.13. 4.23. 3.4.

(In addition to Lovell's own mistake, Franklin's excerpts erroneously put "36." for "3.6.") The correct reading would be as follows.

t  h  e  m  e  r  c  h  a  n  t  &  f  a  r  m  e  r
3. 6.18.23. 3. 4.13. 4.23. 3. 4.

Instructions for Peace Negotiations

In June 1781, Congress resolved to appoint Franklin and others as peace commissioners in addition to John Adams. Peace instructions were sent to Adams, enciphered with keyword CR, but Adams could not decipher it. On the other hand, Franklin had no trouble in deciphering his.

I duly received the two letters your excellency did me the honor of writing to me, both dated the 19th of June, together with the letter addressed to the king and the three commissioners, with the instructions relative to the negociations for peace. I immediately went to Versailles and presented the letter, which was graciously received. I communicated also to Count de Vergennes a copy of your instructions, after having decyphered them.
Franklin to the President of Congress, 13 September 1781

This does not mean, however, that Franklin, being a scientist, was more adept in cipher than Adams. While one authority considers the cipher referred to was Lovell's cipher based on the keyword CR (WE042) (Weber, p.111, n.25), there is reason to believe otherwise. Franklin later confessed his preference of the cipher "in which those instructions were written that relate to the future peace", as discussed below.

Two-part Code (WE008, TH=19')

Morris Sends a New Cipher

In July 1781, Robert Morris told Franklin that he took office of Superintendent of Finances and sent Franklin new code tables.

I shall probably have frequent occasion to address You and shall always be happy to hear from you but the Mischiefs which arise from having letters intercepted are great and alarming. I have therefore enclosed you a Cypher and in the duplicate of my letters I shall enclose another, if both arrive you will use one and in case of your absence leave the other with such person as may supply your place, let me know however which Cypher you use, whether it be No 3. or No 4?
Robert Morris to Franklin, 13 July 1781

Actually, Morris failed to enclose the cipher and sent it with another letter the next day.

Franklin acknowledged receipt of the duplicates on 12 September and Major Franks, who bore the originals, arrived thereafter. Franklin, however, did not like the cipher sent by Morris.

I received the Cyphers; but having perus'd them, I imagine the old one preferable, which I left with you. At least it seems so to me, perhaps because I am us'd to it.
Franklin to Robert Morris, 5 November 1781

The "old one" Franklin mentioned is probably Dumas' passage cipher (WE041), which Franklin left with the Committee of Secret Correspondence, of which Morris was a member, when he left America.

Livingston Prefers Morris' Cipher

Robert R. Livingston took oath as the first Secretary for Foreign Affairs for the United States on 20 October and reported it to Franklin the same day. He would take over the task performed by Lovell of the Committee of Foreign Affairs.

The Americans had just gained an important victory at Yorktown over the British in October 1781. General Lafayette obtained leave to visit his family in France in November and took Livingston's letter of 26 November to Franklin.

On 16 December, Livingston informed Franklin that, of the two ciphers No. 3 and No. 4 sent to Franklin, he would use No. 4.

Since my last of which I send a duplicate by this Conveyance nothing material has happened here, unless it be the evacuation of Wilmington.... The Cipher--which I shall use with you is No. 4 of those sent by Mr. Morris--in that the duplicate is written which went un-Cyphered by the Marquis de la Fayette.
Livingston to Franklin, 16 December 1781

WE008 was a code list of about 660 elements. Though the list is smaller than Dumas' WE004, it is more difficult to break because of the random mapping of code numbers (i.e., two-part code), in contrast to the alphabetical ordering of WE004 (i.e., one-part code). To the original list, Livingston added 16 codes in a postscript to a letter to Franklin dated 13 February 1782.

Franklin, however, repeated his preference of Dumas' cipher to Livingston.

I received the Duplicate of your No 4 [i.e., Livingston's 4th letter to Franklin dated 26 November] in Cypher. Hereafter I wish you would use that in which those Instructions were written, that relate to the future Peace. I am accustomed to that, and I think it very good and more convenient in the Practice.
Franklin to Livingston, 4 March 1782

After receiving Livingston's letter of 13 February, Franklin said again:

I have made the Addition you directed to the Cypher: I rather prefer the old one of Dumas, perhaps because I am more us'd to it.
Franklin to Livingston, 25 June 1782

Thus, Franklin repeatedly expressed his preference of "the old one ..., which I [Franklin] left with you [Robert Morris]", "that in which those Instructions were written, that relate to the future Peace", and "the old one of Dumas", that is, Dumas' passage cipher (WE041).

WE008 Shared with Jay

In June 1782, John Jay, who had been working at the court of Spain, joined Franklin in Paris for the last stages of peace negotiations. In a letter of 8 August to Jay, Livingston used "Doctor Franklin's cipher" (WE008). Livingston enclosed the resolution of Congress of 7 August encoded in both WE007 and WE008. (Originally, WE007 had been sent to Jay but Jay reported that it did not arrive safe. See here for details.)

On 9 August, Livingston used WE008 in his letter to Franklin and expressly disapproved of Franklin's preferred cipher.

{The claims of Spain are the dreams} of one who sighs for what he has no title to, and which, if attained, would only add to the misery he has already hoarded. {The degree of estimation in which she stands with us, you will iudge from the resolution tranmitted to} Mr. Jay.
I write to him in your cipher, being No. 4 of the cipher, which Mr Morris sent you. This is also written in the same cipher. I would wish you to use that, as I have no great reliance upon the one you have written in formerly. It has passed through too many hands.
Livingston to Franklin, 9 August 1782; passages in braces { } were written in WE008.

However, Livingston knew that Dumas' cipher was easier to handle. When he knew John Adams was troubled by Lovell's cipher, he had advised to use Dumas'.

I am very fearful that you will not fully understand the ciphers in which my last letters are written. I had them from the late committee of foreign affairs, though they say they never received any letters from you in them. Mr. Lovell has enclosed what he thinks may serve as an explanation. I would recommend it to you to write to me in M. Dumas' cipher till I can send you or you send me one by a safe hand. Should you be at Paris, Dr. Franklin has Dumas' cipher.
Livingston to John Adams, 26 December 1782

Anyway, to Franklin, Livingston continued to use WE008 (see here).

After Livingston

In June 1783, Livingston resigned and Elias Boudinot temporarily took over correspondence with foreign ministers.

As part of the Resolution of Congress of the 12th instant, enclosed in the above letter is of a secret nature, I have wrote it in Cyphers; and not having Mr. Livingston's, I thought it best to use Mr. Morris's to you, which he has obligingly supplied me with, so that the Commissioners must be indebted to you for the deciphering of it.
Elias Boudinot to Franklin, 18 June 1783

Franklin replied three months later.

I received a few days since the private letter your Excellency did me the honour of writing to me of the 18th. June. I regret with you the Resignation of the late Secretary. ....
We found no Difficulty in decyphering the Resolution of Congress. The Commissioners have taken Notice of it in our public letter.
Franklin to Elias Boudinot, 13 September 1783

Probably, Boudinot did not know the code WE008 used by Livingston was Morris'.

Franklin's Unidentified Cipher (THE=167')

About the time Livingston was writing his disapproval of Franklin's preferred cipher, Franklin on his part used a code in his letter of 12 August to Livingston (see PCC, Roll 127, Page 286 (clean transcript), PCC, Roll 108, Page 207). Somehow, this is different from WE008 of Livingston's recommendation and from WE004 between Franklin and Dumas, still less Dumas' passage cipher WE041.

The translation of the coded portion is known to be as follows (shown in braces { }).

I will only mention {that my conjecture
of that court's design to coop us up within the
Allegheny Mountains is now manifested. I hope
Congress will insist on the Mississippi as the boundary,
and the free navigation of the river, from
which they could entirely exclude us.}
Franklin to Livingston, 12 August 1782; passages in braces { } were written in code.

Using repeated letters and syllables as clues, the plaintext may be speculatively mapped to the codes as follows.

281(that) 599(m) 109(y) 124(con) 481(i) 256[ec]
238[tu] 468(r) '(e) 292(of) 281(that) 551(court) 386(s) 263(d) '(e) 268(si) 173(g) 33(n) 451[to] 440<coop> 399(u) .(s) 453[up] 628[wi] 74[th] 11[in]
167(th) '(e) 415[all] '(e) 576[gh] 187<en> 109(y) 16[mo] 542(un) 347[ta] 37(in) .(s) 481(i) .(s) 648[now] 163[man] 30[if] 112[es] 235[t] 193[ed] 481(i) 346[hope]
428[congress] 143[will] 37(in) 268(si) 414[st] 374[on] 167(th) '(e) 83 [mi] 268(si) 268(si) 654[p] 481(i) 254(as) 167(th) '(e) 315[bo] 542(un) 358[da]
468(r) 109(y) 242[,] 159[and] 167(th) '(e) 119[fre] '(e) 402[navy] 460[ga] 447[tion] 292(of) 167(th) '(e) 170[ri] 399(u) 250[er] 242[from] 479[which]
574[they] 200[could] 64<en> 245[ti] 448[r] '(e) 208[l] 109(y) 371[ex] 408[c] 161[lu] 263(d) '[e] 399(u) .(s)

Assignment in parentheses ( ) are considered probable. Five occurrences of 167(th), four of 109(y), four of 481(i), three of 268(si), etc. comfortably fit the plaintext. Codes sandwiched between such highly probable guesses may also be deemed safe.

Assignment in brackets [ ] are tentative. Separation "navy-ga-tion" may seem artificial but there are examples for such separation (e.g. Livingston's letter in WE007 and Adams' letter in WE013).

Assignment in angle brackets < > are dubious. 440<coop> is classified as dubious because "coop" seems hardly worth listing in a code list of less than 700 elements. 187<en> and 64<en> are dubious because of collision.

It is remembered that Morris sent Franklin two codes: No. 3 and No. 4 and Franklin apparently received both. Though Livingston chose to use No. 4, Franklin may have taken up No. 3 for encoding this passage.

Morris' Code No. 3 is considered to have been prepared on the same printed template as No. 4 with different mapping (see here). Use of the same template is supported by (i) presence of code numbers for all the words and syllables of the assignment including "hope", "fre", "navy", "tion" (but excepting 440<coop>), (ii) lack of a code number for "my" (cf. 599(m) 109(y)), and (iii) lack of a separate code number for "V" from "U" (cf. 170[ri] 399(u) 250[er]).

©2009 S.Tomokiyo
First posted on 2 January 2009. Last modified on 24 October 2009.

Articles on Historical Cryptography
inserted by FC2 system