Codes and Ciphers between Washington and Rochambeau and the Yorktown Campaign

After a vigorous campaign in the Carolinas, Cornwallis invaded into Virginia but he soon found himself bottled up in Yorktown. This was a chance for the Americans. Lafayette kept a close watch on Yorktown. At the end of August, Admiral de Grasse brought the French fleet from the West Indies to the Chesapeake to cooperate with the Americans. If Washington could bring his force around New York down to Yorktown, Cornwallis might be checkmated.

Washington Learns of the Prospect of de Grasse's Arrival


The first time Washington learned of the prospect of de Grasse's arrival in North America was in his conference with French general Rochambeau on 22 May in Wethersfield, Connecticut. Although de Grasse was under instructions to bring his West Indian fleet to North America, Rochambeau mentioned this to Washington only as a possibility. The two generals considered a concerted attack on the British main base in New York under Sir Henry Clinton, Commander-in-Chief and operations in the South in case naval superiority would make it eligible.

While Rochambeau did not give a definite promise of de Grasse's arrival, Washington was informed of it by a letter dated 24 March from John Laurens, minister of the United States at the court of Versailles. The crucial passage was in code, which was the same as the one Laurens used in correspondence with the President of Congress (WE005 as classified by Ralph E. Weber).

Thus, on 26 May 1781, Washington wrote in his diary that Laurens' letter informed him "that a Fleet of 20 Sail of the Line was on its departure for the West Indies 12 of which were to proceed to this Coast where it was probable they might arrive in the Month of July."

Cipher, probably the same WE005, was also used on Washington's side at least in his letter of 9 April 1781 to Laurens.

Washington's Penchant for New York Operations


Although Washington and Rochambeau agreed that they would leave to de Grasse the decision between New York and the Chesapeake, Washington's inclination towards New York was clear. Washington's letters, written in the days after the Wethersfield conference and intercepted by the British on 3 June, convinced Clinton that New York was his objective.

Washington's preference of New York was also clear in his letter of 13 June (image of the trascript) to Rochambeau. In this letter, keywords such as Count de Grasse, New York, and Chesapeake were put in cipher.

Similar use of cipher is noted in another letter of 21 July to de Grasse.

The cipher used between Washington and Rochambeau could be identified to be WE001 from the manuscripts of a letter dated 28 May 1781 from Rochambeau to Washington and another letter dated 12 June 1781.

This code, WE001, was a one-part (i.e., alphabetically ordered) code list of 736 elements including words taken from Entick's Spelling Dictionary and some fifty proper names. Unlike nomenclators, single letters or syllables were not listed. Therefore, words not included in the list had to be enciphered with a simple substitution table for the letters of the alphabet.

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 (plaintext)
e f g h i j a b c d o m n p q r k l u v w x y z s t (cipher)

The code WE001 was originally prepared by Benjamin Tallmadge, Washington's director of the secret service, by July 1779 and was used between Washington, Tallmadge, and his agent (Weber p.76-77). Its use includes letters of 6 August 1779 and 15 August 1779 addressed to "John Bolton" (code name for Tallmadge). Apparently, this communication channel ceased to be used around May 1780 (Weber p.108, n.7).

Washington Decides on the Chesapeake


On 14 August, Washington learned that de Grasse's West Indian fleet would come to the Chesapeake and decided to attack Cornwallis in Yorktown rather than Clinton in New York.

Washington sent words of his new strategy and requested preparation of necessary water transport. In a letter of 17 August to Robert Morris, Superintendent of Finances, he asked for one month's pay for the troops and a certain sum for secret service. In a reply dated 22 August, Morris used code WE011 to explain the critical financial situation. WE011 had 660 elements, which was less than older WE001, but WE011 was a two-part code. Further, single letters and syllables were included in the list, thus eliminating the need of a supplemental substitution table.

Morris would use this code, WE011, also in his letters to Washington during the next year on 17 August 1782 and 16 October 1782.

Washington provided "Mr Morris' cipher" to Rochambeau and Greene.

The Count de Rochambeau will establish winter quarters for his troops at this place, Williamsburg and the vicinity; you will be pleased to open a correspondence with him immediately and continue to keep up such intercourse with him as particular circumstances and operations may require, for which purpose and that you may correspond with secrecy, on any occasion, I have furnished him with copy of Mr Morris' cypher, which has been already communicated to you.
Washington to Greene, 31 October 1781

Apparently, Washington sent the same code that Gouverneur Morris had already sent earlier to Rochambeau (Weber p.87) and Greene (Gouverneur Morris to Greene, 11 September 1781, PRM).

Between Washington, Rochambeau, and Greene, WE011 would be used in letters during the next year such as Rochambeau to Greene, 22 January 1782; Rochambeau to Greene, 6 April 1782; Rochambeau to Washington 29 April 1782; (probably) Washington to Rochambeau 5 May 1782; Rochambeau to Washington 13 May 1782; (probably) Washington to Rochambeau, 24 June 1782; Rochambeau to Washington, 4 July 1782; and George Reid to Washington, 8 June 1782.

Code That Clinton Could Not Read

Rochambeau also wrote about the conference with Washington to Luzerne, French minister at Philadelphia. In the letter, Rochambeau told the minister that he told Washington about de Grasse's coming to North America but had to discuss it only as a possibility because he was ordered to keep de Grasse's instruction to himself. This letter was also intercepted by the British on 3 June.

Clinton's cryptographers, however, could not decipher the numerical cipher used in the letter and Clinton had to send copies to London for deciphering. London had then nearly a century long tradition of systematic cryptanalysis. It would not have taken long for the London experts to decipher them. Deciphered documents in French were sent back on 2 August.

Since my letter to you of the 14th ultimo by the Lively, means have been found to decypher the letter from the Comte de Rochambeau to the Chevalier de la Luzerne, transmitted with the other intercepted letters in your last dispatch, and I have the satisfaction to inclose a complete copy of it for your & Lord Cornwallis's private information, and that of the commander of the King's ships, to whom you will likewise communicate it, with every caution of the utmost secrecy.
Germain to Clinton, 2 August 1781

It was only on 2 September that Clinton realized that Washington's objective was the Chesapeake. Washington made a decisive victory at Yorktown by a strategic surprise.

©2008 S.Tomokiyo
First posted on 26 November 2008. Last modified on 22 July 2009.

Articles on Historical Cryptography
inserted by FC2 system