Diplomatic Codes after the Glorious Revolution

Two diplomatic codes are available at Beinecke Digital Collections.

Printed Template

These specimens show that printed templates for making different versions of code were already used around 1690. (Such printed templates are known to have been used, for example, during the American Revolutionary War (see another article).)

The comparison of the two also shows one developed from the other.

Two-Dimensional Arrangement

In addition, the templates are intended to avoid a purely one-part structure. That is, the entries are not simply written column by column but the entire sheet is divided into 24 horizontal sections for A-Z (no distinction of I/J and U/V). Within each horizontal section, the entries are written column by column.

Thus, when figures are filled in column by column without regard to the sections, the resulting assignment of words/names to figures is not completely in alphabetical order.

Such an idea of introducing irregularity by using two dimensions (e.g., figures progressing vertically and alphabetical listing progressing horizontally) was not uncommon in the 17th century. For example, a cipher (1659) made by Edward Hyde for Charles II and John Barwick is similar to the above structure (see another article). A similar idea is seen in French ciphers during the reign of Louis XIV (see another article, esp. items (3)(4)), a cipher of Cardinal Rohan (1721) (see another article), Spanish Cipher Cg.56 (1698) (see another article), and Dutch ciphers (see Karl de Leeuw, Cryptology and Statecraft in the Dutch Republic p.29).

Diplomatic Code (c.1689-1691)

(Beinecke Rare Book and Manuscript Library, Yale University)

Numbers 15-1508 are assigned to the printed entries. A marginal note says "All numbers under 15 and above 1508 & All numbers ending in 4 and 9 are nulls."

The template leaves many blank entries and some words and names are added by hand.


Handwritten entries may help dating the cipher.

The handwritten entry "James, King" might seem to suggest this was made before the Glorious Revolution. But a closer look shows this belongs to the revolutionary regime. (Even in exile, James might be called "King James." Although there is no entry of "King William", the printed entry "the King" would have sufficed. The absence of "Princess Mary" when there is a handwritten entry "Anne, Princesse" also suggests Mary was already Queen. In addition, there are many Dutch people ("Heinsius", "Dyckvelt", "Heinsius", "Odyke", "Overkerke") who worked with William of Orange, while (as far as I know) there is only one Jacobite "Tyrconnel".)

Besides such indirect evidence, the list includes titles created in 1689 after the Glorious Revolution: Earl of Marlborough, Earl of Monmouth, Earl of Portland, Earl of Torrington. The latest possible year is 1694, in which the Earl of Shrewsbury and the Earl of Devonshire were promoted to dukes. Since the Jacobite Tyrconnel died in 1691, this may be dated between 1689 and 1691.

Diplomatic Code (c.1699-1701)

(The James Marshall and Marie-Louise Osborn Collection, Beinecke Rare Book and Manuscript Library, Yale University)

This belongs to the Manchester Papers related to Charles Montagu, Duke of Manchester.

Numbers 3-2342 are assigned to the printed entries. A marginal note says "The numbers under 3 and above 2343 and all numbers ending in 5 and 9 are Blanks. As a[re] likewise the numbers between 516 and 617."

The template leaves many blank entries and some words and names are added by hand.


The catalog record dates this as "1701 July", which appear to be merely derived from the fact that this is preserved with Manchester's letters in this month (see "diplomatic cipher" at the Full HTML catalog).

Internal Evidence about Dating

But there is some conflicting internal evidence. There are printed entries "late King" and "late Queen", which seem to refer to King William (who died in March 1702) and Queen Mary (who died in 1694). If so, this cannot be dated before March 1702. Dating after March 1702 is consistent with the absence of King James (who died in September 1701), though it is wondered why there is no "Pretender." (There is one other candidate for the "late King", that is, Charles II of Spain, who, however, died in November 1701, again after the catalog dating "1701 July".)

On the other hand, there are handwritten entries for the Duke of "Gloucester", who died in July 1700 and Queen Mary (the above-mentioned "late Queen"!?). "Schoning Gen" (Wikipedia) had died in 1696. Some names (Prince "Apassi" and Count of "Tekeli") seem to suggest years before the peace with the Turks in 1699 (Wikipedia).

Manchester's Letters in Code

Given the conflicting internal evidence, one needs to examine letters in which this code was actually used.

The career of the Earl of Manchester (Wikipedia) (created Duke in 1719) in this period was:

(i) envoy to Venice in 1697-1698,

(ii) ambassador to France from 1699 to the outbreak of the War of the Spanish Succession in 1701,

(iii) Secretary of State for the Southern Department in January to May 1702,

(iv) out of office, and then,

(v) ambassador to Venice 1707-1708.

As far as can be found in the Full HTML catalog of the Manchester Papers, most uses of cipher in his correspondence belongs to September 1699 to December 1700 (i.e., during his embassy in Paris). In a note to one undeciphered letter, it is stated that "No key to the ciphers of the period [i.e., around 1699] is known to exist among the State Papers." Before checking the State Papers, the curator must have confirmed the above code is not the one used in this period.

Apart from the above, use of cipher is recorded only in one letter in 23 March 1702, soon after the death of William III on 8 March 1702 OS. It is yet to be seen whether this letter uses the above code.

Reconstructed Version

(Notes added on 12 January 2019) This code appears to be substantially identical with a reconstructed code ("Duke of Manchester's Cypher") printed in 8th Report of the Royal Commission on Historical Manuscripts, Addendum, 141a, though there are some unmatching entries (13=C, 20=C, 21=D, 37=D, 48=E, 72=K) in the reconstruction. The code was used in letters of Robert Yard and Lord Jersey to the Earl of Manchester in 1699-1700. So, the code was made before "1701 July" as in the catague record (though there remains a possibility that the preserved code is an updated version of the reconstructed version).

Further Sources

A letter of the Earl of Nottingham, Secretary of State, Whitehall, 5 July 1692, in which two passages are encoded, is presented at the Folger Shakespeare Library,

He encoded a military objective as "places to be attempted 823 323 195 164 163 319 325 341 [which are St Malo and Brest]." A marginal note indicates he used "cypher with Mr Duncomb."

In May, the Anglo-Dutch fleet defeated the French at the Battle of La Hogue and the French fleet was dispersed and fled to the ports at Cherbourg, La Hougue, St. Malo, and then Brest (see the map at Wikipedia). In England, a counterattack on the French soil was being discussed. Although the troops were embarked in late July, the plan came to nothing (Zee, William and Mary, p.362-363).

©2018 S.Tomokiyo
First posted on 5 August 2018. Last modified on 12 January 2019.
Cryptiana: Articles on Historical Cryptography
inserted by FC2 system