Deciphering Mary Stuart’s lost letters from 1578-1584

Abstract

Mary Stuart, Queen of Scots (1542–1587), has left an extensive corpus of letters held in various archive collections. There is evidence, however that other letters from Mary Stuart are missing from those collections, such as letters referenced in other sources but not found elsewhere. In Under the Molehill – an Elizabethan Spy Story, John Bossy writes that a secret correspondence with her associates and allies, prior to its compromise in mid-1583, was “kept so secure that none of it has survived, and we don’t know what was in it.” We have found over 55 letters fully in cipher in the Bibliothèque nationale de France, which, after we broke the code and deciphered the letters, unexpectedly turned out to be letters from Mary Stuart, addressed mostly to Michel de Castelnau Mauvissière, the French ambassador to England. Written between 1578 and 1584, those newly deciphered letters are most likely part of the aforementioned secret correspondence considered to have been lost, and they constitute a voluminous body of new primary material on Mary Stuart – about 50,000 words in total, shedding new light on some of her years of captivity in England.

Keywords:

• codebreaking
• England
• Francis Walsingham
• French ciphers
• homophonic ciphers
• Mary Queen of Scots
• Mary Stuart
• Michel de Castelnau Mauvissière
• Queen Elizabeth I of England
• Scotland
• 16th century
• 16th century cryptography

1. Introduction

Numerous collections of historical enciphered letters are held in archives. Many ciphertexts can be attributed to specific historical characters and periods. Attribution is possible if the catalog is complete and accurate, or if the ciphertext is accompanied by a plaintext version. This is also possible if the ciphertext is next to related letters, or if the date, the sender, and the recipients are written in clear. However, there are collections of ciphertexts in archives that cannot be attributed unless they are first deciphered. Making them accessible for historical research requires a systematic and large effort to locate, digitize, transcribe, decipher, and analyze them, such as the DECRYPT Project and the Cryptiana website.¹

As part of those efforts, we came across several collections in the Bibliothèque nationale de France (BnF), primarily Français 2988 and Français 20506, containing more than fifty documents fully in cipher.² Using computerized codebreaking techniques, together with manual textual and contextual analysis, we were able to recover the cipher key and decipher all the letters. To our great surprise, those letters turned out to be from Mary Stuart, from 1578 to 1584, addressed mostly to Michel de Castelnau, seigneur de La Mauvissière, the French ambassador in London between 1575 and 1585. We also found in British Archives plaintext copies of seven of those deciphered letters, from 1583 to 1584, which were apparently leaked to Francis Walsingham,³ Queen Elizabeth I’s secretary and spymaster, by a mole in Castelnau’s embassy.⁴ Those plaintext copies allowed us to definitively confirm the origin of the deciphered letters.

In the BnF catalog, the ciphertext documents are merely listed as “Pièce en chiffre” or “dépêches chiffrées”,⁵ while other items in the same collections are described as originating from the 1520s and 1530s and mostly relating to Italian affairs. Furthermore, the ciphertext documents do not contain any parts in clear, and the letters could not have been attributed without first deciphering them. This may explain why such an important source of material had not been previously associated with Mary Stuart before our work.⁶

The present paper is organized as follows. In Section 2, we provide a brief background on Mary Stuart. Section 3 presents sources for the ciphertexts, other letters from and to Mary, and Mary’s ciphers. Section 4 describes the process of breaking the code and deciphering the letters. Section 5 provides an inventory and summaries of the deciphered letters, highlighting some recurrent topics, and reproducing several letters in full. Section 6 describes Mary Stuart’s secret communication channels and methods, an important aspect highlighted by the decipherment of the letters. Section 7 presents concluding remarks and directions for future research. In Appendix A, we describe the codebreaking algorithm. In Appendix B, we examine a theory as to why the letters include numerous systematic enciphering errors that make their decipherment sometimes challenging. In Appendix C, we provide an up-to-date list of all known letters from Mary Stuart to Castelnau.

2. Mary, Queen of Scots

Mary, Queen of Scots, (1542–1587, see Figure 1) acceded to the throne of Scotland when she was six days old after the premature death of her father James V. Sent to France in 1548, she was brought up in the French court. In 1558, she married the Dauphin Francis,⁷ who succeeded to the French throne in 1559 but died the year after. Mary returned to Scotland in 1561. As a dowager Queen of France, she still had properties in France. Charles IX and Henry III,⁸ who succeeded to the French throne one after the other, and Francis, Duke of Alençon (and, later, Duke of Anjou),⁹ Elizabeth’s suitor, were her brothers-in-law. Catherine de’ Medici was her mother-in-law.¹⁰ The members of the powerful family of Guise were Mary’s relatives through her mother Marie de Guise: the Cardinal of Lorraine,¹¹ her uncle, was her mentor, and Henry, Duke of Guise, her cousin, was an ardent supporter of her cause.¹²

Figure 1. Mary, Queen of Scots (1542–1587) – François Clouet, Public domain, via Wikimedia Commons. https://commons.wikimedia.org/wiki/File:Fran%C3%A7ois_Clouet_-_Mary,_Queen_of_Scots_(1542-87)_-_Google_Art_Project.jpg [Accessed November 29, 2022].

Mary married Henry Stuart, Lord Darnley in 1565, but by the time she gave birth to a son, future James VI, in 1566, she had been disillusioned with Darnley. Her third marriage with the Earl of Bothwell,¹³ too soon after the extraordinary death of Darnley, alienated the Scottish people because Bothwell was suspected by many to have been involved in his murder. Protestant nobles turned against Mary and imprisoned her in Lochleven, a castle on an island in the middle of a lake, where she was forced to abdicate in favor of her infant son in 1567.

In 1568, Mary escaped from Lochleven and fled to England, seeking protection from Queen Elizabeth. If Mary thought Elizabeth would help her regain her throne, she was mistaken. Elizabeth and her council considered Mary to be a threat. Descended from Henry VIII’s sister, Mary had a claim to the English throne, and there were many Catholics who believed that Elizabeth was an illegitimate queen, because, in their eyes, Henry VIII’s divorce of Catherine of Aragon and his marriage with Anne Boleyn, Elizabeth’s mother, were void, and Mary was the rightful queen of England. Rather than restoring Mary to the Scottish throne, Elizabeth preferred James to be King of Scotland with pro-English regents. For Elizabeth, restoring the Catholic Mary to the Scottish throne to the detriment of its Protestant regime or allowing her to claim the English throne from France or elsewhere was too risky.

Mary spent the remaining nineteen years of her life captive in England. For most of this period, Mary was in the custody of the Earl of Shrewsbury,¹⁴ and she lived in his properties such as Tutbury Castle, Sheffield Castle, Sheffield Manor, Wingfield Manor, and Chatsworth House.

Mary managed to keep a correspondence with her contacts in England, such as Castelnau, the French ambassador in London, and those in France, including James Beaton, Archbishop of Glasgow,¹⁵ her ambassador in Paris.¹⁶

Even while in captivity, Mary inevitably continued to be the focus of Catholic plots to put her on the English throne with the help of Catholic powers. The Ridolfi Plot in 1571 resulted in the execution of the Duke of Norfolk,¹⁷ whom Mary had intended to marry. This was not to be the last of such plots.

During prolonged negotiations for a marriage between Elizabeth and the Duke of Anjou, Mary openly professed her support of an Anglo-French alliance, while secretly approaching Spain with her own schemes including marrying her son James to a Spanish princess.

Mary’s relationship with her son was also a matter of concern for the English government. Although their relationship was strained because Mary had revoked her forced abdication and claimed to be the sole queen of Scotland, Mary’s hopes rose when Esmé Stewart came over from France in 1579, became the boy king’s favorite, was created Earl of Lennox in 1580 and promoted to Duke in 1581.¹⁸ Wary of their rapprochement, Elizabeth sent her commissioners Robert Beale¹⁹ and later Sir Walter Mildmay²⁰ to Mary under the pretense of negotiating for her release in return for her renouncing any claim to the English crown.

In the meantime, in August 1582, James was kidnapped by a pro-English faction led by William Ruthven, Earl of Gowrie (the Ruthven Raid), and the Duke of Lennox fell from power. The French royal court tried to renew the alliance between France and Scotland by sending to Scotland de la Mothe-Fénelon²¹ and Sieur de Mainneville²² in January 1583. When the king freed himself from the pro-English faction in July 1583, it was Elizabeth’s turn to send Sir Francis Walsingham to Scotland in September 1583 in an attempt to recover the lost grounds.

The discovery in November 1583 of the Throckmorton Plot,²³ one of several Catholic plots to depose Elizabeth and put Mary on the English throne, led to the tightening of the watch on Mary. Moreover, the Earl of Shrewsbury, Mary’s keeper since 1569, who was occasionally accused of being too lenient with Mary, had to be replaced in those times of increasing Catholic threats. The immediate cause was the domestic scandal caused by the Countess of Shrewsbury’s²⁴ false accusations of her husband having an affair with Mary. Mary was put under the custody of Sir Ralph Sadler²⁵ in August 1584, then taken over by a sterner jailor, Sir Amias Paulet²⁶ in April 1585. Under his uncompromising vigilance, Mary’s private communications with the outside world ceased completely.

A greater blow came early in 1585, when Scotland officially rejected proposals for an “association”, i.e., a joint rule, in which Mary and James would share the crown. Mary deplored that she was betrayed by her beloved son, and her hopes for a positive political outturn were shattered by the Treaty of Berwick made on 6 July 1586, in which James allied himself with England with the prospect of succeeding to the throne of England after Elizabeth.

It was at about this time that Mary fatally committed herself to the Babington Plot. Back in January 1586, a local brewer brought in a note from Gilbert Gifford,²⁷ accompanied by a recommendation from Thomas Morgan, Mary’s trusted agent in Paris.²⁸ Thus a secret communication channel was established whereby the brewer would carry packets of letters mediated by Gifford in a beer barrel. But this was a trap carefully orchestrated by Walsingham to produce hard evidence to incriminate Mary. Gifford had been turned into a double agent by Walsingham when he was captured upon his landing from France.²⁹ He provided every letter which passed through his hands to Walsingham, and those in cipher were deciphered by Thomas Phelippes.³⁰

When Mary received a letter from Babington which included the incriminating passage “The dispatch of the vsurping Competitor” (which implies the assassination of Queen Elizabeth), on 17 July 1586 she made the fatal decision to reply despite her secretaries’ better advice. Her answer was drafted by Jacques Nau,³¹ her French secretary. Then, it was translated into English and enciphered by Gilbert Curll, Mary’s Scotch secretary.³²

Mary’s reply to Babington was handed over to Paulet by Gifford and deciphered by Phelippes, who added a postscript in the same cipher. In this forged postscript, Mary allegedly expressed her wish to “know the names and qualities of the sixe gentlemen, which are to accomplish the designment.”³³ The conspirators were arrested and made full confessions. Mary was brought to trial and was found guilty despite pleading her innocence. Interestingly, the forged postscript was not used against her. She was executed in February 1587 in Fotheringhay Castle.

3. Sources

We list here the sources for the encrypted letters we have deciphered. As almost all of them were written by Mary to Michel de Castelnau, we also list here our sources for previously known letters between Mary and Castelnau, and other related letters. Finally, we mention sources for other ciphers employed by Mary.

3.1. Mary’s letters to Castelnau

Castelnau was the French ambassador in London from 1575 to 1585. Castelnau and Mary regularly exchanged letters, some written in cipher. Previously known letters from Mary to Castelnau originate from various archival collections, including:

• The National Archives (TNA)

  •    ○ SP53

• British Library (BL)

  •    ○ Cotton MS, Caligula C

  •    ○ Harley MS 1582

  •    ○ Add MS 48049

• Hatfield House, Cecil Papers

• Bibliothèque nationale de France (BnF)

  •    ○ Cinq Cents de Colbert (500 de Colbert) 470-471

  •    ○ Français 3158 (fr. 3158; formerly Béthune 8675)

  •    ○ Français 3181 (fr. 3181; formerly Béthune 8690)

  •       ○ Français 4736 (fr. 4736; formerly Supplément Français, no. 593(3))

• The archives of the d’Esneval family.³⁴

Most of the letters known to have been written by Mary were published in 1844 by Labanoff, in his seven-volume Lettres, instructions et mémoires de Marie Stuart, reine d’Écosse. Additional letters are reproduced in Teulet (1859) and Basing (1994) or listed in catalogs, such as the Calendars of State Papers (CSP) or Hatfield Calendar, and in an appendix in Bossy (2001). Some letters can be read in English translation in Strickland (1844), supplemented by Turnbull (1845).

Some of Castelnau’s letters to Mary, to Henry III, and to Catherine de’ Medici, that occasionally reference letters from Mary, are reproduced in Teulet (1862), and originate from various sources (ibid. 1-2n1): BnF fr. 15973 (formerly, Harlay 223), fr. 3305 (Béthune 8808), fr. 3307 (Béthune 8810), fr. 3308 (Béthune 8811), fr. 3377 (Béthune 8880), 500 de Colbert 337, 401, Mélanges de Colbert 11, Archives nationales K.95 (formerly Archives de l’Empire).

3.2. Newly deciphered ciphertexts

The ciphertexts we deciphered are from various collections in BnF:

• BnF Français 2988 (fr. 2988; historically 8513 of the collection Béthune):³⁵ Described in the catalog³⁶ as “Recueil de lettres et de pièces originales, et de copies de pièces indiquées comme telles dans le dépouillement qui suit,” followed by a list of 56 items, 26 of which being ciphertexts we deciphered, simply indicated as “Pièce en chiffre” in the catalog: f.21, f.26, f.30, f.34, f.38, f.42, f.46, f.50, f.54, f.58, f.64, f.69, f.74, f.78, f.82, f.87, f.89, f.96, f.98, f.105, f.109, f.113, f.118, f.123, f.125, f.130.

• BnF Français 20506 (fr. 20506; historically Gaignières 394): Described as “Lettres adressées au grand-maître Anne de Montmorency, dépêches chiffrées et pièces diverses, relatives surtout à l’Italie”, with no detailed listing in the catalog. The 28 letters we deciphered are f.151, f.153, f.158, f.163, f.165, f.170, f.174, f.179, f.181, f.185, f.190, f.194, f.198, f.221, f.223, f.225, f.227, f.229, f.231, f.233, f.235, f.237, f.239, f.241, f.243, f.245, f.247, f.249.³⁷

• BnF Cinq Cent de Colbert 470: There are two letters we deciphered, f.307 and f.308. This volume contains other documents – not in cipher – from Castelnau’s office.

• BnF Français 3158 (fr. 3158): f.57 is a letter from Mary to Castelnau, from 30 October 1584. It is written mostly in clear, but it contains short enciphered passages.³⁸

Except for the last item, all the letters are entirely in cipher, with nothing to indicate the date, the writer, or the recipient. Moreover, the newly deciphered letters in fr. 2988 and fr. 20506 are interspersed with letters in Italian, most of them unencrypted, or with names and dates in clear, originating from the first half of the 16th century, giving the wrong impression that the plaintext documents were the deciphered versions of those letters.

By coincidence, the ranges of folio numbers containing the relevant ciphertexts in the first two collections do not overlap. In BnF fr. 2988, the relevant folios are numbered between 21 to 130, and in BnF fr. 20506 the relevant folios are numbered between 151 to 244.³⁹ Similarly, the relevant folio numbers in the last two collections (307, 308, 57) are unique to those collections. This allows us to employ a convenient naming scheme in this paper: Whenever referring to a certain ciphertext document, we only mention its folio number, preceded with a capital F, such as F123 for f.123 of BnF fr. 2988, or F241 for f.241 of BnF fr. 20506.

3.3. Mary’s ciphers

More than a hundred ciphers related to Mary, Queen of Scots, are preserved in The British National Archives (TNA), mainly in SP53/22 and SP53/23.⁴⁰

• SP53/22 includes “Ciphers, including those for papers seized at Chartley Castle on the discovery of the conspiracy in 1586 which were used, when deciphered, to implicate Mary and bring her to trial.” The ciphers in SP53/22 were seized in Mary’s rooms in Chartley Castle when she was arrested following the exposure of the Babington Plot.

• SP53/23 is a collection of ciphers from 1554–1577, broken by John Somer, a codebreaker in the service of Walsingham before the famous Thomas Phelippes.⁴¹ This volume includes more than sixty ciphers.

• SP12/193/54 includes the most famous cipher of Mary, Queen of Scots, the Babington cipher used in her fatal correspondence with Babington, as well as two other ciphers, one of which is used in Mary’s letter to D. Lewis in 1586.⁴² A copy of the Babington cipher is also in Add MS 27027, f.313v.

• SP106/1-3, a large depot of Elizabethan ciphers, includes an incomplete reconstruction of a cipher used by Mary, Albert Fontenay, and her secretary Nau (SP106/1/48A).⁴³

4. Deciphering the letters

This section describes how we deciphered the letters, combining computerized cryptanalysis, manual codebreaking, and linguistic and contextual analysis. The first step was to transcribe the documents which contain only graphical symbols into a format readable by software programs. Due to a large amount of material (more than 150,000 symbols in total), and since automated transcription such as off-the-shelf OCR software was not applicable,⁴⁴ we utilized a special GUI (graphical user interface) tool developed by the CrypTool 2 project.⁴⁵ After transcribing some documents, we performed an initial computer analysis and decipherment, applying the GUI tool codebreaking function described in Appendix A, identifying the original plaintext language, which turned out to be French, and recovering fragments of plaintexts. We then recovered the homophones – the symbols representing single letters of the alphabet, also identifying special symbols (e.g., a symbol to duplicate the last symbol), and the structure of the cipher. After that, we could identify the symbols for common prefixes, suffixes, prepositions, and words. Based on the partial decipherment of several documents, we were able to attribute the letters to Mary, Queen of Scots, addressed to Castelnau, the French ambassador. By reviewing the text of previously-known letters between Mary and Castelnau, we found several documents matching our decipherments, enabling us to determine or validate the meaning of other symbols. Finally, we identified symbols representing names, places, and the twelve months of the year and completed the transcription and decipherment of all the documents.

In the process, we stumbled upon numerous obscure passages, which were either unintelligible or contained a large number of spelling errors. Upon closer examination, we identified recurrent patterns of enciphering errors unlikely to be just the sporadic errors expected when enciphering long letters. We were able to identify the likely cause of those errors, which can be explained by the concurrent use of multiple cipher tables – for different recipients, as we describe in Appendix B. After identifying those error patterns, we were able to complete the decryption of those challenging passages and interpret them.

The recovered cipher table is similar in design to contemporary ciphers, such as other ciphers used by Mary, Queen of Scots, and Castelnau, the French ambassador.

The codebreaking and decipherment process was primarily iterative. We first transcribed a few documents, recovered some parts of the key, then transcribed additional documents, recovered more parts, and so on, rather than fully completing one step before starting the next one. However, for clarity, we describe the process as a sequence of distinct steps. We illustrate the process using one of the deciphered documents, F38, from 20 January 1580.

4.1. Contemporary ciphers

The basic principle of ciphers remained the same for centuries, each letter of the alphabet being substituted with a symbol, i.e., a substitution cipher. The simplest form was the monoalphabetic substitution cipher, with each letter represented by a unique symbol. As early as the 10th century, Arabic scholars developed effective methods for codebreaking based on frequency analysis. Those methods rely on the fact that some letters of the alphabet, e.g., e or t in English, are much more frequent than other letters of the alphabet and the symbols representing them are likely to be the most frequent in ciphertexts and easy to identify. To thwart codebreaking based on frequency analysis, homophonic ciphers were introduced, in which, for the most common letters, e.g., e, there was a choice of several symbols (a.k.a. homophones) to encipher that letter. Because common words or syllables were also likely to reveal recurring patterns useful to codebreakers, a nomenclature, consisting of a list of symbols for frequently used words, names, or syllables, was also included in ciphers.⁴⁶ Null symbols, or nulls, that should be ignored during decryption, were often included in ciphers, as well as special symbols to repeat the preceding symbol or to delete the preceding symbol. The symbols were either arbitrary symbols, including geometrical shapes, Latin or Greek letters, alchemy and astronomy symbols, letter variants, and Arabic figures. Homophonic substitution ciphers, usually accompanied by a nomenclature, remained the standard for centuries.⁴⁷

Before starting the codebreaking and decipherment processes, we were expecting the cipher to be homophonic, with a nomenclature. However, we did not know – a priori – the language of the original unencrypted letter.

4.2. Transcribing the letters

The letters we deciphered consist only of graphical symbols, without any text in clear. An example of an enciphered letter – F38 – is shown in Figure 2.

Figure 2. Sample ciphertext – F38 (Source: gallica.bnf.fr/BnF fr. 2988 f.38).

In all the ciphertexts, we have identified 219 distinct graphical symbol types listed in Figure 3. It can be seen that some symbols are variants of other symbols with the addition of various diacritics, such as the highlighted examples. Based on our knowledge of contemporary ciphers, it was expected that those diacritics would be significant, completely changing the meaning of the symbols.⁴⁸ To obtain an accurate transcription, it was essential to capture the symbols with diacritics separately from the same symbols without them.

Figure 3. Full set of graphical symbols.

In Figure 4, a screenshot of the transcription GUI tool is shown. Using computer mouse gestures, symbols can be marked with a surrounding box, and assigned to specific symbol types (at this stage we did not know yet the meaning of those types, so we assigned them some arbitrary numbers).

Figure 4. GUI Transcription tool (Source for archive image: gallica.bnf.fr/BnF fr. 2988 f.38).

4.3. Initial decryption

To recover and map the homophones, i.e., the symbols representing the letters of the alphabet (a to z), we employed the codebreaking algorithm of the GUI tool, which is described in Appendix A. At first, we had no way of knowing which symbols are homophones and represent single letters and which symbols represent other language entities such as words or names. We, therefore, assumed that any symbol could represent any letter of the alphabet and applied codebreaking under this assumption. We first assumed that the language was Italian, as most of the other documents in the archive collections are in Italian, but obtained no meaningful results. Next, assuming that the language was French, we obtained a tentative decryption, as shown in Figure 5.

Figure 5. F38 – Initial decryption.

It is possible to discern multiple modern French words (“prochaine”, “beau-frere”, “irriter”, “persister”, “moins”, “temps”, “se voir”),⁴⁹ Middle French words (“bruict”, “mectre”)⁵⁰, or plausible parts of words (“catholi…”, “craignan…”, “persecut…”).

Next, we marked those plausible fragments in the GUI tool, confirming the assignments of the homophone symbols that form those fragments, which now appear in capital letters, as shown in Figure 6, in which we also highlight a few interesting sequences. We now focus on the unconfirmed symbols (in lower case in the decryption) which were likely to have been wrongly assigned.

Figure 6. F38 – Decryption with some symbols confirmed (in capital letters).

From the sequence SVRLARIVEPROCHAINE, and from other sequences, we hypothesized that the symbol may have the effect of duplicating the last letter instead of representing one of the letters of the alphabet, because under this assumption, SVRLARIVEPROCHAINE reads as SURLARRIVEEPROCHAINE (“sur l’arrivée prochaine” – about the upcoming arrival), which is plausible.

The expression “sur l’arrivée prochaine” is expected to be followed by the word “de” (of, in English), rather than a word starting with the letter R as in the tentative decryption in Figure 6. We hypothesized that the symbol that follows R may indicate that the previous symbol should be erased. This hypothesis was confirmed when analyzing other sequences. For example, under this assumption, the sequence MAVISIRERE reads as MAVVISSIERE, Mauvissiere being a plausible French name (and in fact, the French ambassador to England and the recipient of the letter).⁵¹

It can be seen in Figure 6 that symbols with diacritics, when interpreted by the tool as homophones representing single letters of the alphabet, are usually not part of contiguous plausible segments. Furthermore, based on our knowledge of contemporary ciphers, we also expected that symbols with diacritics would rather represent words or entities other than single letters. For example, the sequence CATHOLIS is likely to be CATHOLIQUES (Catholics), the symbol representing the word QUE (that) rather than one of the letters of the alphabet.

4.4. Recovering the homophone symbols

To recover the mapping of the homophones more accurately, we excluded all the symbols with diacritics, so that only those without diacritics are considered to be homophone candidates by the tool’s codebreaking algorithm. We thus obtained an improved decryption, as shown in Figure 7. There are significantly fewer sequences that do not make sense compared to the initial decryption in Section 4.3, and most of the recovered segments contain either complete words or plausible parts of words.

Figure 7. F38 – Decryption after identifying the homophones.

The recovered mapping of the homophones is shown in Figure 8. It can be seen that there are two homophones for most letters of the alphabet, and only one for the remaining letters.⁵²

Figure 8. Homophones.

4.5. Recovering the symbols for words and parts of words

We illustrate here the generic process to recover the meaning of the symbols which are part of the nomenclature, i.e., they represent words or parts of words, e.g., prefixes and suffixes. As mentioned in Section 4.4, we expected those symbols to include diacritics. The process was performed manually, with the aid of the GUI tool, based on linguistic analysis of the deciphered fragments.⁵³

In Figure 9, we highlight the occurrences of the symbol , of which we don’t yet know the meaning.

Figure 9. F38 – Identifying common words and parts of words - 1.

The first occurrence follows L’ARRIVEE PROCHAINE (the upcoming arrival) and we expect this expression to be followed by “de” (of). The third instance (on the fifth line) precedes CA, so if is “de”, this could be “deça” (over here), which is also plausible. Other instances of , including in other ciphertext documents, also look plausible when interpreted as “de” and we can assume that means “de”.

We manually enter the assignment of as “DE” in the GUI tool, and obtain the decryption shown in Figure 10, in which we also highlight the symbols , , and , which we want to interpret next.

Figure 10. F38 – Identifying common words and parts of words - 2.

• appears in ADVAGE, which is likely to be “advantage”, and therefore, is likely to represent the common French suffix ANT. also appears in AVT, which results in AVTANT (“autant”, so much), and in VOY – resulting in VOYANT (“voyant”, seeing), both also being plausible.

• appears in RABLISSEM, likely to be RESTABLISSEMENT (the Middle French form of “rétablissement”, reestablishment). So would be EST (“est”, is), and the common French suffix ENT. also appears in INT, which could be INTENT, and as it is followed by , another symbol with diacritics, INT may represent INTENTION, representing ION, which is also compatible with PERSECUT, “persecution”, in the last line.

• appears in MBEAVFRERE, which could be “mon beau-frere” (my brother-in-law), and thus, likely represents ON, which is also a French word (loosely translated to “one” or “someone”). This is also consistent with SEMEDIVERSBRUICTS, or “on seme divers bruicts” (someone is spreading various rumors) in the second line.

Using the GUI tool, it is possible to test new hypotheses and discard those that are not plausible. New hypotheses, if found to be correct, will in turn lead to the successful interpretation of new symbols, creating an “avalanche effect” and allowing for the recovery of the meaning of the vast majority of the symbols. The process is somewhat analogous to solving a large crossword puzzle, except that the tool’s codebreaking algorithm provides a headstart, by automatically identifying and mapping the homophones – the symbols that represent the letters of the alphabet.

4.6. Recovering the symbols for names and places

We identified most of the symbols representing people and places using contextual analysis, initially recognizing the characters, events, and places referred to in the ciphertext based on additional sources and the historical context. We could do that only after we had made significant progress with deciphering the documents and could read most of the deciphered text.

We illustrate the process with the symbol , which we expected to represent a proper name. This symbol appears in “sur l’arrivée prochaine de mon beau-frere” (about the upcoming arrival of my brother-in-law). After we were able to establish that the author of the letters was Mary, Queen of Scots, we considered the historical figures who could have been her brother-in-law mentioned in the letters. There were three candidates, as Mary had been married to the late French King Francis II, who had three brothers when he died in 1560. One brother, Charles IX, was King of France from 1560 until his death in 1574. He was succeeded by Henry III, another brother who reigned from 1574 to 1589. The third brother, Francis, Duke of Anjou, was engaged in negotiations to marry Elizabeth I, Queen of England, and made several visits to England to meet her. Since Mary was in captivity in England when she wrote about the upcoming arrival of , her brother-in-law, it is clear from the historical context that the symbol must represent Francis, Duke of Anjou.

Other names and places were less straightforward to identify, some of them being mentioned only once or very few times in the collections of ciphertexts. The discovery of plaintext copies of several deciphered letters in archives was a great aid in the process, as they mentioned names which could be matched with certain symbols.⁵⁴ Eventually, we were able to recover the meaning of almost all the 219 distinct symbol types, apart from a handful of less-frequent ones. In Figure 11, we show our final decipherment of F38. We highlight an enciphering error (an error made by the secretary who enciphered the letter in 1580), OFFPSEZ instead of the expected OFFENSEZ (offended), likely due to the fact that the symbol for the letter p, , is similar to the symbol for EN, , but without the dot.

Figure 11. F38 – Final decryption.

The complete plaintext reformatted with spaces and punctuation may be found in Section 5.3 (F38), together with a tentative translation.

4.7. Recovering the symbols for the months

The deciphered letters usually included, near the end, the day of the month, spelled out in full, and a symbol representing the month. We recovered the symbols representing the months of the year by examining letters for which a copy of the plaintext was available, and by analyzing the contents of deciphered letters taking into consideration the historical context, with the following series of deductions:

• By comparing our decipherments to known plaintext copies of those letters, we identified the symbols for February (F58), March (F96), April (F96), July (F46), August (F69), and September (F69).

• F87 dates from “ce deuxiesme , jour de ma delivrance de Loklin”, i.e., the anniversary of Mary’s escape from Lochleven Castle, which was on 2 May 1568, hence the symbol stands for May.

• F98 is from the last day of July, and acknowledges the receipt of Castelnau’s last letters from the 26th of , which must be 26th of June (May and April already being assigned to other symbols).

• In F50, from the last day of , Mary reports that the Earl of Shrewsbury has read to her a letter from Walsingham with answers from Queen Elizabeth to du Ruisseau’s⁵⁵ requests on Mary’s behalf. Another letter (a draft in Nau’s hand) from 2 November 1582 dates this specific event to 31 October,⁵⁶ from which it follows that stands for October.

• In F21, from 9 January (“janvier” being spelled out), Mary acknowledges the receipt of Castelnau’s letters from the 9th and 13th of , and in F123, from the 12th of , she confirms the receipt of Henry Howard’s⁵⁷ letters from the 24th of , from which it follows with high confidence, considering the months already assigned, that is December and is November.

• The only remaining month, January, fits the last unassigned month symbol, , as there is thematic overlap between the deciphered F38, from the 20th of , and a letter to Beaton from 20 January 1580,⁵⁸ so both letters are likely to have been written on the same day.

Figure 12 shows the recovered month symbols (applying French spelling that is consistent with contemporary letters written by Nau).

Figure 12. Mary-Castelnau cipher – Month symbols.

4.8. The complete reconstructed Mary-Castelnau cipher

We show in Figures 13 and 14 our reconstruction of the cipher used by Mary Stuart to communicate with Castelnau.

Figure 13. Mary-Castelnau cipher – 1.

Figure 14. Mary-Castelnau cipher – 2.

In Middle French writings, the letters u and v were not distinguished as they are today (v for the consonantal [v], u for the vocalic [u] and the semi-vocalic [ɥ]). They could be used interchangeably, while they often appear as positional variants, with v used initially (e.g., “vser”) and u medially (e.g., “auoir”), without reference to their phonetic value.⁵⁹ Similarly, the letters i and j could both be used for both vocalic [i] and consonantal [ʒ] (e.g., “ie”).⁶⁰ Furthermore, some letters found in modern French, such as k and w, were not part of the basic alphabet although they might have been needed to spell some foreign names. Those particularities were also reflected in cipher tables, and in the reconstructed Mary-Castelnau cipher, as described below.⁶¹

• The homophones and interchangeably represent the letter u as well as the letter v.

• Similarly, the homophones and may both be used for the letter i as well as for the letter j.

• The letter y has a distinct homophone, .⁶²

• The letter k has distinct homophones, and , needed to spell English names such as Norfolk.

• Despite the need to spell names like Walsingham, there is no homophone for the letter w. For example, Walsingham is most often spelled as “vvalsingham”, and in some cases, “valsingham”.

• Accents are disregarded while enciphering, as there are no dedicated symbols to represent letters with accents. Similarly, spaces and punctuation marks are generally not enciphered.

4.9. Comparing the Mary-Castelnau cipher with related ciphers

As mentioned in Section 3.3, the State Papers at the British National Archives contain dozens of cipher keys used by Mary and her associates that were captured or reconstructed from intercepted messages by Walsingham’s and Burghley’s agents.⁶³ The famous cipher used in the Babington plot is of the simplest kind, a monoalphabetic substitution cipher with a small nomenclature. However, most of the other ciphers in those archives are homophonic and include a nomenclature with names, places, and either French or English common words. Most ciphers employ graphical symbols, sometimes with diacritics.

The reconstructed Mary-Castelnau cipher does not appear in the archive collections listed in Section 3.3, but other ciphers share several features with it. One of them is a homophonic cipher between Mary and Châteauneuf, French ambassador in London after Castelnau.⁶⁴ Some parts of the cipher table are shown in Figure 15. It includes nulls and symbols for repeating or deleting the last symbol. It also includes symbols for punctuation, months, and numbers. Its nomenclature is extensive, with many entries also appearing in the Mary-Castelnau cipher. It also employs diacritics, some similar to those of the Mary-Castelnau cipher.

Figure 15. Cipher between Mary and Châteauneuf – selected parts (Source: TNA SP52/22/22).

The reconstructed Mary-Castelnau cipher also resembles a cipher, shown in Figure 16, held in a collection of Castelnau’s papers from his time as ambassador in London.⁶⁵ The two ciphers have similar diacritics and similar nomenclature vocabularies.

Figure 16. A cipher found in Castelnau’s papers (Source: gallica.bnf.fr/BnF 500 de Colbert 472, p. 347).

Interestingly, diacritics were not very common in French diplomatic ciphers of the time. For example, contemporary ciphers used by the French ambassadors in Rome, Venice, Spain, and Constantinople did not feature diacritics.⁶⁶ It is unclear whether the Mary-Castelnau cipher was designed by Mary and her associates or was based on other French cipher designs, and assessing its origin requires further research.

5. The deciphered letters

In this section, we examine the inventory of the newly deciphered letters from Mary to Castelnau, we briefly highlight some of the topics those letters deal with frequently, and we provide a summary of the contents of each letter, with some letters reproduced in full.

5.1. Inventory and timeline

Among the 57 letters we deciphered as part of this work, 54 are from Mary to Castelnau. Two additional letters (F307, F198) are from Mary to de la Mothe-Fénelon, Henry III’s envoy to Scotland toward the end of 1582. A third one is from Jacques Nau, Mary’s secretary, to Jean Arnault (F249).⁶⁷ All letters except one (F308) are dated, that is, the month and the day of the month are mentioned – in cipher, and we were able to attribute the year based on contents and historical context. The newly deciphered letters were written while Mary was under the custody of the Earl of Shrewsbury, from Sheffield or from another of his properties. Among the letters from Mary to Castelnau we deciphered, the earliest is from May 1578 (F87), the latest fully in cipher is from May 1584 (F221), and the latest – only partially enciphered – is from 30 October 1584 (F57). For comparison, the earliest known letter in clear from Mary to Castelnau is from March 1576,⁶⁸ and the latest is from March 1586,⁶⁹ after Castelnau had already returned to France.

We needed to determine whether there were other letters from Mary to Castelnau, apart from those we deciphered, that were originally sent in cipher, although only a plaintext copy may be found in archives, e.g., a leaked copy of the original plaintext. This process can be challenging,⁷⁰ but we were able to establish that three of the previously known letters were likely to have been sent in cipher,⁷¹ while their ciphertext version is not in the BnF letters we deciphered. In addition, we have been able to locate the plaintext of seven newly-deciphered letters in British archives (F42, F46, F58, F69, F78, F96, F221), which were probably leaked from Castelnau’s embassy. Apart from those ten letters in cipher (seven plus three mentioned above), it is reasonable to assume that the remaining previously known letters were sent in clear, though this is not certain except for those which are known to be the original, e.g., those that have an autograph signature. To the best of our knowledge, the remaining 45 newly deciphered letters from Mary to Castelnau do not appear in archival sources surveyed as part of this work (listed in Section 3).⁷²

Often, letters written in cipher from Mary to Castelnau would refer to previous letters sent to him in cipher. In addition, at least eight letters from Castelnau to the French court refer to letters in cipher that Castelnau had received recently from Mary.⁷³ We have identified three enciphered letters from Mary to Castelnau that match those referenced letters. The remaining five referenced letters could not be identified, indicating that they might have been lost.

In Appendix C, we list the newly deciphered letters together with previously-known letters found in archives, and together they constitute an up-to-date combined corpus of letters from Mary to Castelnau, with almost one hundred letters in total. To contextualize our new decipherments, we show in Figure 17 a breakdown of the letters in the combined corpus over 6-month periods from 1576 to 1586, dividing the letters into three categories:

Figure 17. Mary’s letters to Castelnau – 1576–1586.

• Newly deciphered letters and previously unknown.

• Newly deciphered letters for which a plaintext copy was previously known.

• Known letters.

We can roughly divide the combined corpus of Mary-to-Castelnau letters into five periods:

• 1576-1579: It is not fully clear why there are so few letters in this four-year period, with only four newly deciphered letters and four previously-known letters in clear. A secure channel to convey enciphered letters could have faced some challenges, as evidenced by Mary complaining in May 1578 that she had not received any letters in the previous eight months (F87). Also, two enciphered letters are referenced in other letters but we could not find them in the combined corpus,⁷⁴ hinting at the possibility that additional letters from this period were exchanged but may not have survived.

• 1580-1581: There are 11 newly deciphered letters from this period in addition to four previously-known letters written in clear. Only six of nine enciphered letters referenced in other letters are included in the combined corpus, so some letters exchanged during that period may also have been lost.⁷⁵

• 1582-mid-1583: At this stage, the confidential communication channel between Mary and Castelnau was probably secure and stable, involving multiple trusted couriers, and the volume of letters significantly increases, with 29 newly deciphered letters. In addition, there are six letters written in clear and most probably delivered under Walsingham’s supervision. Nine enciphered letters are referenced in other letters, and as all except one can be matched with letters in the corpus,⁷⁶ we estimate that the majority of the enciphered letters from Mary to Castelnau from that period are included in this corpus.

• Mid-1583-1584: In mid-1583, Walsingham was able to recruit a mole in the French embassy and he obtained leaked copies of letters from and to Castelnau.⁷⁷ The plaintext of most of our newly deciphered letters from that period, seven of eight, can be found in British archives, in addition to the plaintext copies of three letters originally written in cipher for which we did not find a ciphertext in the BnF archives. It would seem that the leak from the embassy was quite effective and comprehensive.⁷⁸

• 1585-1586: Following the discovery of the Throckmorton plot, Castelnau had apparently agreed to discontinue the exchange of enciphered letters with Mary. All the letters from that period were previously known, and all of them are fully in clear. Some of them were sent by Mary to Castelnau after he was replaced in September 1585 as French ambassador to England.

In summary, our new decipherments significantly enlarge the corpus of letters from Mary to Castelnau available for historical research, accounting for 45 of a total of 99 letters. For the period from 1580 to mid-1583, our new decipherments account for about 80% of the corpus. They also account for the vast majority of the letters sent in cipher (45 of 55) and for all the letters in cipher before July 1583.

5.2. Main topics covered in the letters

Mary’s secret letters highlight a multitude of topics, of which only a few are mentioned in this section.

A major recurrent topic has to do with Mary’s efforts to maintain a secure communication channel with Castelnau, and through him, with her network of associates and allies, mainly in France. Careful precautions were taken to conceal and protect this critical channel, the deciphered letters showing that it was in place as early as 1578 and active until at least mid-1584. This confidential channel operated in parallel with an official channel under Walsingham’s supervision through which, in Mary’s own words, she would never write anything that she did not want even her worst enemies to be able to read.⁷⁹

Another recurrent topic in the letters is the proposed marriage between the Duke of Anjou and Queen Elizabeth. While Mary pledges her support for the marriage, often vehemently defending herself against accusations she is in fact opposing it, she constantly warns Castelnau that the English side is not sincere with their negotiations, their only purpose being to weaken France and counter Spain, by encouraging the duke to attack Spain in the Low Countries.⁸⁰ After the duke’s campaign in Flanders ends up in disaster, as she had been warning all along,⁸¹ Mary offers to help in reconciling the duke with the king of Spain.⁸²

In several letters, Mary expresses a strong animosity toward the Earl of Leicester,⁸³ a longtime favorite of the queen, and toward other members of the Puritan faction, whom she accuses of fomenting plots against her, her son James, and even against Queen Elizabeth, such as a scheme to marry Leicester’s son to Arbella Stuart,⁸⁴ granddaughter of Mary’s hostess the Countess of Shrewsbury, with the intent of claiming the succession to the English throne. Mary is often writing to Castelnau on those alleged plots, advising him to report them to the queen, but without mentioning that the information came from her.⁸⁵ Francis Walsingham is also frequently mentioned in the letters, Mary warning Castelnau of his schemes in France and Scotland,⁸⁶ describing him in negative terms, as a cunning person, falsely offering his friendship while concealing his true intentions.⁸⁷

A series of letters from the second half of 1582 highlights Mary’s frantic response to the news on the abduction of her son James by a Scottish faction (the Ruthven Raid),⁸⁸ desperately asking for help from France. When the French king finally sends an envoy to Scotland, Mary expresses her dissatisfaction at the results and her feeling that she and her son have been abandoned by France.⁸⁹

Several letters refer to negotiations about Mary’s release and her reestablishment to the Scottish throne in association with her son, in return for her giving up all claims of succession to the English throne. The letters report on several visits by Robert Beale and other commissioners on behalf of Queen Elizabeth, Mary initially hoping that an agreement may be achieved, but bitterly reaching the conclusion that the commissioners either didn’t have the required mandate, or that those negotiations were not in good faith, and no more than an attempt to gain time or extract intelligence from her.⁹⁰

Mary’s complaints about her conditions in captivity, and her requests to improve them are frequently mentioned in the letters. The letters also deal with matters related to her dowry in France, and her efforts to ensure that her servants and allies are financially rewarded. From time to time, she suggests enticing various people with financial rewards so that they would switch sides, or soften their attitude toward her.⁹¹ She also asks for Castelnau’s assistance in recruiting new spies and couriers, while sometimes she warns him – rightly – that some people working for her might be Walsingham’s agents.⁹²

By nature, the contents of letters exchanged in cipher via a confidential channel are expected to be more revealing than the contents of official letters. However, while some of the letters were exchanged at the time of the Throckmorton Plot in 1583, even mentioning Francis Throckmorton as a trusted courier,⁹³ they do not contain any details about the plot, which is only indirectly alluded to after it had been exposed, Mary deploring Throckmorton’s suffering after his arrest.⁹⁴

5.3. Summaries of the letters

In this section, we provide preliminary summaries of the contents of each newly deciphered letter, and in some cases, we provide a full decryption together with a tentative translation.⁹⁵ The plaintext of seven of the deciphered letters appears in known sources.⁹⁶ For those letters, only the parts that differ from or are missing from the known copy are summarized.

We also tentatively dated the letters. Since the month and the day of the month are written in the letters (in cipher), we only needed to identify the correct year of each letter, based on contents and historical context.⁹⁷ For each document, we provide evidence for the assigned year in a footnote. All years are considered to start on January 1st. Also, although the Gregorian calendar was introduced in October 1582 and adopted in France in December, with a 10-day shift, there is evidence that Mary continued to use the old calendar style long after October 1582,⁹⁸ and for convenience, all the dates given below follow the old style.

Most of the letters were written in Sheffield, as indicated by a special symbol in the enciphered letter. The other locations, Chatsworth, Wingfield, and Worksop, are spelled out in cipher.

Mary’s letters to Castelnau often enclosed letters to be forwarded to her contacts, mostly in France. These enclosures and their intended recipients were usually mentioned – in cipher – at the end of the main letter, together with a special symbol indicative of each recipient, matching the same symbol used to mark the relevant enclosure so that its recipient could be identified.⁹⁹ Since we consider the knowledge of these enclosures to be of importance for research, for each letter we list the recipients of enclosures referenced in the ciphertext. In the archive collections in which we found the enciphered letters from Mary to Castelnau, we could not identify any enciphered copies of letters that would qualify as such enclosures, which may suggest that Castelnau would have forwarded them as requested without keeping copies.

We list the letters and their summaries in chronological order.

Unless mentioned otherwise, “the king” refers to King Henri III of France, and “the queen” to Queen Elizabeth I of England. “The Queen Mother” refers to Catherine de’ Medici.

F87 − 2 May 1578, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 87.¹⁰⁰

Contains an enclosure for James Beaton.¹⁰¹

Mary complains she has not received any letter for more than seven months. Robertson,¹⁰² a trusted courier, is afraid of being exposed because of news of an alleged interception. Mary proposes to reassure Walsingham of her good intentions, and if her succession rights to the English throne are preserved, as well as those of her son James,¹⁰³ she agrees to staying in captivity. She acknowledges that Walsingham is a clever man and may detect any attempt to write letters with contents that are too positive, so she will also include a rebuttal of his accusations against her in a letter she is about to send via the official channel. She claims that Walsingham and the Puritans are acting against the queen, supporting the succession claim of the Earl of Huntingdon,¹⁰⁴ therefore seeing Mary and her son as their enemies. Mary wishes to bring England back to Catholicism, but not by force. She gives her consent for her official correspondence to be read by Walsingham. Mary fears that she might be moved to someone else’s custody, instead of the Earl of Shrewsbury whom she finds decent, and in such a case, her life would be in great danger in the event that Elizabeth dies. Mary thanks Castelnau for his advice on her affairs in Scotland, asking that the king of France support her allies, the Earls of Argyll¹⁰⁵ and Atholl,¹⁰⁶ who recently recovered some of their authority. She warns Castelnau that Moulins¹⁰⁷ and Cockburn¹⁰⁸ are spies in the service of Morton.¹⁰⁹ She is considering sending a token to the Earl of Leicester, following Castelnau’s advice.

F227 − 12 June 1578, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 227.¹¹⁰

Contains enclosures for Beaton and Francisco Berthy.¹¹¹

Mary thanks Castelnau for interceding with the queen on behalf of the French envoy Hieronimo Gondy,¹¹² and for mitigating the queen’s bursts of anger against Mary, despite Mary’s respect for her. She also thanks Castelnau for his good services with the Earl of Leicester, Lord Burghley, Walsingham, and Wilson.¹¹³ The Earl of Leicester was at Sheffield this week, but her captivity conditions have worsened instead of being improved as she had been told. Mary fears again that she will be transferred, under the false pretext of reports on the activities in Ireland of Catholics who had been banned, while at the same time, the Earl of Leicester is about to start persecuting the many Catholics in England. Mary wishes to win over Wilson’s wife with presents and money, but this must be done in secret. She is not happy with the plans of the Duke of Anjou in the Low Countries. The queen and her council are planning to support the party opposing those favored by France in Scotland, to prevent France from gaining more influence. The Earls of Argyll and Atholl have been reconciled with Morton, raising Mary’s fears for the safety of her son who is now in the hands of the unreliable Earl of Mar.¹¹⁴

F249 − 03 October 1579, Jacques Nau to Jean Arnault, Sheffield

Reference: BnF Fr. 20506 f. 249.¹¹⁵

As Nau is sending this letter to Jean Arnault with a new courier, he only writes things of little importance, which is what Arnault is also supposed to do in his reply. Nau reports that letters from Arnault and de Chaulnes¹¹⁶ sent to Sheffield were opened before they came into the hands of Wilson. Adam Blackwood¹¹⁷ from Mary’s council is trying to get his father-in-law to succeed Dolu¹¹⁸ as Mary’s treasurer in France, having also written directly to Walsingham. Mary, however, favors de Chaulnes, Arnault’s brother-in-law, and Nau regrets that some who had previously recommended de Chaulnes now support Blackwood’s claim.

F241 − 8 November 1579(?), Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 241.¹¹⁹

Mary has now two or three new ways to convey secret letters. She asks Castelnau to update her on the progress of the negotiations on the marriage between the queen and the Duke of Anjou. The anti-marriage faction turns against Mary because she has written to the queen in support for the marriage, also claiming that Mary has thus harmed the public interest. She asks Castelnau to convince the queen that she has no one in her realm to better rely upon than Mary, who in turn has no interest in losing the queen’s goodwill. Mary can’t imagine that the Duke of Anjou has any intentions toward her which are not positive. She thanks Castelnau for helping John Hamilton.¹²⁰

F233 − 26 November 1579, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 233.¹²¹

Contains enclosures for Beaton and Robertson.

Castelnau has informed Mary of a calumny against her, being apparently accused of having spoken against the Duke of Anjou.¹²² She claims that this is coming from the faction in England which is against the Elizabeth-Anjou marriage, this faction trying to weaken Mary’s friendship with the Duke of Anjou, hoping to attract her to their side, which they could not achieve by their previous mischievous actions, on which Mary promises to provide more details later. Mary does not believe the calumny was invented by the queen. She asks Castelnau to plead that the queen investigate the origin of this slander.

F179 − 6 January 1580, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 179.¹²³

Contains an enclosure for Beaton.

Although Fowler¹²⁴ had started the slander by reporting accusations against Mary,¹²⁵ she believes that it was concocted by the faction opposing the marriage. She requires again that Castelnau ask the queen to investigate the matter and Fowler in particular, who knows about intrigues by the Puritans against the marriage and against Queen Elizabeth. Wilson has written to the Earl of Shrewsbury, denying the accusations against Mary. Mary has learned that the Earl of Leicester has been selling his properties, and some think he may retire if the marriage takes place. The Puritans, currently too weak to act, are allying with the Scottish rebels, openly objecting to the rule of women. It is feared they may capture both the queen and Mary.¹²⁶ Mary wants Castelnau to ask the queen to write a letter to her host, the Earl of Shrewsbury, recounting her trust in him, to prevent him from joining the Puritan faction. Although the Earl of Leicester has denied to the queen that he had married the Countess of Essex,¹²⁷ the latter has been signing her secret letters as “L. Leycester” for more than a year, and Castelnau should use this information against him. The Earl of Leicester told Mary that she owes him her life as the queen once wanted her dead. Mary has arranged for costly silverware to be given to Wilson. She complains about an apothecary, who was sent to her by Adam Blackwood, having worsened her condition.

F231 − 7 January 1580, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 231.¹²⁸

Contains an enclosure for Beaton.

Mary asks Castelnau to find out who has convinced the queen to have her letters to him opened in the Privy Council, although she is not concerned about their contents, as in such letters sent via the official channel, she would never write anything she doesn’t want even her worst enemies to be able to read. Castelnau should reassure the queen of Mary’s affection and sincerity. Mary is happy about the possibility mentioned by Castelnau that the queen may convert (to Catholicism) if the Elizabeth-Anjou marriage takes place. She warns Castelnau that Walsingham falsely presents himself as a friend, hiding his real intentions. She fears that her host and hostess may join the Earl of Leicester and the Earl of Huntingdon against her, and she asks Castelnau to find out more from Wilson.

F181 − 15 January 1580, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 181.¹²⁹

Contains an enclosure for Robertson.

This letter is being delivered by a new trusted courier, and Mary is asking Castelnau to report to her via him on the latest developments around the Elizabeth-Anjou marriage, and the upcoming session of the parliament, fearing that her enemies may drive decisions against her and the Catholics. She would like Castelnau to recommend her cousin John Hamilton, who has been opposing Morton, to the king of France. The Earl of Leicester has fallen from grace, for which Mary feels sorry. The anti-marriage faction is causing disturbances in France to divert anger directed at them. Representatives from England were sent to the (Protestant) synod of Montauban and to another location,¹³⁰ Walsingham driving those efforts, with the ultimate purpose of converting all Frenchmen to be Huguenots. The faction opposing the marriage also spreads rumors that James is very sick. Mary informs Castelnau that Fowler has admitted that his (false) testimony was given at the request of the Earl of Leicester. She pledges that any further negotiations she may undertake would only be conducted via Castelnau.

F38 − 20 January 1580, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 38.¹³¹

We reproduce here the full decryption and a tentative translation.

Monsieur de Mauvissiere, d’autant que sur l’arrivee prochaine du duc d’Anjou mon beau frere en ce royaume on seme divers bruictz autant a son desadvantage que pour irriter ses meilleurs amyz contre luy et les mectre en deffiance de sa bonne et droicte intention vers eulx, je vous diray seulement que vous debvez adviser ledict sieur duc d’Anjou lorsqu’il sera par deca de ne persister en la requeste qu’on dist qu’il a faicte pour le restablissement du comte de Leicester duquel il n’a besoing de se fortiffier et moins de craindre la force et puissance en y pourvoyant de bonne heure. Je scay que quelques catholiques en ont ja esté fort offensez, craignans de se voir derechef soubs les persecutions du temps passé ce qui les rendra constamment affectionez au duc d’Anjou d’autant plus qu’ilz le verront animé contre leurs communs ennemis.

Excusez moy vers luy de ce que je ne luy ay encores escript, craignant de le mectre en soubcon et de nuire a ses negociations, lesquelles cependant je ne fauldray d’assister de tout ce que je pourray par mes amyz. Comme cy devant je vous ay mandé, je remectz a sa courtoisie et a vostre bonne intercession les bons offices que j’espere de luy pour mon traictement en ceste captivité, la conservation de ma personne et droictz en ce royaume.

Je desirerois infiniment pourchasser une seconde visitation vers mon filz considerant l’estat present des affaires d’Escosse qui s’y offre fort a propos, mais j’apprehende le soubcon qui en pourra naistre en divers endroictz de sorte que je suis en opinion d’actendre jusques apres les nopces qu’on m’asseure debvoir estre avant ce karesme.

Si en ce prochain parlement on traicte de la succession, souvenez s’il vous plaist d’en parler a la royne d’Angleterre de ma part et d’en faire les mesmes remonstrances que je vous ay aultresfoye escriptes sur le mesme subject et occasion, a quoy me remectant je ne vous feray ceste plus longue que pour prier Dieu vous avoir en sa saincte garde.

Escript a Sheffeild ce vingtiesme janvier.

Translation:

Monsieur de Mauvissière, since with the forthcoming arrival of the Duke of Anjou, my brother-in-law, in this kingdom various rumors are being spread as much to his disadvantage as to irritate his best friends against him and to put them in distrust of his good and upright intention toward them, I will only tell you that you must advise the said Duke, when he is here, not to persist in the request that he is said to have made for the rehabilitation of the Earl of Leicester, whom he does not need in order to fortify himself, and he needs even less to fear [Leicester’s] strength and power if he takes care of it in good time. I know that some Catholics have already been greatly offended by this,¹³² fearing to see themselves again under the persecutions of the past. Such fear will make them constantly affectionate toward the Duke of Anjou, all the more so if they see him animated against their common enemies.

Apologize to him on my behalf for not having written to him yet, for fear of putting him under suspicion and of harming his negotiations, which however I will not fail to assist with everything I can through my friends. As I have already written to you, I leave to his courtesy and your kind intercession the good services I hope for from him for [the improvement of] my treatment in this captivity, the preservation of my person and of my rights in this kingdom.

I would like very much to pursue a second visit to my son, considering the present state of affairs in Scotland, which seems quite appropriate, but I fear the suspicions that may arise in various places, so I am of the opinion that I should wait until after the wedding, which I am being assured will be before this Lent.

If in this next parliament the succession is dealt with, please remember to speak to the queen of England on my behalf and to make the same pleas that I previously wrote to you on the same subject and occasion, by repetition of which I am not going to prolong this letter, but for praying to God to keep you in his holy protection.

Written in Sheffield this twentieth of January.

F105 − 20 February 1580, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 105.¹³³

Contains an enclosure for Beaton.¹³⁴

Mary has informed Castelnau via Robertson that her host, the Earl of Shrewsbury, is sick. The Elizabeth-Anjou marriage project seems to be postponed or uncertain, cooling down those who support the Duke of Anjou. Mary is threatened by the party of the Earl of Leicester who is acting jointly with the Earl of Huntingdon, her mortal enemy. Mary is being accused of being too supportive of French interests. The Earl of Leicester is openly working against France, and Stafford,¹³⁵ goaded by Lady Sheffield¹³⁶ who is now reconciled with Leicester, is spreading some sinister rumors about the duke, whom Castelnau should warn. Rumors of a quarrel between the Earl of Leicester and the Earl of Oxford¹³⁷ are saddening Mary, because of her good memory of the Duke of Norfolk,¹³⁸ and Castelnau should convey to the Earl of Surrey¹³⁹ that Mary considers him as her second son. Mary is also pleased that the latter’s sister has married Lord Buckhurst.¹⁴⁰ While the Earl of Shrewsbury was not as close to death as they say at court, she nevertheless expects him to die in the near future. She therefore asks Castelnau to ensure her safety by secretly trying to influence the choice of her new guardian, and Robertson, who knows very well the nobles of this country, can advise him on that matter. Mary has met Myldemur¹⁴¹ the day before, discussing matters related to the security of the queen and of Mary. Castelnau should thank the queen for granting a passport to Mary’s physician and plead for Mary if he meets Claud Hamilton.¹⁴²

F237 − 16 September 1580, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 237.¹⁴³

Contains enclosures for Beaton and Robertson.

Mary asks Castelnau to support Robertson on his way to France. She fears that she might be removed from Sheffield and asks Castelnau to discuss the matter with Lord Burghley.

F239 − 27(?) September 1580, Mary to Castelnau, Sheffield(?)

Reference: BnF Fr. 20506 f. 239.¹⁴⁴

Some parts of the manuscript are damaged.

Mary asks Castelnau to further assist Robertson, so that he is presented to the king and the Queen Mother, and granted a pension, and to plead to the Duke of Anjou and to Simier¹⁴⁵ on Robertson’s behalf, but without herself being mentioned. The Earl of Huntingdon, supported by the Earl of Leicester, is trying again to become her new keeper. Mary fears that France is neglecting Scottish affairs because of the marriage negotiations, and as a result, James might be driven toward a course different from hers and her predecessors’.¹⁴⁶

F98 − 31 July 1581, Mary to Castelnau, Chatsworth

Reference: BnF Fr. 2988 f. 98.¹⁴⁷

Contains enclosures for Beaton, for a relative of Thomas Morgan, and for Françoys Levin.¹⁴⁸

Mary is disappointed by the lack of support from French envoys, which would be a good reason for her to negotiate with Spain, but she denies rumors to that regard. She would not seek protection from powers other than France unless she is forced to, and it is in the interest of the king to maintain her support and the support of her friends. Mary warns Castelnau that despite the Duke of Anjou’s trying to accommodate the faction against the marriage, those of this faction are mocking him. Also, the king should be warned of Walsingham’s hostile activities while in France. Mary learned that the Earl of Leicester and Lord Burghley are acting in concert with Spain against the Duke of Anjou in the Low Countries. Mary’s enemies are also trying to assume control over her son in Scotland. She fears that Castelnau might be replaced by another ambassador who would not support her. Her movements have been further restricted, with negative effects on her health. Mary asks Castelnau to plead with Lord Burghley that she be allowed to exercise and to use her carriage, as she cannot walk or ride anymore.

F185 − 18 September 1581, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 185.¹⁴⁹

Contains enclosures for Beaton¹⁵⁰ and Thomas Morgan.

Mary has learned from Castelnau and others that France is aware of Walsingham’s actions in Scotland in concert with the Earl of Angus.¹⁵¹ She complains that the queen is becoming more hostile toward her, despite Mary not giving her any reason. As a result, a letter sent by Mary to the queen has been delayed. Her host is allowing her to exercise again. James has asked Mary to recognize him as king of Scotland and has sent George Douglas¹⁵² to France. Mary is waiting for a letter from the king and the Queen Mother before answering James, and she asks Castelnau to find out their opinion on the matter. She would also like him to plead with the queen to accept a request by James to send an envoy to Mary, and if James makes such a request, to also allow her secretary Nau to visit James in Scotland. Castelnau had asked Mary to be granted the post of seneschal of Poitou, which is part of her dowry, but she cannot grant it as it is currently occupied by Viverox,¹⁵³ son-in-law of the late Puiguillon.¹⁵⁴

F194 − 20 October 1581, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 194.¹⁵⁵

Mary is worried that an important letter she sent to Beaton via Germain Herton¹⁵⁶ has not been delivered, as it contains details about negotiations for an association whereby James and Mary are to jointly hold the Scottish crown. She asks again that Castelnau plead for the queen’s permission to send Nau to Scotland, and in that case, he may be accompanied by a representative of the queen. Mary is again worried she will be removed from her current host’s custody and asks Castelnau to find out more on that subject, and on the planned visit to her by Robert Beale.

F82 − 6 November 1581, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 82.¹⁵⁷

Mary is writing in haste so that the courier who just delivered Castelnau’s letter to her may carry back her response. She complains that a previous letter from her to Castelnau (F98) was delayed by several months (from July to November). She thanks Castelnau for his latest negotiations with the queen and the advancement of Mary’s affairs. Mary mentions that in the past, when the Earl of Leicester was closer to the queen, he had advised Mary to write openly and freely to her and told her that while the queen might sometimes get offended and angry, she eventually cools down. Mary has tried so far to demonstrate her goodwill and positive intentions toward her. She asks Castelnau to plead again with the queen for sending someone to her, hearing that since the arrival of the Duke of Anjou (in England), there is no plan to send Beale or anyone else to her. Mary threatens to revoke her prior commitments if she is not immediately granted permission to send someone to James. The queen is trying to alienate Mary from France by offering her an equivalent in England of her dowry in France and by preventing France from interfering in Mary’s or James’s affairs. Mary pledges her loyalty to France. She complains about her health, not being sure she will survive unless she is allowed some exercise, and her conditions of captivity are improved. She thanks Castelnau for meeting Lord Burghley and the Earl of Leicester, but the latter’s recent inclination toward the Earl of Huntingdon makes her doubt his former promises. Mary asks Castelnau that as part of the marriage negotiations, persecutions against Catholics in England be eased. Castelnau should also plead with the Duke of Anjou so that legal proceedings by the Sieur de Rosne¹⁵⁸ against her be stopped. She also asks that any correspondence with Thomas Morgan be concealed from anyone else. She is happy that Mildmay has been commissioned by the queen to visit her, but while in the past he had left a positive impression, he has since become closer to the Earl of Huntingdon. She also mentions the late “traitor” Morton, who had been supported by the queen, and the recent plots by Thomas Randolph.¹⁵⁹

F21 − 9 January 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 21.¹⁶⁰

Contains enclosures for Beaton and Thomas Morgan.

Her health has improved but she is disappointed that her requests for better conditions have not been answered and asks Castelnau to intervene on her behalf with Walsingham and Beale. During his last visit, Beale had urged her against granting her son’s request to be named King of Scotland, and insisted that instead, she and her son should sever her prior alliances with other princes, relying on the queen to put things in order in Scotland. Beale also assured her she would be treated well while in captivity, and that she would be allowed to send someone to James. Mary thinks Beale was sincere and hopes that he will be sent with her own envoy to Scotland. But meanwhile, to counter France’s influence in Scotland, the queen is trying to convince the Spanish king that an agreement between Mary and James would be against Spanish interests. Members of her council profess their affection to France and the Duke of Anjou, but in fact they are acting against the proposed marriage, spreading rumors that the duke plans to usurp the English throne, and if the marriage takes place, they are ready to take arms. Her hostess is trying to bring her husband into the faction against the marriage. Catholics in England hope that the Duke of Anjou will act in their favor, as several Jesuits have been recently executed, but the Earl of Huntingdon is trying to convince some Catholics that the duke is not really interested in religious matters, and that those supporting the marriage will be punished anyway. Castelnau should warn the queen against those actions, but he should make sure not to reveal that the information came from Mary. Mary thanks Castelnau for his assistance regarding the legal action of the Sieur de Rosne against her and asks that the Duke of Anjou pay him the agreed sum. Mary has not written via the official channel, as the queen is currently fully preoccupied with the duke’s visit.

F235 − 19 January 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 235.¹⁶¹

Referring to her prior letter from 9 January 1582 (F21), Mary adds that her hostess has been won over by the faction against the Elizabeth-Anjou marriage, being promised that her granddaughter Arbella Stuart would be married to one of the Countess of Essex’s sons.¹⁶² Mary is worried that this faction would try to remove her from Sheffield, and asks Castelnau to prevent this. He should also warn the Duke of Anjou against this faction. Mary mentions the positive answer sent to her by the king via Castelnau about what she previously wrote regarding her and her son. She also recommends that the king support the Duke of Lennox, as the best means to restore France’s influence in Scotland.

F109 − 23 February 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 109.¹⁶³

Contains an enclosure for Beaton.

The Duke of Anjou has left England. Mary has heard that the Earls of Leicester and Hatton¹⁶⁴ have convinced the queen to renew an alliance with Spain against France, because after having offended the duke and refused an offer for a defensive and offensive league with France against Spain, England is now left without any support. She claims that the queen would eventually support Spain, regardless of the marriage taking place or not. Mary asks Castelnau to plead for Nau being allowed to go to Scotland. She does not want to wait anymore, as the queen is meanwhile acting against her in Scotland by supporting the Earl of Angus. She asks again that the Duke of Anjou pay the agreed sum as part of the settlement of Rosne’s claim against Mary.

F163 − 28 February 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 163.¹⁶⁵

Contains an enclosure for Beaton.

Mary is asking again Castelnau to plead for Nau being allowed to go to Scotland with someone of the queen’s choice, claiming this would benefit all parties. Mary’s enemies in Scotland are working against her son, but if she could communicate with him, and if he trusted her, he would surely be able to counter those factions with help from France. Mary assures Castelnau she wouldn’t do anything without the king’s advice. She thanks Castelnau for obtaining a passport for du Ruisseau, so that he can assist her in her affairs as well as in the exercise of her religion. She also asks Castelnau to provide her with some money. Mary would like her own physician to be sent to her, rather than those commissioned by Walsingham, the Earl of Huntingdon, and the Earl of Leicester. The latter, having recently accompanied the Duke of Anjou to Flanders, is bragging of having convinced the duke of his being his best friend in England, and of encouraging him to stay in Flanders thus committing to protect the “drunk” Flanders people and renewing the League of Burgundy.¹⁶⁶ Mary is ashamed that the duke’s enemies, who are now in a stronger position, are mocking him. Mary asks Castelnau to obtain an attested copy of a memorandum that Beale received, to ensure that her host will follow the queen’s instructions, rather than Walsingham’s.

F125 − 7 April 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 125.¹⁶⁷

Mary complains that several of her letters have been delayed. Using the codename “la Tour” she introduces a new courier, Francis Throckmorton, whom she holds in high esteem, but without mentioning his name in the letter. He resides in London and will be able to handle urgent communications between her and Castelnau. Mary is sorry that the Duke of Anjou is fighting Spain for the benefit of (Dutch) Protestants who will fail to support him in the future, thus badly tarnishing his reputation. The queen had previously sent Greville¹⁶⁸ to the Prince of Orange,¹⁶⁹ and having learned of the attempt to assassinate the latter, she and her council will try to bring some of the Dutch provinces under her protection if he dies from his wounds, forcing the Duke of Anjou to share his control over the Low Countries. The duke was wrong to neglect his relations with the Earl of Sussex¹⁷⁰ and Lord Burghley, and instead rely on the false promises of the Earl of Leicester, who is in fact his worst enemy. Mary also writes that the duke took too much on his shoulders, more than he can carry. Mary has asked her staff to hold prayers for the duke as ordered by the king of France. Mary asks for clarifications on a proposal that she transfer her rights to the Scottish and English thrones to her son. Her host has informed her that her requests to send two physicians, to allow du Ruisseau to visit her, and to allow her outside exercise have been granted. Beale had previously advised her to write to James asking him to apologize for his refusal to let Captain Errington,¹⁷¹ the queen’s envoy, enter Scotland, as a precondition for Nau being allowed to go there. Castelnau should thank Beale, and assure him of Mary’s intention to reward him, also asking him to be the one to join Nau in Scotland. Mary is unhappy that de Chaulnes has not reimbursed Castelnau, who has apparently paid for Mary’s carriage.

F165 − 16 April 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 165.¹⁷²

Contains an enclosure for Beaton.

Referring to her letter from 7 April 1582 (F125), in which she wrote that her host had told her of several of her requests being granted, she excuses herself that a pain in her right arm has prevented her from writing to James as discussed before. Mary found out that Walsingham had been acting against James, and that he and the queen’s council want to stir up a rebellion in Scotland. They want James to move away from France, claiming that the king of France is in poor health, with no child to succeed him, and that his successors would be inclined toward the English queen. Another false claim is that the king of France’s hatred toward the House of Guise will prevent the Guises from assisting James, who should not rely on assurances from such a weak country as France. Mary writes that a letter, signed by five members of the queen’s council and asserting these claims, was sent to Scotland. Castelnau should warn the king that the negotiations for the marriage are not sincere, the queen and her council undecided on whether to satisfy the king and the Duke of Anjou or to reconcile with Spain. Mary is worried she might be removed to Windsor or to the Tower of London, which would put her life in danger, and asks Castelnau to intervene on her behalf with Lord Burghley. Mary also heard that in the past, some members of the queen’s council had pleaded with the queen to name her successor, and she asks Castelnau to act in her favor. Asking for the first name of a potential new courier named Bezet mentioned by Castelnau, she writes to him that if this is Henry Bezet,¹⁷³ Castelnau should not trust him as he recently traveled with the Earl of Leicester to Flanders, but if this is Patrick Bezet,¹⁷⁴ he may be tasked to carry letters to Scotland.

F243 − 22 April 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 243.¹⁷⁵

Contains an enclosure for Beaton, and a package for Bodin.

Referring to her last letter from the previous week (F165), Mary mentions rumors about discussions on the succession to the English throne and asks Castelnau to make sure nothing is decided which would be against her interests. She has sent the letter to James that Beale had requested,¹⁷⁶ so that James should apologize and accept the queen’s envoy, and asks Castelnau to find out whether it has been delivered. Castelnau should not disclose that Mary has already agreed to an association with James to jointly hold the Scottish crown, and she also wants him to find out more about the dealings of the Earl of Leicester with her hostess. Mary sends with this letter a package to Sieur Bodin,¹⁷⁷ asking him to write arguments in favor of Mary’s rights to the succession to the English throne. She also asks him to promptly forward any secret letter from Beaton or Morgan.

F223 − 27 April 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 223.¹⁷⁸

Mary did not get definitive answers from the queen and her council regarding her prior requests. She asks Castelnau to find out more about their intentions and those of the Earl of Leicester. She also asks Castelnau, in case Nau is allowed to go to Scotland, to instruct him on what he is allowed to negotiate, and to provide him with a letter of recommendation to the Duke of Lennox.

F170 − 10 June 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 170.¹⁷⁹

Contains an enclosure for Beaton.

Mary is concerned that she did not get several secret letters she was expecting from her ambassador, asking Castelnau to deliver them as soon as he receives them. Mary discovered that the two physicians sent to her are against her and depend on Walsingham. One of them, named Smyth, is a follower of Paracelsus¹⁸⁰ who recommended to her the drinking of gold water. Mary asks Castelnau not to forward to her any medicine prescribed by those two physicians. She is still waiting for an authorization for Nau to go to Scotland. She fears her liberty might be further restrained, and she might be removed to the Tower of London, in which case, she would want the Duke of Guise to negotiate on her behalf. Mary is satisfied that Castelnau is now aware that the queen, given England’s poor state of affairs after having offended almost all the princes of Christendom, is trying to fool him with false promises. She warns him that the queen may try to stir things up in France. Mary and her son are currently too weak, and they need the help of the king and the Duke of Anjou, and meanwhile, she will wait patiently. She also asks Castelnau to diligently forward to her any letter from George Douglas.

F158 − 31 July 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 158.¹⁸¹

Contains enclosures for Beaton and for Henry Howard (a cipher key).

Mary thanks Castelnau for warning her son and the Duke of Lennox of a plot by the Earl of Angus. If a rebellion takes place in Scotland, she wants France to intervene, in the name of the old league between Scotland and France, which France should value, Mary being loyal to France and being able to ensure her son’s loyalty. France should prevent Scotland from falling under England’s control, and Castelnau should also warn the queen not to intervene in Scotland, trusting James will be able to crush any rebellion. Some in the queen’s council are acting against the Duke of Anjou in Flanders, fearing him more than they fear Spain, and they are trying to restore the League of Burgundy. Mary asks Castelnau to find out more about the Earl of Angus’s plot, and to plead to the king for her protection and the protection of her son. Walsingham has been withholding a letter from her son for more than three weeks.¹⁸² She asks Castelnau to find out more from Walsingham about the developments in Scotland that Castelnau had reported to her via Nau. She also wants to know of the latest developments on the Elizabeth-Anjou marriage, although she does not believe it will take place. She is expecting a visit by Beale, followed by a trip by him to Scotland together with Nau, but she does not want such a trip to take place, being concerned that this would allow England to interfere against her and her son.

Postscript from Nau: Mary responds favorably to the king’s request to grant a prebend of St. Quentin to the Sieur de Saint-Prix.¹⁸³

F229 − 2 September 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 229.¹⁸⁴

Contains an enclosure for Beaton.

Mary just learned of the capture of her son by the partisans of the Earl of Angus. There is talk of them moving toward the Duke of Lennox, who may be assembling an army, but she cannot know for sure, being restrained, and unable to send any letter, such as a letter to the Duke of Lennox she wishes to send. She asks Castelnau to find out more about the recent developments, and to inform her of any new developments as soon as possible. Mary also wants Castelnau to write to the king to make sure the queen does not get further involved in Scotland, and she asks that the king allow her family, the Guises, to secure James and his servants.

F151 − 10 September 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 151.¹⁸⁵

Contains an enclosure for Beaton.¹⁸⁶

Mary thanks Castelnau for having warned the Duke of Lennox of plots against him, and for informing her of the developments in Scotland, but she feels helpless, with her son in the hands of their worst enemies. She thanks him for also informing the king and the Queen Mother, hoping they will show compassion and assist her. She wants the king to send someone of quality to Scotland.¹⁸⁷ She warns that the Duke of Anjou should not rely on the queen’s assistance in the Low Countries, as, if he is successful, England would eventually ally with Spain against him. Mary asks Castelnau to find out more about the actions of the queen and her council, and the results of George Carey’s¹⁸⁸ visit to Scotland. Mary suspects that Archibald Douglas¹⁸⁹ is working with Walsingham, but he still may prove his loyalty to her by obtaining intelligence on the plans of her enemies in Scotland. Mary refuses to assist the illegitimate daughter of the late Kirkcaldy of Grange,¹⁹⁰ and her mother in particular. Mary asks Castelnau to send her a letter via the official channel describing the current situation in Scotland, so that she has a reason to write to the queen.

F118 − 24 September 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 118.¹⁹¹

Contains an enclosure for Beaton.

Mary thanks Castelnau for informing her about recent developments in Scotland and the arrest of Archibald Douglas.¹⁹² She has learned that the Duke of Lennox is still in Scotland, being safe at this stage, but she fears he may be captured by the rebels, and she asks Castelnau to try to obtain the support of the king and the Duke of Guise for his restoration. Du Ruisseau has been arrested, further restricting her ability to communicate with the outside world, and she asks Castelnau to intervene in his favor, as his detention is unjustified. She even suggests that France should reciprocate by arresting some Englishman in France, to preserve the king’s honor. She also asks Castelnau to write to her via the official channel. In a postscript, Nau asks Castelnau to assist du Ruisseau, who is Nau’s brother-in-law.

F245 − 13 October 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 245.¹⁹³

Contains enclosures for Howard and du Ruisseau.

Du Ruisseau, who has left Sheffield, will verbally provide Mary’s answers to Castelnau’s latest letters. Mary asks Castelnau to prevent any further intervention by the queen in Scotland, and if he is allowed to go to Scotland, which is unlikely, he may refer to James as king, but only for that specific visit. Mary now agrees to assist the illegitimate daughter of the late Kirkcaldy of Grange and recommends that she be introduced to her aunt in St. Pierre.¹⁹⁴ Castelnau should find out whether Archibald Douglas can be trusted and what he has learned about English plots against her and her son, but she warns again Castelnau that he may be a double agent.

F74 − 25 October 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 74.¹⁹⁵

Contains enclosures for Beaton, for du Ruisseau,¹⁹⁶ and for Henry Howard.¹⁹⁷

Mary refers to her previous letter (F245), as well as to verbally answering Castelnau via du Ruisseau. She finds it strange that Castelnau did not request the queen’s permission to go to Scotland despite being instructed to do so by the king. The king’s declared interest in protecting Mary and her son was enough to prevent her enemies in England from intervening in Scotland, and while she respects Castelnau’s trying to maintain good relations with the queen by not making requests that would irritate her, she writes that this would only have the effect of the queen’s feeling free to act against France’s interests. While Mary was given all sorts of excuses related to the situation in France, nothing in her eyes justifies abandoning her and her son to the rage and cruelty of her enemies. Mary reiterates her request that the king and the Queen Mother send someone of quality to Scotland, asking Castelnau to let her know of their intentions in that regard. She also wished the little liberty she previously had would be restored, having been severely restrained following the incident with du Ruisseau. With her health deteriorating, she fears that she soon may die, while she is even not allowed to write and ask for help. Regarding her letters sent via the official channel to Castelnau being shown to the queen and her council, she argues that while she does not write anything sensitive in those letters, exposing her dealings about her dowry estates would be prejudicial to the king, and therefore she needs to maintain a confidential channel with Castelnau and France. As previous letters from Castelnau had been held for two months, she asks him to report on du Ruisseau and her requests from the queen. Morgan has informed her that he has notified Castelnau of Parry¹⁹⁸ being a spy working for Walsingham. Mary reiterates her request to obtain intelligence from Archibald Douglas. She will also ask her treasurer to reimburse Castelnau.

F50 − 31 Oct 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 50.¹⁹⁹

Mary thanks Castelnau for assisting du Ruisseau and forwarding to him her instructions about her estate and her requests to improve her captivity conditions. She has from both du Ruisseau and her host that the queen’s answers to her requests were somehow positive, but Walsingham, after his mischievous actions in Scotland, should not expect that those answers will appease Mary, and she is resolute not to let her enemies abuse her patience. Mary has also been informed that the Queen Mother has instructed Castelnau to proceed with his recent overture to negotiate her release, and Mary asks him to pursue this diligently, and not settle with evasive responses. Castelnau should not call James King of Scotland, as her agreement that he be called as such by the French king’s ambassador when he goes to Scotland is only for that specific visit and because of the pressing necessity to get access to James. This title of king can be granted and referred to by foreign princes only after the association between James and Mary to jointly hold the Scottish crown is published by the parliament. Mary thanks Castelnau for sending her James’s portrait and wants to know when it was painted and by whom because she finds it different from other portraits she had already received. Mary complains of lax communication security protocols, after a box sent via the official channel by du Verger²⁰⁰ containing an enciphered letter from Beaton together with a letter in clear from Castelnau was inspected by her host, but luckily, the letter in cipher was not discovered.

F123 − 12 December 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 123.²⁰¹

Contains enclosures for Beaton, and for de la Mothe-Fénelon (a letter (F307), and a cipher key).

This is essentially the cover letter of an attached letter (F307) to de la Mothe-Fénelon, to whom Mary asks Castelnau to hand over its deciphered version. She is also sending a new cipher table to be used by de la Mothe-Fénelon to communicate with her. Castelnau should also give de la Mothe-Fénelon a copy of Mary’s latest letters to the queen, so that he can act on them on behalf of the king. Mary complains she has been in captivity for 14 years, and she fears she may not be able to survive her miserable conditions. She wants de la Mothe-Fénelon to plead for her with the queen and her council, so that she and her son get assurances from them that they won’t act in mean ways against them, stating that a good agreement can erase the past, with the queen in a strong enough position to force Mary to abide by any agreement they may be able to reach. Refraining from pleading in order to not irritate the queen will only result in the queen not respecting her promises, and the Duke of Anjou should not rely on her false friendship. Mary can’t believe that out of respect for the queen, she and James will be abandoned to the mercy of their enemies, who would benefit from the king’s lack of support to Mary and James, and she hopes that de la Mothe-Fénelon will demonstrate the contrary by encouraging Mary’s friends. She also asks that 100 ecus be paid to Mosman,²⁰² and 40 ecus to a lady called Paterson.²⁰³ She also warns Castelnau not to trust Fowler.²⁰⁴

F307 − 12 December 1582, Mary to de la Mothe-Fénelon, Sheffield

Reference: BnF 500 de Colbert 470 f. 307.²⁰⁵

Mary is satisfied that de la Mothe-Fénelon has been selected as the king’s envoy to Scotland, given his experience and past good service, and she provides him with a detailed list of the topics she wishes him to take care of:

• Even though the main purpose of his visit is to see James, de la Mothe-Fénelon should also be allowed to visit Mary, to learn about the true state of her affairs, and discuss what to do in Scotland for James.

• Nau should be allowed to join him, as he is familiar with Scottish affairs, and the queen should have no reason to refuse this request, because she also wants to send someone on her behalf with de la Mothe-Fénelon.

• De la Mothe-Fénelon should congratulate James in person on the agreement on the association with Mary, pressing him to make public its articles as they were communicated by Mary to the Duke of Guise, as it is James’s only means to justly acquire the title of King of Scotland, and to be recognized as such by the princes of Christendom.

• De la Mothe-Fénelon should negotiate with those near James to ensure that the reestablishment of his freedom is made public. For that purpose, if there are any armed forces guarding him, they should be dismissed immediately.

• Any nobleman or person of quality who was not previously in his council or his court should be sent away.

• The Duke of Lennox should be reinstated in his previous charges. When de la Mothe-Fénelon enters Scotland, he may communicate with the duke through the Laird of Ferniehirst,²⁰⁶ whom Mary has ordered to meet de la Mothe-Fénelon at the border. De la Mothe-Fénelon should oppose the Duke of Lennox’s leaving Scotland, who should rather proceed and act with his armed forces. The Earl of Arran²⁰⁷ should be released and returned to his role as the guard’s captain.

• George Douglas, brother of William Douglas of Lochleven, and others still in prison should be released and any charge against them dismissed. Archibald Douglas should also be released.

• The conspirators may be pardoned by the parliament, but they first must agree to retire to their homes. De la Mothe-Fénelon may also secretly negotiate with James, given the French king’s approval, on the renewal of the old league between France and Scotland. He may also approach the Earls of Argyll, Huntly,²⁰⁸ Athol, and other noblemen supporting France. Mary was informed that the son of Lord Hunsdon²⁰⁹ was recently sent to Scotland to convince James to abandon his alliance with France, offering him an annual pension of 20,000 pounds, promising to him that he would be declared second in the line of succession to the English throne, and proposing that he marry the daughter of the Earl of Derby²¹⁰ or the Countess of Essex. De la Mothe-Fénelon should remind James of his promise not to marry anyone without Mary’s advice and agreement. The Queen Mother had previously mentioned the Princess of Lorraine,²¹¹ whom Mary sees as a suitable candidate.

• If de la Mothe-Fénelon is unable to obtain James’s release, he should work with those of the “good” party, so that James may escape and return with Lennox. James should agree to that, given the current risk on his life or of his being removed to another place. To bring the conspirators to reason, de la Mothe-Fénelon should threaten them that the king will stop any commerce between France and Scotland. He should also reassure those of the good party with pledges of assistance and financial support if they are attacked by the conspirators. De la Mothe-Fénelon may remind the Earl of Bothwell, the son of Mary’s illegitimate half-brother,²¹² of his previous promises of loyalty toward Mary, so that if he is currently acting with the conspirators, he should detach himself from them.

Mary also asks de la Mothe-Fénelon to insist with the queen on the points mentioned in a letter Mary had sent to her, of which Castelnau has a copy. She would also like the king to appoint a permanent ambassador to Scotland. This should not offend the queen, who has already appointed an ambassador to represent her in Scotland.

F190 − 18 December 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 190.²¹³

Contains enclosures for Beaton, for de la Mothe-Fénelon (a letter – F198, and a cipher key), and for Mosman.

Referring to her letters from December 12 (F123 to Castelnau, and F307 to de la Mothe-Fénelon), Mary writes that she has learned that de la Mothe-Fénelon has not yet been authorized to pass from England to Scotland. She asks Castelnau to be informed on any development regarding the negotiations carried by de la Mothe-Fénelon, and, if he is still in England, Castelnau should hand over to him an enclosed letter (F198).

F198 − 18 December 1582, Mary to de la Mothe-Fénelon, Sheffield

Reference: BnF Fr. 20506 f. 198.²¹⁴

Mary hopes that not having been allowed to go to Scotland, de la Mothe-Fénelon has already advised the king to send someone else (directly) by sea. She also would like him, when he is back in France, to plead with the king and the Queen Mother about her conditions in captivity, and the state of her dowry. She is disappointed that she and her son have received little support from France, despite their loyalty. Thus, they have been abandoned to the cruelty of their enemies, and she would be grateful to de la Mothe-Fénelon if he is able to relieve them from their current misery.

F225 − 31 December 1582, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 225.²¹⁵

Contains enclosures for Beaton, for the Duke of Lennox, and for Henry Howard.

Mary has learned that de la Mothe-Fénelon has finally been allowed to go to Scotland but finds it strange that he has been further delaying his trip. Mary now considers this trip useless, the delay having enabled the queen to act in concert with the conspirators in Scotland, and at this stage, de la Mothe-Fénelon should expect Lennox to be expelled. James cannot be safe without Lennox, and the latter’s reestablishment should be the king’s top priority. If Lennox is already gone before de la Mothe-Fénelon arrives in Scotland, de la Mothe-Fénelon should not enter into an agreement with the traitors, as any such agreement would prejudice Mary. As Lennox is about to travel to France via England, she asks Castelnau to hand over to him an attached letter. Mary expects the queen to appoint 100 horsemen to guard James more closely, not showing any respect for the king’s request to the queen to release James and improve his conditions. Walsingham has been holding all (official) letters from Castelnau to Mary for the last two months, and Mary asks Castelnau to press Walsingham to deliver them.

F153 − 15 January 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 153.²¹⁶

Since de la Mothe-Fénelon’s arrival in England, Mary has not received any news from him or from Castelnau, neither via the official nor via the secret channel, and she fears something bad has happened with Castelnau’s letters. She can’t imagine that de la Mothe-Fénelon, sent on the king’s orders after Mary’s urgent request to assist her, has willingly neglected to write to her during his long stay with the queen. Given the long delays, and now that Lennox is out of the country, Mary does not think anything positive may result from his upcoming trip. Mary has been told that Castelnau will be replaced by another ambassador and asks him to clarify on the subject. Mary would like to know how the Duke of Lennox was treated during his sea journey, whether he has received her letter, and whether he has already replied to it. Mary also asks Castelnau to enquire with the queen on the two women servants she requested, as they are needed for the winter, and Mary is feeling better for the first time in quite a while. Castelnau should talk to Beale about his negotiations with Mary during the previous winter, and she understands the subject has been discussed with the queen and her council. Castelnau should thank Beale for his good services and his honorable report, mentioning her positive opinion of his sincerity toward her. Castelnau should claim that he has heard those positive words from du Ruisseau, and not directly from her, in order not to expose their secret communications. The persecution of Catholics in England is getting worse, under the cover of the queen’s apparent friendship with France, and it is even said that some Catholics were discovered with the help of members of Anjou’s council. Castelnau should ask the king not to let her enemies think she is being abandoned and neglected.

F247 − 20 January 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 247.²¹⁷

Mary has learned that her two previous letters had been delayed (F190 and F225), therefore Castelnau is now receiving those two and the current one together. Mary thanks Castelnau for discussing her release and the improvement of her conditions with the queen, but she feels that de la Mothe-Fénelon did not act according to the instructions he was given by the king and the Queen Mother. She hopes that when de la Mothe-Fénelon is back from Scotland, he and Castelnau jointly promote these topics with the queen. Some of Mary’s friends are puzzled by the fact that de la Mothe-Fénelon went to Scotland almost as a prisoner, accompanied by William Davison,²¹⁸ the queen’s envoy, thus limiting his ability to act freely and negotiate with anyone other than James, which has been detrimental to the prestige of the French king. Mary is unhappy that her letters have not reached the Duke of Lennox on time, asking Castelnau again to deliver them to him, and to let her know of the duke’s discussions with the queen and her council.²¹⁹ Mary writes that what the queen is claiming, based on an intercepted and deciphered copy of a letter from Mary to Beaton, is completely false, denying ever writing “such things” to him,²²⁰ and Castelnau should insist on seeing the original letter in order to verify its authenticity. Castelnau, who has informed Mary that her request for two chambermaids and two officers has been approved, is to ensure that there is no difficulty with their dispatch.

In an addendum (presumably written after January 20), Mary reports that she was informed of Lennox’s negotiations with the queen. She warns Castelnau of (William) Fowler, whom she knows to be dependent on (Thomas) Fowler who once served the Countess of Essex but is now in the service of the Earl of Leicester and had started the slander against her.²²¹ Since the mother of the illegitimate daughter of the late Kirkcaldy of Grange has refused to support her own daughter, Mary does not want to assist her, accusing her of being a man chaser who only wishes to move to another country. Rather, Mary would have her whipped while tied to the back of a carriage if she had the authority to do so.²²²

F64 − 20 February 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 64.²²³

Contains an enclosure for Beaton.

Mary has not received any letter from du Ruisseau or from her council in France since his departure (in October 1582). She asks Castelnau to enquire with Walsingham about the delivery of (official) letters, as the matter is important for the management of her affairs in France. There should be no legitimate reason to deprive her of such communications, but the queen, wishing to ruin her and her son, wants to deny her any intelligence. Castelnau should plead with the queen that those communications are essential in order for her to manage her dowry. The disastrous failure of the Duke of Anjou in the Low Countries is being badly interpreted in England, his enemies taking advantage of the news to spread rumors that the duke, during his visit in England, had planned to take the Tower of London by surprise and make himself the master of the city. There is little faith in everything that comes from the queen and her council. Mary regrets seeing the duke, her brother-in-law, having been fraudulently manipulated so that he has lost the armies he had employed for the benefit of (Dutch) heretics rather than for the protection of those of his religion in England. The Earl of Leicester has started to form a party, in order to usurp the English crown in imitation of his father²²⁴ by marrying his son to Arbella, granddaughter of Mary’s hostess, the Countess of Shrewsbury. Mary fears they would also usurp the Scottish throne, and James may be killed. Castelnau should warn the queen of those plots and make clear to the queen and her council that she has not renounced her rights to the crown of Scotland. Castelnau should also inform Lord Burghley that the Earl of Leicester has sent portraits of his son to Mary’s hostess, without letting him know this information came from Mary. Mary has learned that a French gentleman on his way to meet de la Mothe-Fénelon in Scotland has been arrested near Wetherby, 20 miles from Sheffield in the direction of Berwick, as he had no passport with him. Mary was told that de la Mothe-Fénelon has proposed a marriage for James in France, and the renewal of the alliance between France and Scotland, but she is displeased. Any such marriage would need her permission, and any alliance should include herself as the association stipulates shared ruling, and for now any decision on an alliance should be made solely by her, and any alliance done without her would be nullified.²²⁵ Castelnau should plead with the king that her agreeing to grant the title of king to James should not be turned against her.

F174 − 28 February 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 174.²²⁶

Contains enclosures for Beaton, for lord Henry Howard,²²⁷ and for de la Mothe-Fénelon.²²⁸

Mary informs Castelnau that she is sending a letter to the queen via the official channel.²²⁹ She complains to Castelnau that she has not received anything from her ambassador in the last five months. More than one year ago, she had warned the Duke of Lennox that the Earl of Leicester and Walsingham were plotting with Colonel William Stewart²³⁰ against him, but she thinks that it should now be possible to restore Stewart’s affection to those bearing his name, and if the king agrees, a good pension would convince him to switch sides. Castelnau should ask de la Mothe-Fénelon to write to Mary about his negotiations in Scotland, and if he is not willing to do so, at least, to let her know of her son’s health. Mary heard that de la Mothe-Fénelon is to sail to France together with Colonel Stewart, appointed to be James’s ambassador in France, in which case she would like Castelnau to find out the state of affairs in Scotland, and the intentions of the conspirators. Castelnau should convey to Stewart her good opinion of him, after her reconciliation with the late Lady Lennox, and the marriage of Lady Lennox’s son with the daughter of Mary’s hostess, which Stewart had mediated.²³¹ If the king would declare his good intentions toward herself and James by advancing her affairs, she would be most grateful, but since she is not aware of any efforts in that direction, she will have no choice but to conclude that she is being neglected and abandoned. Castelnau should press for obtaining a passport for her two chambermaids and for her officers,²³² so that they can arrive at Sheffield before she leaves for the baths, as she is about to request permission for. Mary thanks Castelnau for sending her some wine, which she considers to be the best, which was just in time because the wine ordered by her host for Lent has not yet arrived from France. If Baldwin²³³ comes to Sheffield, he should bring with him the 2,000 ecus she had requested from Castelnau, otherwise Castelnau should send the money by the first occasion, since she is in great necessity, and even 1,000 ecus would do.

F89 − 10 April 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 89.²³⁴

Expecting Beale to arrive, Mary will promptly notify Castelnau of anything of importance, asking him to find out the real intentions of the queen and her council regarding the negotiations. Mary’s couriers, de la Tour²³⁵ and George More,²³⁶ are ready to fetch letters from Castelnau at any time. Castelnau should also find out why her host, the Earl of Shrewsbury, is being called to the court. Mary thanks Castelnau for the news about her son and the present state of affairs in Scotland, not having heard anything from de la Mothe-Fénelon. If the king does not decide soon to support those in Scotland who remain unwilling to accept the conspirators tyrannical rule, Mary fears they might be forced to comply, also seeing that the queen of England is helping the conspirators. The visit of de la Mothe-Fénelon and Mainneville²³⁷ would bring little benefit and would rather trigger the queen to act against her and her son. The queen’s pernicious designs will then extend to the king and the Duke of Anjou, and the recent events in the Low Countries are only a proof of what she had been warning all along, that is, trying to please the queen would only strengthen her and give her more confidence to act against the interests of the king and the Duke of Anjou, the grandeur and prosperity of whom she fears, especially if the king brings back Scotland under his influence. Knowing her natural character, it is better to keep her in fear of not being able to rely on any of the other princes of Christendom. Castelnau should plead with the king about Mary’s miserable condition and the conditions of her son, who is further restrained by the conspirators and by the new guard the queen has raised. Also, Castelnau should prevent James’s removal to another place by the conspirators, who accuse their opponents of taking arms. Mary knows that the Earl of Leicester is behind those developments, driven by his old ambitions, and Castelnau should warn the queen of the latter’s plans together with those of the Countess of Shrewsbury, her hostess, to marry Leicester’s son to Arbella Stuart, thus being able to claim the English throne as well as the Scottish throne by marrying James and the Countess of Essex’s young daughter.²³⁸ But the queen should not know that this information is coming from Mary. Mary thanks Castelnau for his financial assistance, as she is not receiving from France what is owed to her, her dowry having been reduced to half of what she owns. Mary heard that the Duke of Anjou is going to meet the Queen Mother in Calais. Castelnau should thank the duke for the recommendation he recently made for her affairs in France by M. de Rambouillet,²³⁹ and he should also ask him to discuss the matter with the Queen Mother. Mary would like to get a copy of Rerum Scoticarum Historia, written by George Buchanan²⁴⁰ and printed after his death. If the book is already in circulation, it should be forbidden, as it hurts the honor of her son, of herself, and of her predecessors. Because she is now assured of her son’s devotion toward her, she provides her consent that from now on, Castelnau may grant him the title of King of Scotland on all occasions. She had previously delayed this nomination, fearing that it would benefit the conspirators in Scotland. She also asks Castelnau to write to her as often as possible while Beale is in Sheffield, and to provide her with his advice.

Postscript by Nau: Nau mentions a letter sent recently by Mary via the official channel.²⁴¹ Since Mary has granted Edward Foster²⁴² the pension of the late Andrew Hamilton,²⁴³ Nau would like to employ him, but he would first need to know his age, qualities, and religion. Nau cannot directly write to Foster, as he would not want to send a letter in clear together with the current package of letters in cipher. He also writes that he is trying to satisfy Jerome Pasquier’s requests.²⁴⁴

Additional postscript by Mary: She acknowledges receipt of a packet from Francis Throckmorton. She fears that she may be moved to another location, putting her life in greater danger. Regarding the Jesuit²⁴⁵ recently captured in England, Mary is sure nothing will be found to connect him with her.²⁴⁶ She asks Castelnau to assist in her request to go to the Buxton baths, and thanks him for the 200 pounds she received from him.

F130 − 16 Apr 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 130.²⁴⁷

Contains enclosures for Beaton, and Throckmorton.²⁴⁸

We reproduce here the full decryption and a tentative translation.

Monsieur de Mauvissiere, j’eusse bien desiré que ce porteur eust peu actendre encores quelques jours pour partir affin que par luy je vous peusse donner quelque certaineté des negociations de Beale par deca esquelles il n’est jusques icy entré que superficiellement m’ayant faict quelques ouvertures pour ma delivrance, restablissement des affaires en Escosse et d’une bonne amitié entre sa maistresse, moy et mon filz et noz estatz, mais le tout si en general qu’il ne s’en peut prendre aulcun fondement, m’estant de mon costé retenue pareillement d’entrer en aulcunes particularitez, voyant que ledict Beale n’est credité ny aucthorisé pour ce regard par aulcunes lettres, memoires ou instructions de la royne d’Angleterre ny de ceux de son conseil de sorte que, comme je luy ay dict, tout traicté entre luy et moy de ceste facon ne pourroit estre solide ny substantiel, ains serviroit seulement a descouvrir inutilement mes intentions, comme il sembloit estre sa principale charge ce que je n’estois deliberee de faire de bonne asseurance d’en venir a quelque effect, me souvenant par trop bien du peu qui reussit de mes conferances avec luy en son dernier voyage ores que je creusse qu’il y eust sincerement procedé et en homme de bien.

Voyla a quoy nous en sommes demeurez ; je ne fauldray de vous tenir adverty de ce qui en en suivra cy apres et cependant je vous remercye de vos bons et amples advis par vos dernieres que ledict Beale m’a apportees, suivant lesquelles, si les choses preignent quelque acheminement par deca, je pourray bien requerir qu’il me soit permis d’envoyer Nau vers la royne d’Angleterre pour luy proposer et remonstrer de ma part ce que je ne vouldrois commectre qu’a elle mesme, et ne trouverrois poinct hors de propos que des ceste heure vous en feissiez comme de vous mesme instance, mais s’il fault a bon escient entrer en traicté et qu’il y aye esperance d’en venir a conclusion, il me conviendray employer quelques personnaiges de plus grande qualité avec lesquelz je le pourray joindre pour leur plus ample instruction sur toutes occurrances necessaires dont il est de longtemps assez informé. Je ne scay comme assez vous remercier du soing, vigilance et entierement bonne affection avec lesquelz je voy que vous embrassez tout ce qui me concerne et vous prie d’y continuer plus vivement que jamais, specialement pour madicte delivrance a laquelle je voy la royne d’Angleterre assez enclinee. S’il en reussit aulcune chose, asseurez vous que ce ne sera de recongnoistre aultant qu’il sera en moy vos bons offices en cest endroit et l’obligation que je vous ay du passé.

Sheffeild, ce seiziesme apvril.

La lettre marquee est pour le sieur de la Tour et l’aultre marquee pour monsieur de Glasgo, ambassadeur de Sa Majesté. Nau baise bien humblement les mains a monsieur de Mauvissiere, remectant a ce porteur, qui a charge de communiquer avec luy sur toutes occurrances, de luy faire entendre toutes nouvelles de deca.

Translation:

Monsieur de Mauvissière, I would have very much liked this bearer to have been able to delay his departure for a few more days so that through him I could give you some certainty on the negotiations with Beale, into which he has so far entered only superficially, having made to me some overtures for my deliverance, restoring the affairs in Scotland and a good friendship between his mistress, myself and my son and our states, but in too general terms for one to rely upon, myself being similarly restrained from entering into any particulars, seeing that the said Beale is not credited nor authorized for this purpose by any letter, memorandum or instruction by the queen of England nor of those of her council so that, as I told him, any treaty between him and me [achieved] in this way would neither be solid nor substantial, but would only serve to unnecessarily uncover my intentions, which seemed to be his principal charge; I was not inclined to give him assurances this would come to any effect, remembering too well the little that had resulted from my conferences with him on his last visit, to believe he was sincerely proceeding as a man of goodwill.

This is where things stand; I will not fail to keep you informed of what will follow hereafter, and nevertheless I thank you for your good and ample advice in your last (letters) which the said Beale brought me, according to which, if things make some progress hereafter, I may well request that I be allowed to send Nau to the queen of England to propose and explain to her on my behalf what I would like to commit only to her, and I would not find it inappropriate for you to make an appeal for it right now, but if it is necessary to enter into a treaty and if there is hope of coming to a conclusion, it will be convenient for me to employ several people of greater quality whom I can send to join him for further instructions on all the necessary matters of which he has long been well informed. I cannot thank you enough for the care, vigilance and entirely good affection with which I see that you embrace everything that concerns me and I beg you to continue to do so more strongly than ever, especially for my said release to which I see the queen of England quite inclined. If it brings any success, be assured that I will acknowledge as much as I can your good services with her, and the obligation I have for you from the past. Sheffield, this sixteenth of April.

The letter marked is for the Sieur de la Tour²⁴⁹ and the other marked for Monsieur de Glasgow, ambassador of His Majesty. Nau humbly kisses the hands of Monsieur de Mauvissiere, asking this bearer, who is in charge of communicating with him on all matters, to let him hear all the news from here.

F54 − 5 May 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 54.²⁵⁰

Contains an enclosure for Henry Howard,²⁵¹ and a cipher key for the Sieur de Mainneville.

We reproduce here the full decryption and a tentative translation.

Monsieur de Mauvissiere, vous m’avez faict tres le plaisir de m’informer amplement, comme vous avez faict par voz deulx dernieres, de voz procedures sur ceste nouvelle ouverture et negociation de ma liberté, de laquelle je ne vous puis mander aulcune particularité solide jusques a ce que je voye quel pied et fondement aura pris le raport que Beale fera de ce qui s’est passé entre luy et moy sur ce subject, m’ayant promis a son partement de m’en donner advis le plustost qu’il pourra et d’y faire sincerement tous bons offices. Or, si d’avanture il se retenoit de m’escripre si ouvertement qu’il seroit de besoing dont mesmes je ne le vouldrois presser, je vous prye d’entrer en conference avec luy et de retirer indirectement, si vous pouvez, l’intention de ceste royne et de ceux de son conseil en cest affaire, car je pense que ledict Beale le vous commectera plus volontiers qu’a sa plume.

J’actenderay encores a requerir le voyage de Nau jusques a ce que les choses, comme je vous mandois dernierement, soient plus advancees, luy mesme m’en ayant suppliee.

Le voyage de mon hoste est a ce que j’entendz retardé ou plustost rompu, mais je ne scauroys imaginer sur quelle occasion si ce n’estoit qu’ilz le voulzissent avoir compris en la commission de ceux qu’ilz deputeront pour traicter avec moy, car jusque icy j’en apercoy poinct que leur intention soit de m’avoir par dela comme il semble que cest espion, dont vous m’escripvez en vostre seconde lettre, vous ayt²⁵² raporté. Mais a la verité, je n’entendz pas bonnement ce qu’il veult dire par ces motz que leur intention est de m’attirer par dela et ne conclure rien par deca. Je vous prye me les esclaircir par les premieres que vous m’escriprez.

Je ne trouve qu’a propos l’intention de monsieur de Manneville, mais prenez garde qu’en intimidant vainement ceste royne et ne succedant²⁵³ pas ma delivrance, vous ne luy faciez prendre, pour s’asseurer, ordre plus exact et rigoureux en ma garde et traictement par deca. Je vous remercye du bon advis que me donnez pour requerir la royne mere, sy son voyage de Callais continue, d’envoyer quelque gentilhomme expres par deca pour m’assister de sa part en ceste demande de ma liberté. Vous ferez s’il vous plaist ce bon office pour moy vers la royne mere, m’excusant si je ne luy en escripz moi mesmes, n’osant par ceste voye rien commectre hors de chiffre, et par l’ordinaire il ne pourroit faillir a estre descouvert.

J’ay escript a la royne d’Angleterre tout de la mesme facon que vous me conseillez, et de bouche j’ay chargé Beale de faire entendre a Burghley, Leicester et Walsingham et aultres du conseil, selon qu’il verra estre a propos, la sincerité de mon intention vers leur dicte royne, eulx mesmes et cest estat, avec asseurance de ma part contre toute innovation a l’advenir si ce traicté prend effect. Vous leur pourrez encores confirmer ; et pour le regard du roy, monsieur mon beau frere, le meilleur office que vous me scauriez faire en son nom seroit, en insistant vifvement pour madicte liberté, de promectre qu’il interviendra a l’accord et traicté qui se passera sur ce entre la royne d’Angleterre, moy et mon filz et le validera par sa propre signature, demeurant comme respondant d’une part et d’aultre pour la perfourmance d’iceluy.

Si le sieur de Manneville passe, comme on dict qu’il pretend, par ce pays, je vous prye de le remercyer de ma part pour ses bons offices en Escosse dont je me sens bien obligee vers luy et retirez soigneusement le plus particulier compte et discours que vous pourrez de ses negociations tant avec mon filz que avec ceux du bon party audict Escosse et en quel estat il y a laissé les affaires. S’il desire m’escripre luy mesme, vous luy delivrerez l’alphabeth de chiffre cy enclos affin que cy apres il s’en puisse aider s’il est d’avantage employé en mes affaires.

N’oubliez ce que cy devant je vous ay mandé pour entrer en pratique avec le colonel Stewart, et a cest effect je pense que quelque offre d’une bonne pension serviroit grandement, s’il plaisoit au roy de France a l’en gratiffier. Ramantevez luy la bonne souvenance et opinion que vous avez entendu qui me demeurent tousjours de luy pour le bon debvoir qu’il feist entre feu Madame de Lenox et moy et que, s’il me veult asseurer de sa fidelité vers moy et mon filz pour en dependre entierement d’icy en avant selon son debvoir, je luy moyenneray plus de bien honneurs et grades qu’il n’en recepvra jamais de ceste royne, l’usant seulement pour le temps.

Tesmoignez a Archibal Duglas le ressentiment qui me demeure de l’indigne et mauvais traictement qu’il a receu a mon occasion, ce que je recepveray pour une entiere preuve de sa fidelité et sincere intention vers moy affin de l’en recompenser si Dieu m’en donne jamais les moyens. Des conditions qui luy ont esté offertes par Valsingham je trouve qu’il feroit mieux d’accepter sa retraicte en Escosse ou il ne laissera d’entretenir telle intelligence qu’il vouldra avec ledict Walsingham pour en tirer ce qu’il pourra et ne sera pas en son danger ny demeurera tant obligé vers ceste royne comme elle pretenderoit luy donnant entretenement par deca, mais qu’il advise bien avant la main de pourvoir a sa seureté en Escosse, ou estant il me feroit un grand plaisir de regaigner s’il pouvoit le comte d’Angus et de bien asseurer mon nepveu de Bothovel.

Donnez de ma part a Foular dix livres Streling et l’asseurez que, continuant a vous inpartir fidellement les advertissements qu’il aura d’Escosse et de servir pour moy ou vous l’enployrez, vous luy moyennerez une honneste recompense laquelle de vray je ne vouldrois luy refuser si j’estois une fois bien asseuree de sa fidelité et mesme pourrois je luy accorder quelque pension annuelle.

J’ay escript a monsieur de Glasgo qu’il retire ceste jeune fille de laird de Granges et l’apoincte avec quelqu’une de mes parentz en France ; cependant, je remercye madame de Mauvissiere de sa demonstration de bonne volonté en cest endroit qui m’oblige tousjours de faire pour tous les vostres ou l’occasion s’en presentera ; sur quoy, me recommandant a vos bonnes graces, je prie Dieu qu’il vous aye, monsieur de Mauvissiere, en sa saincte et digne garde. Sheffeild, le cinquiesme may selon la vieille computation.

Ce qui suit est de Nau a monsieur de Mauvissiere.

Monsieur, je vous baise les mains pour la favorable souvenance qu’il vous plaist avoir de moy qui demeureray tousjours vostre fidelle serviteur. J’envoye a cest honneste homme Fauster le brevet de sa pension ne restant rien sinon que, s’il vous plaist le retenir plus longuement par dela, je vous supplie l’advouer pour vostre domestique affin de l’exempter de la malveillance et recherche d’aulcuns qui le pourroient troubler sans vostre protection.

La lettre marquée est pour le frere de Sa Majesté, celle sans marque est un alphabet pour Monsieur de Manneville.

Translation:

Monsieur de Mauvissière, you have given me great pleasure in informing me amply, as you have done in your last two letters, of your proceedings on this new offer and negotiations for my freedom, on which I cannot give you any solid details until I see the outcome of Beale’s report on his dealings with me on this subject, [Beale] having promised me upon his departure to inform me about it as soon as he can and to sincerely do everything in his power to help. Now, if beforehand he refrains from writing to me as openly as would be necessary on this matter – on which I would not wish to press him, I beg you to confer with him and to indirectly find out, if you can, the intentions of this queen and those of her council on this matter, for I think that the said Beale will commit it to you more readily than to his quill.²⁵⁴

I will still wait before requesting [permission for] Nau’s journey until the things which I wrote to you about recently are further advanced, since he has begged me to do so.

The journey of my host, as I understand it, has been delayed or rather canceled, but I could not imagine any reason for this, other than their wanting to include him in the commission they will send to deal with me, for so far I do not see that their intention is to have me there as it seems that this spy, of whom you wrote to me in your second letter, has reported to you. But in truth, I do not quite understand what he means by these words, that their intention is to pull me there and not to conclude anything here. I beg you to clarify this for me in the first letters you write to me.

I do not find in Monsieur de Mainneville intentions anything but appropriate, but beware that by intimidating this queen in vain and not succeeding in obtaining my release, you might cause her to issue more precise and strict orders for my custody and treatment here. I thank you for the good advice you gave me to request the Queen Mother, if she proceeds with her travel to Calais, to diligently send some gentleman to assist me on her behalf in this request for my freedom. Please kindly apologize on my behalf to the Queen Mother that I don’t write to her directly, as I do not dare to commit anything not in cipher via this channel, and via the ordinary one, my letter would not fail to be discovered.

I have written to the queen of England as you have advised me, and by word of mouth I have instructed Beale, as he sees fit, to convey to Burghley, Leicester and Walsingham and others of the Council the sincerity of my intention toward their said queen, themselves, and this state, with assurances on my part against any innovation in the future if this treaty takes effect. You can also confirm this to them; and as far as the king, my brother-in-law, is concerned, the best thing you could do for me in his name would be to promise that he will take part in the agreement and treaty which will take place on this matter between the queen of England, myself and my son, validating it by his own signature, remaining as the guarantor on either side for the performance of the treaty.

If the Sieur de Mainneville passes through this country, as it is said he would do, I would ask you to thank him on my behalf for his good services in Scotland, for which I feel very much obliged to him, and to make sure to obtain from him a report as detailed as possible on his negotiations with my son and with those of the good party in Scotland, and on the state in which he has left matters there. If he wishes to write to me directly, you will deliver to him the enclosed cipher alphabet so that he can then use it if he is further involved in my affairs.

Do not forget what I have previously sent you about entering into business with Colonel Stewart, and to this end I think that an offer of a good pension would be of great use, if it would please the King of France to grant him one. Bring back to him the good memory and opinion of him, as you have heard, that always remain with me for the good service he did between the late Madame de Lennox and me and that, if he is willing to assure me of his loyalty and entire devotion toward myself and my son, in accordance with his duties, I will give him more honors and grades than he will ever receive from this queen who uses him only for the moment.

Convey to Archibald Douglas the resentment which I still have for the unworthy and bad treatment which he has received [because of his actions] on my behalf, which I will receive as a full proof of his loyalty and sincere intentions toward me, so that I shall reward him if God ever gives me the means for it. About the conditions which have been offered to him by Walsingham I think that he would be better off agreeing to retire in Scotland where he will not fail to maintain such intelligence as he may want with the said Walsingham, to obtain what he can out of it, and he will not be in danger nor remain so obliged toward this queen as she would pretend to support him there, but he should take good precautions beforehand to provide for his safety in Scotland, where it would give me great pleasure if he could win the Earl of Angus back, and [re]assure my nephew of Bothwell.

Give Fowler ten pounds Sterling on my behalf and assure him that if he continues to faithfully impart to you the intelligence he will obtain from Scotland, and to serve me wherever you will employ him, you will give him an honest reward which indeed I would not refuse him once I were well assured of his fidelity, and I could even grant him some annual pension.

I have written to Monsieur de Glasgow to take hold of this girl of the Laird of Grange and appoint her to one of my relatives in France; however, I thank Madame de Mauvissière for demonstrating her goodwill in this matter, for which I will forever be obliged to do the same for all your relatives whenever I will have the occasion; whereupon, commending myself to your good graces, I pray to God that he may keep you, Monsieur de Mauvissière, under his holy and worthy protection. Sheffield, the fifth of May according to the old computation.

The following is from Nau to Monsieur de Mauvissière.

Monsieur, I kiss your hands for the favorable remembrance you have of me, who will always remain your faithful servant. I am sending to this honest man Foster the patent for his pension, having nothing further to write than if it pleases you to retain him longer, I beg you to accept him as your servant, in order to exempt him from malevolence and from being chased by others who would trouble him without your protection.

The letter marked is for Her Majesty’s brother; the unmarked one is an alphabet for Monsieur de Mainneville.

F26 − 31 May 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 26.²⁵⁵

Mary is awaiting the arrival of the queen’s commissioners, Mildmay and Beale. She fears that a separate agreement with James may have been already concluded, in which case little will be achieved during the commissioners’ visit. She had been insisting on their dealing with herself and James together, to preserve his safety, and to secure the support of France. She asks Castelnau whether the Hamiltons’ representative has informed him about the negotiations for their reestablishment, and how the Duke of Lennox’s man²⁵⁶ and Colonel Stewart have been dispatched (to London), as it is said that the latter’s requests and proposals have been accepted. She asks Castelnau to reassure Lord Burghley and Walsingham of her sincere intentions in the negotiations and of her affection for the queen, as her enemies are trying to create a divide between herself and the queen. Castelnau should mention Mary’s good opinion of Lord Burghley, as a wise councilor loyal to his country. Even though Lord Burghley had been against Mary in the past, Castelnau should try to convince him that she holds no grudge against him regarding the death of the Duke of Norfolk. Mary heard that Colonel Norris²⁵⁷ and all English forces in the Low Countries have been decommissioned,²⁵⁸ and that the queen has decided not to further interfere in the Duke of Anjou’s affairs there, being informed that the king of Spain has made extensive war preparations. This is exactly what Mary and others predicted would result from the duke being lured into this country by the Earl of Leicester’s schemes. Mary expects that the queen will accommodate the king of Spain at the expense of France. Castelnau should inform Archibald Douglas that she has written to James in his favor and instruct him that once he is established in Scotland, to keep his promises of loyalty, he should set up a communication channel via Fowler or another courier and provide intelligence from time to time. Castelnau should assure him that the king will grant him a pension. Mary asks Castelnau to continue finding out as much as he can about the actions of her hostess and her children against herself and against her host, and to inform her of anything he may learn.

Postscript from Nau: Nau has presented Mary with the words of praise Castelnau had sent to him. Foster should have received a document via More,²⁵⁹ and Castelnau should assure him that he will be remembered in a beneficial way.

Mary continues the letter: She informs Castelnau that she is writing to her ambassador, asking that her aunt from St. Pierre²⁶⁰ send the young daughter of the Laird of Barnbougle²⁶¹ to Mary, instead of another young woman previously agreed upon. It would be even better if she sent both, along with a third one, Rallay’s niece.²⁶² Since Walsingham had agreed only two, Castelnau should discuss the topic with him, arguing that the daughter of the Laird of Barnbougle is destitute of any support from her father, and that the ladies would replace as many servants so as not to increase the size of Mary’s household.

F113 − 1 June 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 113.²⁶³

Contains an enclosure for Beaton.

Mary has received a letter from Colonel Stewart together with Castelnau’s postscript. She has also received a letter from Howard, but she has no time to answer him at the moment. Castelnau should not trouble himself about a ring missing from a package of hers that he had forwarded to Howard, as the ring is still with her.²⁶⁴ Mildmay and Beale have arrived this evening, handing over letters from the queen, but as they provided no details about the negotiations, Mary cannot report anything yet. She asks Castelnau to pay in cash for the book of flowers mentioned by Beale. Castelnau should thank Archibald Stewart²⁶⁵ for his new offers and promises about the release of her son, and if he can make this happen, he will be rewarded. Castelnau should try to win Colonel Stewart over, as Mary heard he is not pleased with the queen and her council. The king should demonstrate the negative effects of Scotland not having an alliance with France, by banning any commerce between Scotland and France until James is released. But if the king is renouncing his attempts to renew the league of Scotland with France, Mary finds that an initiative to reconcile the Duke of Anjou with the king of Spain would be highly beneficial for the whole of Christendom, and she will do her best to promote this initiative, promising utmost secrecy. For that purpose, she would need to know the terms proposed by the duke to the pope and the Spanish king. She asks Castelnau to let her know whether the Duke of Anjou intends to move forward with this initiative, and whether she is allowed to intervene in its favor. If Castelnau finds it appropriate, Mary will send him letters to be delivered to the Spanish ambassador to England,²⁶⁶ in which she will start proposing the reconciliation in general terms. If Castelnau does not want to handle those letters, Mary could instead venture to send them to the Spanish ambassador via a gentleman appointed by the latter’s predecessor. Mary will make sure to update Castelnau on those negotiations.

F34 − 15 June 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 34.²⁶⁷

Contains an enclosure either for Beaton (most likely) or for the Duke of Guise.²⁶⁸

As Mary expected, there was no resolution as a result of the visit of Mildmay and Beale, which confirms her suspicions that their only purpose was to gain time. They claim the queen cannot make any progress with an agreement unless she first knows the intentions of her son and of those around him. Mildmay has been ordered to return home, and while the Earl of Shrewsbury remains in Sheffield, Beale has been instructed to meet the queen and report to her the details of the discussions. Mary does not regret this state of affairs, as the conditions offered to her were no less hard than those of her current prison, where at least she is free from any obligation toward the queen, and she has more power to manage her affairs than in the state into which they want to reduce her, under the pretext of releasing her from captivity. Mary asks Castelnau to confer with Beale, and to inform her as soon as possible of the intentions of the queen and of her council, and of Beale’s report on the negotiations. To advance the negotiations, Mary suggests that Castelnau instill some fear in Beale of some large enterprise by France or Spain if the negotiated treaty is not concluded, since she had felt some apprehension from their side. Mary is waiting for Castelnau’s answer on what she wrote (in F113) about the reconciliation of the Duke of Anjou with the king of Spain, because she does not want to engage in a matter of such importance without being sure that she will not be disavowed. For that purpose, Castelnau should obtain a declaration from the Duke of Anjou on his intention on the matter, while she pledges again that she would keep the matter secret. Mary heard that Charretier,²⁶⁹ the duke’s secretary, has been captured by de la Motte²⁷⁰ on his way back to Flanders. Mary suggests that letting Charretier pass as a prisoner of the Spaniards would help in handling the reconciliation in a secure and secret manner, as Charretier is familiar with his master’s affairs, and no one else could better handle the matter, Mary being ready to assist as much as she can. Mary apologizes to Howard and Throckmorton for not having the time to write to them at this time, being occupied with the departure of the commissioners.

F46 − 3 July 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 46.

Known plaintext copy: BL Harleian MS 1582/320–321.

Contains an enclosure for Beaton and, also enclosed, a summary of the status of negotiations between Mary and the queen’s commissioners.²⁷¹

The deciphered letter is practically identical to the copy in BL Harleian MS 1582, except for the last paragraph which is missing from the Harleian copy. In this paragraph, Mary mentions the enclosure, and she asks Castelnau not to send the courier back to Sheffield without letters she is expecting, and to inquire with Beale about the status of the negotiations. The deciphered version also includes the date (July 3rd), while the Harleian copy is undated.

In the d’Esneval papers (Ms. 190r-192r), a letter dated 14 July 1583 shares most of its contents with the newly deciphered letter and the plaintext Harleian copy, but also significantly differs from them in other places, as discussed by Bossy.²⁷²

F78 − 17 Aug 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 78.²⁷³

Known plaintext copies: Hatfield Calendar v. 3, 12 (Cecil Papers Volume 162, no. 23–24).²⁷⁴

Contains enclosures for Gray,²⁷⁵ Archibald Douglas, and Throckmorton.

In the Cecil Papers copy, a surname is left blank (“Charles …”) which in our decipherment reads as “Charles Cavendish.”²⁷⁶ Mary’s last sentence mentioning Howard (“mon frere”, to whom she apologizes for not having the time to write) and Nau’s short postscript, in which he thanks Courcelles,²⁷⁷ are also missing from the copy in the Cecil Papers.

F69 − 3 September 1583, Mary to Castelnau, Worksop

Reference: BnF Fr. 2988 f. 69.

Known plaintext copies: Hatfield Calendar v. 13, 234–35 (Cecil Papers Volume 133, no. 31).²⁷⁸

Contains an enclosure for Beaton.

The last sentences of Mary’s letter are not included in the Cecil Papers copy: Mary thanks Castelnau and his wife for assisting (Mary) Seton,²⁷⁹ and she will not forget, when she is able to do so, to compensate him and his relatives, including her goddaughter,²⁸⁰ about whom she hears much of goodness and virtue. Following Castelnau’s advice, Mary will write to the main members of the queen’s council, including the Earl of Leicester. She will also reply to Thomeson,²⁸¹ from whom she had received a letter the day before. Castelnau should convey her regards to Beale, and that Mary remembers his good services.

A postscript by Nau is not included in the Cecil Papers copy: Nau informs Castelnau of a positive report by a gentleman who has accompanied Mary Seton on behalf the Earl of Shrewsbury, the latter praising Castelnau’s favors to him, and Mary is also grateful for that. Nau also informs Courcelles that he has conveyed his useful information to Mary. Mary desires to help (Archibald) Douglas being pardoned.

F30 − 10 October 1583, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 30.²⁸²

Mary has received from Castelnau some news from Scotland which relieves her from the anxiety she felt since Walsingham went there, having feared that he would stir up more trouble. Those fears did not materialize, and James is now protected from his enemies in Scotland and England. Mary would like Castelnau to plead for the help of the king and of the Queen Mother, pledging to preserve the old alliance between France and Scotland, and thus ensuring her son’s devotion to France. If current necessities prevent such assistance, her Guise cousins should be allowed to assist her, and Mary is willing to allocate up to 2,000 pounds of her dowry money for that purpose. Castelnau should also oppose any attempts by the queen to send armed forces to Scotland, as rumors claim she is planning to do so, possibly driven by the apprehension of a Spanish intervention, in which case, the best course of action for the king would be to ensure the loyalty of Scotland and of James. This has been Mary’s constant wish, but if there is some extreme necessity, or if she is abandoned by her friends, she would feel free from any obligation.²⁸³ Some Scottish noblemen among her most loyal partisans have cooled down toward France in recent years, because their devotion to France has not been rewarded, and nothing came from the nice words and promises of the previous ambassador, de la Mothe-Fénelon. Those noblemen as well as James may fall for overtures from Spain, as being deprived of the support of France, they might look for it elsewhere, which Mary would prefer rather than seeing them succumb to their enemies. It is still in the king’s power to keep James close to France. As for Walsingham’s claims that there is intelligence between her and James, she can no longer say there is none, but when Castelnau is confronted by Walsingham or others of the queen’s council about it, he should mention that those near James who are loyal to Mary are advising him without Mary being involved, relying on their own experience with similar past incidents, with the sole purpose of ensuring James’s safety. Castelnau should thank Archibald Douglas for the information he has provided and for his pledge to continue to do so, asking him to probe Walsingham about the results of his recent visit. Mary would be ready to help him financially, but the state of her affairs may prevent her from doing so currently. Castelnau is to inquire of John Hamilton on what has been going on between him, his brother, and the queen, mentioning that Mary heard negative things and that they had better resolve this if they expect any favor from Mary. Mary is feeling much better now, compared to recent years, but does not want to reopen the negotiations with the queen. Postscript by Nau: 100 ecus should be paid to Mosman, 50 ecus to Fowler, three angelots to More’s man. Mary needs an additional 200 ecus from Castelnau.

F58 − 25 February 1584, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 58.

Known plaintext copies: BL Harleian MS 1582/311–313.²⁸⁴

At the end of the deciphered letter, an acknowledgment of receipt of Castelnau’s letters and a short postscript from Nau, asking Courcelles to continue to update Mary on relevant developments, are both missing from the BL Harleian MS 1582 copy, in which, there is instead a famous postscript written to Walsingham by his agent in the French embassy.²⁸⁵

F96 − 30 April 1584, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 96.

Known plaintext copies: BL Harleian MS 1582/321bis–323.²⁸⁶

F42 − 23 May 1584, Mary to Castelnau, Sheffield

Reference: BnF Fr. 2988 f. 42.

Known plaintext copies: BL Harleian MS 1582/315–317.²⁸⁷

F221 − 31 May 1584, Mary to Castelnau, Sheffield

Reference: BnF Fr. 20506 f. 221.

Known plaintext copies: BL Harleian MS 1582/404.²⁸⁸

F57 − 30 Oct 1584, Mary to Castelnau, Wingfield

Reference: BnF Fr. 3158 f. 57.²⁸⁹

This letter is written mostly in cleartext with some ciphered passages, the meaning of which was unknown prior to our present work. We reproduce and translate the enciphered parts, and the surrounding cleartext parts.²⁹⁰ The boldface used below represents portions originally in cipher.

……

Le principal de quoy j’ay maintenant à vous escripre est: 1° mon changement hors d’icy, qu’on m’a rapporté estre résolu, soit entre les mains de ce bon homme sir Ralf Sadler ou du comte de Rutland ou de mylord de Sainct Jon.

……

Cependant, poursuivez, je vous prie, avec toute la diligence qu’il vous sera possible, le voyage de Nau par delà, et parlez en de ma part particulièrement à monsieur le comte de Leicester, luy ramantevant la promesse qu’il m’ha faicte de le mener a effect, ayant mesmes a luy mander de bouche par ledict Nau ce que je ne puis et n’ose commectre au papier. Je vous recommande ledict Gray et toutes ses negociations par delà, à ce qu’où il aura besoing de vostre ayde et assistance et de l’intercession du Roy, monsieur mon bon frère, vous l’en subveniez ;

……

Ce qu’advenant, n’oubliez de bien informer le dit Courcelles de toutes vos conceptions, bons avis et conseil pour mon estat et affaire par delà. Donnez vous de garde de Archibal Duglas, car il n’a procedé en tout si sincerement comme il vous a voulu faire croire, et surtout qu’il ne scache rien par vous des negociations dudict Gray.

……

Et estant tout ce que je vous puis mander à présent, me trouvant pressée du partement de ceste dépesche affin qu’elle puisse trouver Gray en temps par dela, je finiray par mes bien affectionnées recommandations ; priant Dieu qu’il vous aye, monsieur de Mauvissière,…

Translation:

……

The main topics I am writing to you about:

1° My transfer from here,²⁹¹ which I was told was firmly decided, either to Sir Ralph Sadler, a good man, or to the Earl of Rutland,²⁹² or to Lord Saint Jon.²⁹³

……

However, please continue to pursue as soon as possible [an authorization for] Nau’s travel over there,²⁹⁴ and in particular, raise the topic on my behalf with the Earl of Leicester, reminding him of the promise he made to make this happen, and even have Nau tell him verbally what I cannot put in writing. I recommend assisting the said Gray for his negotiations there, in case he needs your assistance or the assistance of the king.

……

Do not forget to inform the said Courcelles of your ideas and good advice about my estate and affairs there. Beware of Archibald Douglas, for he has not been as sincere as he would like you to believe, and in particular, make sure he doesn’t learn anything from you about Gray’s negotiations.

……

This is all I can report on now, being under pressure to dispatch this letter, so that it can reach Gray there in time, and I will conclude with my affectionate recommendations etc.

F308 – Date and place unknown, Mary to Castelnau

Reference: BnF 500 de Colbert 470 f. 308.²⁹⁵

This deciphered letter is reproduced here in full, with a tentative translation.

Faictes toutes les demonstrations d’amitié que vous pourrez a ce porteur et luy tenez le plus honorable langage de la comptesse sa maistresse, Madame de Shereusbury, luy discourant, comme vous scavez bien faire, la fiance que vous scavez que j’ay en sa bonne volonté et amitié et que reciproquement en elle doibt continuer les bons offices que aultrefoys elle m’a faictz dont vous avez esté adverty lorsqu’elle estoit en court. Vous luy remonstrerez aussi comme a mon occasion seule vous congnoissez plusieurs seigneurs en ce royaulme qui sont affectionnez a ladicte comtesse, lesquelz ont extreme regret de la division qui estoit entre elle et moy, et pour conclusion, que pour son bien propre elle me doibt aymer et conserver son debvoir reservé plus que personne d’Angleterre. Je ne vous ramantevray point de luy declarer la meilleure volonté du roy, monsieur mon beau frere, vers moy et l’estat qu’il faict de mon droict par deca, m’asseurant que vous n’obmectrez rien qui puisse servir a ce propos ; et gardez vous soigneusement de passer oultre, ny vous fyer en facon que ce soit a ce porteur d’aucune chose d’importance, mesmement que vous ne veuilliez estre sceue par le comte de Leicester, soit de mes affaires ou du roy, monsieur mon beau frere. Trouvez moyen de communiquer avec mylord Burghley et me mandez son advis je vous ay escript nagueres et pense que dans jeudy au plus tard vous recevrez la depesche.

Translation:

You should express your friendship in every possible way toward this messenger, and speak highly of his mistress, the Countess of Shrewsbury, mentioning my confidence in her friendship and goodwill, and that in return, she should continue with the good services she did for me in the past, of which you have learned when she was at the court. Also mention that you know of several noblemen in this country who are devoted to the countess but are deeply sorry that there is a divide between me and her, and in conclusion, for her own good, the Countess should appreciate and preserve the duty assigned to her,²⁹⁶ more than anyone else in this country. Being sure that you will not omit anything useful, I do not need to remind you to let him know of the goodwill toward me of the king, my brother-in-law, who also puts forward my rights in this country.²⁹⁷ Make sure not to reveal anything of importance to this messenger, regarding my affairs or those of the king, as any such would be passed on. Please find some way to communicate with Lord Burghley, and let me know of his opinion on what I wrote to you in a letter you should receive by Thursday.

6. Mary Stuart’s secret communications

The existence of a secret communication channel between Mary and Castelnau has been well-known to historians, and it was even known to the English government at the time.²⁹⁸ Even though its existence was known, Bossy states that for some time before mid-1583, the channel was so secure that its contents have been lost.²⁹⁹ With our new decipherments, we provide evidence that such a secret channel was already in place as early as May 1578. Also, while some details were already known, our new decipherments provide further insights into how this channel was operated, and on the people involved.

6.1. Official and secret channels

During her captivity, Mary was allowed to write and receive letters via the official post under Walsingham’s supervision, which Mary referred to as “la voye ordinaire”, “la voye de Valsingham” or the like,³⁰⁰ but sometimes the delivery of letters was significantly delayed (F82, F125, F247, F64),³⁰¹ and there were times when she was not allowed to send letters.³⁰² This official channel was used, apart from communications to Queen Elizabeth or council members, for less sensitive matters like her properties in France.³⁰³ In several letters we deciphered, Mary writes explicitly that there is usually nothing in her official letters that she would need to keep secret even from her enemies (F74, F231). At one time, Mary said Archibald Douglas might write unimportant matters to her via this channel.³⁰⁴

In addition to official channels, Mary managed to keep up secret correspondence via confidential channels, mentioned in her letters as “la voye secrette” (the secret channel) or, simply “ceste voye/commodité” (this channel/commodity). A letter can be considered to have been confidential, and thus sent in cipher, if it mentions such “voye” or refers to another secret letter (or, needless to say, some other sensitive contents), even if only a plaintext copy of this letter has been preserved in archives.³⁰⁵ On the other hand, it is more difficult to tell with certainty that a particular letter preserved in clear in archives was not originally sent in cipher.³⁰⁶ The letters we have deciphered include further direct references to such channels, which clearly indicate that confidential channels were operated in parallel with the official one (F153) and Mary used the official channel for inconsequential matters (F165), sometimes sending another letter – encrypted – on the same day (F174). When de la Mothe-Fénelon was sent to Scotland in late 1582, Mary not only sent him secret letters via Castelnau, but she also wrote via the official channel to avoid any suspicion (F123).

Secret channels were not immune from interception or perhaps the casual prying eyes of the bearer. Thus, when Mary heard that Castelnau would write to Catherine de’ Medici to request her assistance, Mary asked him to also convey her regards, apologizing for not writing herself,³⁰⁷ because she did not dare write anything not in cipher (to Catherine de’ Medici) to be delivered by the current messenger (“par ceste voye”) who was also carrying the present enciphered letter to Castelnau (F54). Nau penned his frustration on the same subject in a postscript, saying that he wanted to reply to a certain Foster, but did not dare write anything not in cipher via a secret channel (F89).

Our decipherments revealed one incident in which Castelnau was careless enough to forward a letter in cipher from Beaton in a packet sent via the official channel. Fortunately, Shrewsbury, who opened the packet, did not take the trouble of inspecting it, and presumably, did not notice the letter in cipher. Mary told Castelnau to be more careful (F50).

Castelnau is known to have forwarded confidential letters between Mary and her associates in Continental Europe such as Beaton in Paris. A diplomatic bag was the safest means to avoid their interception at the ports.³⁰⁸

Bernardino de Mendoza, Spanish ambassador in London, is also known to have forwarded at least one letter from Mary to Beaton.³⁰⁹ As to delivering her letters to the Spanish ambassador, in June 1583, Mary mentions a hazardous channel via a gentleman appointed by his predecessor³¹⁰ (F113).

6.2. Delivery and interception of letters

Mary had to face multiple challenges in order to maintain a secure and secret communication channel with the outside world, including with Castelnau, even though Shrewsbury’s vigilance was sometimes lax.³¹¹ Mary once wrote that sending someone bringing her letters directly could be hazardous, and a channel via London was safer because, once a letter reached London, she had friends whom she trusted to safely deliver it.³¹²

Mary’s secret letters were often intercepted, at times on their way out of her residence.³¹³ Sometimes, letters she sent could not be securely delivered, and they were returned to her. One of Mary’s letters went twice as far as London but could not be delivered and was returned, nearly two months after its sending. Upon receiving it back, Mary wished to update it by writing on the latest topics, but satisfied herself by simply resending it because the bearer could not wait (F82). When a messenger felt an imminent danger, he might burn the packet, as in the case of one entrusted to a certain Jacson by John Hamilton for delivery to Morgan, at the time in London.³¹⁴

Incoming letters were also detained. Sometimes, letters were delivered after being read, but letters in cipher were confiscated.³¹⁵ When secure delivery means were not available, packets for Mary would accumulate at the French embassy.³¹⁶

On occasion, Mary was allowed to receive visits from her associates, who may bring letters to her, or carry her letters to other destinations. For example, when Andrew Beaton, her maître d’hôtel and a brother of her ambassador James Beaton, was sent to France, he carried with him secret letters for her relations.³¹⁷ Certainly, on such occasions, utmost precaution was necessary to avoid being discovered.³¹⁸ When she received a visit from du Ruisseau, a member of her Council in France, he managed to bring in letters but he was arrested by Shrewsbury on his way back and Mary’s letters were confiscated.³¹⁹ A letter we deciphered includes Mary’s complaints about du Ruisseau’s arrest (F118).

But visitors were not always allowed. In 1581, Mary’s request to send for physicians and surgeons was frowned upon because it was suspected that she was seeking to contact, via those physicians, the Duke of Anjou, who was expected to pay the queen a visit in London.³²⁰ Even if visitors were allowed, she might have been closely watched during the meeting, but somehow letters could still be sent out.³²¹

In 1585, she was under the more stringent custody of Sir Amias Paulet, who, informed that coachmen and laundresses might be carrying letters, took strict measures to prevent this.³²²

6.3. Dissimulation

Mary took great care to ensure that her secret communication channel with Castelnau would not be discovered.

In 1583, Mary wanted to deliver a cipher table to Archibald Douglas via Castelnau, but in order to conceal the existence of her secret channel with Castelnau, she asked him to delay its delivery to Douglas until the women (probably her servants) who were visiting the embassy had left London, so that Douglas would think those women had brought the cipher table from Mary to Castelnau, but he would not be able to verify this with them. She also asked Castelnau to make Douglas believe that what Douglas would write in the cipher would be delivered by ordinary packets.³²³

The letters we deciphered contain additional references to such precautions.

When du Ruisseau was arrested in Sheffield after visiting Mary, she asked Castelnau to protest to the queen and Walsingham. But she apparently assumed that the arrest had been kept secret, for she instructed Castelnau to claim that he had not heard from du Ruisseau for two months, although the latter had told him he intended to stay in Sheffield for only three weeks, and that consequently, he concluded that du Ruisseau was being held against his will (F118, 24 September 1582).

At one time, after du Ruisseau was released, Mary asked Castelnau not to forward an enclosed letter for Beaton before he had news that du Ruisseau had reached Calais, obviously, in order to make it seem as if the letter was carried by du Ruisseau (F245, 13 October 1582). At another time, Mary asked Castelnau to convey her words of thanks to Beale, instructing Castelnau that in order to conceal their confidential channel, he could explain to Beale that Mary’s words were conveyed to him by du Ruisseau (F153, 15 January 1583). There were other occasions when Mary asked Castelnau to provide Queen Elizabeth or Lord Burghley with some information but without letting them know it came from her (F21, F64, F89).

6.4. Secret messengers

Mary’s confidential channels of communication were operated by several people over her years in captivity.

A certain Singleton was in principal charge of Mary’s secret correspondence for seven or eight years, but in September 1580, he had to leave England in view of tightened control over Catholics in England. This Singleton appears to be identical with Robertson mentioned in several letters we deciphered.³²⁴

Robertson appears to have had troubles in delivering letters several months before Mary wrote in May 1578 of her concern of not having received anything from Castelnau or from France for the last seven or eight months. To show Robertson that she would not fail to compensate him, she sent him a chain of jewels (F87). Robertson’s troubles, and thus the interruption in Mary’s communications, might have something to do with the (alleged) discovery of some of her servants in England, as reported by Thomas Morgan on 25 December 1577.³²⁵ When Mary wrote to Castelnau that she had to find alternative means to carry letters because the person she had employed was not available (F241), she may have referred to Robertson.

There were continuing efforts to recruit new bearers. Precautions were taken to try a new bearer by initially entrusting him to carry non-sensitive messages.³²⁶ In October 1579 (F249), Nau sent a note to Arnault via “a man of law” (F179), in which he asked his colleague to write to him on some unimportant matters in order to ascertain whether the bearer was reliable. The following month, Mary may have referred to this bearer when she wrote about “ceste voye que je n’ay encores bien experimentée” (this channel which I have not yet checked well-enough), refraining from mentioning sensitive matters even in her enciphered letter (F233). Sometimes, Mary had to rely on someone she could not fully trust to carry a secret letter, such as a servant of the Countess of Shrewsbury, asking Castelnau not to reveal anything of importance to that person, who would report everything he may hear to Leicester (F308). In January 1580 (F181), Mary wrote to Castelnau that she was trying a “voye nouvelle” (new channel). In 1581, Godfrey Foljambe³²⁷ was mentioned as being always able to provide a courier service “en default de ceste voye” (when this channel is not available) (F82, F235). In April 1582, Mary noticed the name Bezet in a letter from Castelnau and asked him to try to set up a channel via him to deliver her letters to Scotland, but only if he was the Patrik Bezet she knew (F165).

According to Bossy, Claude de Courcelles, Castelnau’s secretary, carried packets and controlled a secret communication channel from approximately the end of 1581. He met with Mary’s people, had contacts in London, and worked closely with Mary’s party in Paris.³²⁸

In mid-1583, Walsingham was finally able to infiltrate the secret communication channel between Mary and Castelnau by recruiting a spy in the French embassy. It was long considered that the spy who supplied Walsingham with copies of letters between Mary and Castelnau was Chérelles, based on a postscript added to one of the leaked letters. Bossy reached a different conclusion, and suggested that Laurent Feron, a clerk in the embassy, was the mole, rather than Chérelles, whom he identified to be the same person as Jean Arnault.³²⁹

One of Courcelles’ contacts in London was Sir Francis Throckmorton,³³⁰ who was originally cultivated by Singleton in May 1581 and got acquainted with Mary’s agents such as Thomas Morgan and Charles Paget in Paris.³³¹ In a letter from April 1582 (F125), following several of her packets being delayed, Mary introduced Throckmorton to Castelnau. Throckmorton became intimate with the ambassador and carried letters, but he was arrested in November 1583 at the discovery of the Throckmorton Plot.

Probably captured on this occasion was a letter from Castelnau to Mary dated 5 November 1583, which is extant only in a copy in the hand of Phelippes. As it is likely that Castelnau used the same cipher in his letters to Mary as the one in Mary’s letters to him, this suggests that the Mary-Castelnau cipher might have been broken by Phelippes or leaked by Feron.³³²

After the discovery of the Throckmorton Plot, Mary, deploring “the discovery of all my contacts who have frequented your house,” suspected that there was a spy in Castelnau’s embassy and asked him to make sure that her messengers are received by his servants of assured loyalty at some place outside of the embassy. In reply, Castelnau assured her that in the embassy, only three people knew of their secret correspondence. Unfortunately, the three included Feron, whom Castelnau did not suspect of being behind the leakage because he “never leaves my chamber and writes everything in front of me and in my presence.” The other two were Courcelles and Arnault.³³³ Back in January 1583, Mary gave similar advice to Castelnau (F247). That is, she advised him that the couriers must not know each other; that especially nobody except Castelnau should be allowed to know who the current messenger is; and that those in his office who already knew should be bound to absolute secrecy.

Another bearer, Thomas Baldwin, was the Earl of Shrewsbury’s London agent and ran a carriage service between London and Sheffield. He had been recruited by Courcelles and delivered Mary’s letters as early as October 1581 (F194). In February 1583, he was to bring 2,000 ecus to Sheffield (F174). Even after Throckmorton’s arrest, he managed to carry Mary’s confidential letters until his own arrest in October 1584.³³⁴

The discovery of the Throckmorton Plot in November 1583 put Castelnau in a delicate position. He had to promise Walsingham that he would show him all the letters to and from Mary, though, just before making the promise, he sent Mary a packet containing the letters for her that had piled up in the embassy. In the meantime, the ambassador’s extensive involvement with Mary was questioned by the French court.³³⁵ Castelnau was recalled in November 1584, and left England the following year.

With Castelnau’s replacement, Guillaume de L’Aubespine, baron de Châteauneuf, it appears Mary no longer communicated via the official channel under Walsingham’s supervision.³³⁶ Some letters between Mary and Châteauneuf are held in British archives, with some in the hand of Phelippes, suggesting that their confidential letters were intercepted and deciphered by the English authorities. There is even a letter from Walsingham to Phelippes, attached to intercepted letters from d’Esneval and Châteauneuf, in which Walsingham asks to know their contents.³³⁷ These letters are all from the period after Mary was made to believe a safe channel was established by Gifford, who was in fact Walsingham’s double-agent (see Section 2).

6.5. Codenames and aliases

In Mary’s secret correspondence with Castelnau, not only was the text enciphered, but also some codenames or aliases were used for additional security. For example, “M. de la Tour” or “Sieur de la Tour” referred to Francis Throckmorton. The identification was so obvious for historians that Bossy even assumes it was known to Walsingham.³³⁸ A letter from 25 February 1584 even provides a direct link: it refers to “du Sieur de la Tour, du Comte de Northumberland et de mylord Henry Hovard” and a few lines below “ledict Throkmorton et Hovard”, linking “de la Tour” with Throckmorton.³³⁹ In one of the letters we deciphered, Mary explicitly introduces the codename: “ce gentilhome qui s’apellera entre nous ‘la Tour’.” (this gentleman, who will be called between us “la Tour”) (F125, 27 April 1582).

Henry Howard, mentioned with Throckmorton in the quote above, was also given an alias. He was called “mon frere” (my brother) by Mary, “vostre frere” (your brother) by Castelnau,³⁴⁰ and “le frere de la royne d’Escosse” (the brother of the Queen of Scots) (F74) or “le frere de Sa Majesté” (the brother of her Majesty) (F225) by Nau. According to Bossy, the alias refers to the fact that Mary had intended to marry his brother, the Duke of Norfolk.³⁴¹ Mary also called Howard’s nephews, that is, Norfolk’s sons, “her children”.³⁴² The earliest use we found of this alias is in F74 (October 1582),³⁴³ in which Nau proposed a new cipher symbol for Howard, but this new symbol is never used in the subsequent ciphertexts we examined. Howard is explicitly spelled in F158 (July 1582), F245 (October 1582), and F58 (February 1584). The appellation “frere” may have been introduced in October 1582 and fallen into disuse after Howard’s arrest in late 1583.

Additional aliases are known: According to Bossy, “Monsieur de la Rue” was an alias for Mary’s clandestine Jesuit servant Henri de Samerie and “Banque” may have been Baldwin.³⁴⁴ Robertson, mentioned in Section 6.4, may also have been a codename for Singleton, or vice versa.

7. Conclusion

The fact that the letters were fully in cipher, stored in archive collections with unrelated material and incomplete catalog information, may explain why those ciphertext documents had not been previously attributed to Mary Stuart.³⁴⁵ Therefore, the present work would not have been possible, without prior systematic efforts to map, digitize, and decipher encrypted letters held in archives, such as the DECRYPT project and the Cryptiana website.³⁴⁶ It is possible that an additional effort to examine other collections in BnF and in other archives may unearth additional ciphertext documents from Mary Stuart, taking into account that we only had access to the BnF collections that are available online, and that there is evidence that some enciphered letters known to have existed are still missing.³⁴⁷ Furthermore, inspecting the physical documents in the collections covered by the present work may help fill some gaps in deciphered texts, caused by the low quality of the scans or of the documents, also enabling the examination of physical properties such as ink, paper, and handwriting. In addition, it would also be interesting to understand why almost all the ciphertext documents ended up in mostly unrelated BnF collections.

In two of his books,³⁴⁸ Bossy provides a detailed account on how Walsingham and his agents infiltrated Castelnau’s embassy in mid-1583, and compromised Mary’s secret communication channel with Castelnau.³⁴⁹ Our present work confirms Bossy’s hypothesis about the secret channel being in place for a while before mid-1583, showing that it was established already in 1578, and that it was most active in 1582 and 1583. The newly deciphered letters also highlight extensive efforts by Mary and her associates to recruit messengers and ensure communication secrecy and security. While we were able to observe some variety in handwriting styles and choices of specific homophones during encryption, we could not establish who actually enciphered Mary’s letters. Further research, including comparative handwriting analysis, combined with tools like statistical analysis, might allow for the identification of the various “hands” involved and possibly the specific persons who did the enciphering work for Mary.³⁵⁰ Another direction for further research is around the role of historical codebreaking with regards to the Mary-Castelnau correspondence. While at least one letter in cipher seems to have been intercepted and deciphered in late 1583 by Phelippes,³⁵¹ we did not find the key for the Mary-Castelnau cipher among the dozens of cipher keys related to Mary held in British archives, nor did we find any document encrypted with this cipher in those archives.

While in this paper we dealt in more detail with Mary’s communication channels, there are plenty of other topics for which new insights may be obtained upon further study of the deciphered documents, such as specific events or people.³⁵² Another possible direction for future research could be a systematic comparison of the writing style, structure, and subjects of the enciphered letters, with the non-enciphered letters that were mostly sent via official channels.

Due to the sheer amount of deciphered material, about 50,000 words in total and enough to fill a book, we have only provided preliminary summaries of the letters, as well as the full reproduction of a few of them, hoping to provide enough incentive to historians with the relevant expertise to engage in in-depth analysis of their contents, to extract insights that would enrich our perspective on Mary’s captivity during the years 1578-1584.

Scholars interested in accessing our deciphered material are welcome to contact the authors. The authors are also interested in partnering with scholars to produce an annotated edition of all the newly deciphered letters.

For primary sources, see Section 3

Abbreviations

• CSP Vatican refers to Rigg (1926).

• CSP Spain (Simancas) refers to Hume (1892-1899).

• CSP Scotland refers to Boyd (1903-1915) or Boyd and Meikle (1936).

• CSP Foreign: Elizabeth refers to Butler and Crawford Lomas (1913).

• Hatfield Calendar refers to Historical Manuscripts Commission (1899) or Salisbury (1915).

Acknowledgments

We would like to thank the DECRYPT Project team for their support, and in particular, Crina Tudor, Nikolina Milioni, and Beáta Megyesi from the University of Uppsala for transcribing parts of the material. We would also like to thank Katy Gibbons from the University of Portsmouth and Camille Desenclos from the University of Picardie Jules Verne for their kind assistance and advice.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work of one of the authors has been supported by the Swedish Research Council, grant 2018-06074, DECRYPT – Decryption of Historical Manuscripts.

Notes on contributors

George Lasry

George Lasry is a computer scientist in the high-tech industry and a member of the DECRYPT Project. He obtained his Ph.D. with the Applied Information Security research group at the University of Kassel. He specializes in cracking historical ciphers. In 2013, he solved the Double Transposition cipher challenge. He also deciphered German ADFGVX ciphertexts and diplomatic codes from World War I; World War II messages encrypted using the German Siemens and Halske T52 teleprinter encryption device; papal ciphers from the 16th, 17th, and 18th centuries; French homophonic ciphers from the 16th and 17th centuries; and transposition ciphers from the Biafran War in the late 1960s.

email: [email protected]

Norbert Biermann

Norbert Biermann is a pianist, musical coach and arranger, and Professor of Repertoire at the Berlin University of Arts. Cryptology is his hobby, with a particular interest in historical ciphers. In 2018, he solved three cipher problems by G. B. Bellaso, a 16th-century Italian cryptologist, and in 2021 published a working paper on letters in cipher sent from the French court to the Count of Lauzun in Ireland in 1690.

email: [email protected]

Satoshi Tomokiyo

Satoshi Tomokiyo majored in astrophysics in the University of Tokyo. Master of Science. Member of the Physical Society of Japan. He currently works for a patent firm in Tokyo. He devotes his private hours to research of historical cryptography. He maintains Cryptiana, a website on historical cryptography, since 2008 (http://cryptiana.web.fc2.com/code/crypto.htm).

email: [email protected]

Notes

1 Megyesi et al. (2020) and Tomokiyo (2019–2022). Those projects also cover ciphertexts that are already attributed.

2 Except one document which has parts in clear and part in cipher, which we found at a late stage in the project – see Section 5.1.

3 Sir Francis Walsingham (c. 1532–1590).

4 See Section 6.4.

5 “Item in cipher”, “enciphered messages”.

6 The cipher is somewhat complex, but a 16th-century trained codebreaker would have certainly been able to break it with a moderate effort. This is also the case for a motivated modern scholar, but such motivation could only have arisen from a prior understanding that those letters are from Mary Stuart, which was not possible until they were deciphered.

7 Francis II (1544–1560), King of France (1559–1560).

8 Charles IX (1550–1574), King of France (1560–1574); Henri III (1551–1589), King of France (1575–1589).

9 Francis, Duke of Anjou and Alençon (1555–1584), the youngest son of King Henry II of France.

10 Catherine de’ Medici (1519–1589), Queen of France (1547–1559), and mother of the kings Francis II, Charles IX, Henry III, and the Duke of Anjou.

11 Charles, Cardinal of Lorraine and Archbishop of Reims (c. 1525–1574).

12 Henry I of Lorraine, Duke of Guise (1550–1588). He and his brothers Louis II of Lorraine, Cardinal of Guise (1555–1588), and Charles II de Lorraine, Duke of Mayenne (1554–1611), were the most prominent members of the Guise family (the “House of Guise”) at the time of our deciphered letters.

13 James Hepburn, 4th Earl of Bothwell (c. 1534–1578).

14 George Talbot, 6th Earl of Shrewsbury (c. 1522/8–1590).

15 James Beaton (1517–1603), archbishop of Glasgow.

16 See Section 6.

17 Thomas Howard, 4th Duke of Norfolk (1536–1572), 3rd Earl of Surrey.

18 Esmé Stewart, 6th Seigneur d’Aubigny (c. 1542–1583).

19 Robert Beale (1541–1601). Brother-in-law to Walsingham.

20 Sir Walter Mildmay (before 1523–1589), Chancellor of the Exchequer to Queen Elizabeth, and brother-in-law to Walsingham.

21 Bertrand de Salignac de La Mothe-Fénelon (1523–1599), had been French ambassador to England before Castelnau.

22 François de Roncherolles, Seigneur de Mainneville (1551–1589). Variously spelled: Maineville, Maigneville, Meyneville, Manneville.

23 Sir Francis Throckmorton (1554–1584).

24 Shrewsbury’s wife was Bess of Hardwick, Countess of Shrewsbury (c. 1527–1608).

25 Sir Ralph Sadler (1507–1587).

26 Sir Amias Paulet (1532–1588).

27 Gilbert Gifford (c.1560–1590).

28 Thomas Morgan (1546–1606), secretary to the Earl of Shrewsbury, had been won over to her cause by Mary, emigrated to France and worked in Paris for her in “the position of a virtual secretary of state” (Bossy 2001, 15).

29 Alford (2012, 196).

30 Thomas Phelippes (1556–1625). A gifted codebreaker as Phelippes was, he was sometimes helped in his work by the capture of cipher keys (Pollen 1922, lv; Guy 2004, 480; Fraser [1969] 2015, 691).

31 There is some confusion about his first name. We follow Strickland (1854, v. 6, 134 fn 1), who states that two brothers served Mary, the elder of whom, named Claude Nau, was her secretary when she came to England and authored a biography of Mary (Stevenson 1883), while from 1575, her secretary was Jacques Nau, his first name explicitly recorded in the official examination record at the time of the Babington Plot (Pollen 1922, 140-144; Kahn 1967, 122). Stevenson (1883), however, also assigns Claude Nau to this time period, and so do various modern sources such as DNB (1894) as well as ODNB (2004), s.v. Nau.

32 Pollen (1922, cxxix-cl).

33 Pollen (1922, 45). This famous postscript was mentioned in William Camden’s Annales published during James I’s reign. The 19th century historian Patrick Fraser Tytler found its draft in the archives (Tytler 1887, 127; Pollen 1922, 45; Read 1909, 132; now TNA SP53/18/55).

34 According to Bossy (2001, 67, fn10), copies of letters of Castelnau and others in the d’Esneval papers were made in France after Castelnau’s return in October 1585 and before d’Esneval’s departure in December as ambassador to Scotland.

35 Dictionnaire des fonds spéciaux et des principales collections et provenances, URL: http://comitehistoire.bnf.fr/dictionnaire-fonds-sp%C3%A9ciaux-principales-collections-provenances.

36 Catalogue général des manuscrits français (1868) t.1.

37 BnF fr. 2988 (f.2, f.9) and fr. 20506 (f.136, which is a copy of f.9 of fr. 2988) both include letters entirely in cipher (apparently Venetian) with the signature of Hieronimo Ranzo, who was active around the 1520s. Both also include English letters in cipher, which appear to address Mary (f.1 in fr. 2988 and f.146 in fr. 20506). Curiously, these are filed some sheets before Mary’s letters in both volumes. Hopefully, these may help determining how Mary’s letters ended up in volumes of apparently unrelated papers.

38 Those passages in cipher could not be deciphered before our work, and are marked as ellipses in Labanoff (1844, v. 6, 45).

39 Therefore, for example, a folio numbered 123 (f.123) would only be relevant in BnF fr. 2988, and f.123 would either not exist or be irrelevant in BnF fr. 20506.

40 Kahn (1967, 996), an authoritative work in the field of cryptography, cites Schooling. 1896. Secrets in Cipher. Pall Mall Magazine, VIII, to give examples of cipher keys used by Mary. The keys reproduced in Schooling are taken from these volumes.

41 Somer’s deciphering sheets are found in Cotton MS, Caligula C III and probably in TNA SP53/11/35.

42 Richards (1974, no.18).

43 Richards (1974, nos.11-17).

44 Optical Character Recognition (OCR) works primarily on documents with printed letters of the alphabets and other standard symbols, rather than on documents with handwritten graphical symbols.

45 CrypTool 2 – E-learning tools for cryptography (Kopal 2018).

46 A cipher with a list of symbols and their meaning, often on a single sheet of paper, is referred to as a nomenclator by some authors.

47 For more details on the evolution of historical ciphers, see Kahn (1967).

48 While we identified early on that dots on the side or on the bottom of symbols should be considered as diacritics and not as stand-alone symbols, at first, we wrongly identified the / and the ) signs as stand-alone symbols. After progressing with codebreaking and decipherment and comparing with other ciphers used by Mary and Castelnau, we recognized them as diacritics as well and corrected our transcription database accordingly.

49 Next/upcoming, brother-in-law, to irritate, to persist, less, time, to see oneself.

50 Noise/rumor, to put.

51 The actual decipherment process involved different plausible plaintext sequences. Those shown here were selected for clarity and illustrative purposes. Furthermore, the process involved multiple hypotheses and iterative trial-and-error testing.

52 The key initially recovered was slightly different, with some errors corrected later.

53 We also employed contextual analysis, after we identified the origin and the subjects of the letters.

54 Those plaintext copies were also useful in providing interpretations or confirming hypotheses about other symbols, such as those representing common words.

55 Du Ruisseau, Nau’s brother-in-law, was, according to CSP Scotland, “intendent of the [Scottish] queen’s affairs in Vermandois” (v. 5, 40–41 no. 46). He visited Mary in 1582 and was then allowed to speak to the queen on Mary’s behalf (Leader 1880, 509–12).

56 CSP Scotland v. 6, 194, no. 199; cf. also Leader (1880, 512).

57 Henry Howard, 1st Earl of Northampton (1540–1614), younger brother of Thomas Howard, 4th Duke of Norfolk, who had been executed in 1572 for having participated in the Ridolfi plot. In her letters to Castelnau, Mary usually identified Howard using the codename “mon frere” (Bossy 2001, 68).

58 Reproduced in Labanoff (1844, v. 5, 114).

59 “user” and “avoir” in modern French; “to use” and “to have” in English, respectively.

60 “je” in modern French, “I” in English.

61 Those conclusions are in line with the structure of another cipher used by Castelnau (see Figure 16 in Section 4.9).

62 The letter y was often written in Middle French in places where i would be written in modern French (e.g., ‘roy’ in Middle French, ‘roi’ in modern French, ‘king’ in English).

63 William Cecil, 1st Baron Burghley (1520–1598), Lord High Treasurer and chief adviser of Queen Elizabeth I.

64 TNA SP 52/22/22. This may be the one Mary sent early in 1586 (Mary to Châteauneuf, 31 January 1586, reproduced in Labanoff 1844, v. 6, 258; Guy 2004, 480).

65 BnF 500 de Colbert 472, p.347. Also transcribed in Dewaël, W. Table de chiffrement ou Chiffre de Michel de Castelnau de la Mauvissière (1520–1592) durant son ambassade en Angleterre (1575–1585). Accessed November 17, 2022. https://manuscrit-esperluette.jimdofree.com/chiffre-ambassade/

66 Tomokiyo (2019–2022).

67 Jean Arnault, Seigneur de Chérelles, secretary to Castelnau.

68 TNA SP 53/10/78, Labanoff (1844, v. 4, 307).

69 BnF 500 de Colbert 470/185, Labanoff (1844, v. 6, 267).

70 We used a number of criteria, the most obvious being that it is likely that confidential contents would have only been sent in cipher. When a letter contains references to communication channels or routes (which would indicate the existence of a parallel and unauthorized communication channel), references to other enciphered letters, negative statements against the queen or members of her council, or the like, it is likely that the letter was sent in cipher. Conversely, letters sent via the official channel, under Walsingham’s supervision, would not include contents which would harm Mary if they were read by the queen’s agents. In addition, Mary may have included in those letters contents that she actually wished them to read, such as requests she had made or was about to make, and denials of accusations made against her.

71 BL Harley MS 1582/306 from 22-25 July 1583, TNA SP 53/13/1 from 5 January 1584, and BL Harley MS 1582/313 from 21 March 1584.

72 After excluding F57, a letter from 30 October 1584 with mixed ciphertext and plaintext passages, and F308, an undated letter.

73 Castelnau to the king, 30 September 1578 (Teulet 1862, v. 3, 37), 8 February 1580 (ibid., 64), 25 September 1580 (ibid., 75), 10 February 1581 (ibid., 87), 24 April 1581 (ibid., 107), 13 September 1582 (ibid., 142); Castelnau to the Queen Mother from 28 September 1582 (ibid., 151); Castelnau to the king, 31 July 1583 (ibid., 226).

74 Castelnau to the king, 30 September 1578, reproduced in Teulet (1862, v. 3, 37), and Mary to Beaton, 9 May 1578, reproduced in Labanoff (1844, v. 5, 39), refer to letters from Mary to Castelnau that neither appear in our newly deciphered letters, nor in other known sources.

75 Three letters from Mary to Castelnau are referenced by Castelnau in his letters to the king, on 25 September 1580 (Teulet 1862, v. 3, 75), 10 February 1581 (ibid., 87), and 24 April 1581 (ibid., 107), but appear neither in our collection of newly deciphered letters, nor in other known sources. Castelnau’s letter to the king on 8 February 1580 (ibid., 64) has a reference to newly-deciphered letters (F181, F179), and five of the newly-deciphered letters reference other newly-deciphered letters from that time period.

76 Seven of the newly deciphered letters reference other newly deciphered letters from that time period. Castelnau’s letter to the Queen Mother from 28 September 1582 (Teulet 1862, v. 3, 151) refers to F151. Castelnau’s letter to the king from 13 September 1582 (ibid., 142) refers to a letter which appears neither in our collection of newly deciphered letters, nor in other known sources.

77 Section 6.4.

78 One letter, F46, is referenced in another source, in Castelnau to the king, 31 July 1583 (Teulet 1862, v. 3, 226).

79 See Section 6. Mary’s comments are from F74 and F231.

80 For example: F227, F233, F231, F38, F98, F21.

81 F64, F89.

82 F113, F34.

83 Robert Dudley, 1st Earl of Leicester (1532–1588).

84 Lady Arbella Stuart (1575–1615) was the daughter of Charles Stuart (hence, Mary was her aunt) and Elizabeth Cavendish (hence, Bess of Hardwick, Countess of Shrewsbury, was her grandmother). Arbella had a potential claim to the English throne as a descendant of Margaret Tudor, daughter of Henry VII.

85 Against Leicester: F179, F105, F163, F64. On the scheme involving Arbella: F235, F64, F89. Mary also reports to Castelnau (in F179) on Leicester’s marriage with the Countess of Essex, which was kept secret from Queen Elizabeth.

86 For example: F30, F181, F98, F185, F165, F174.

87 F231, F87.

88 On 22 August 1582, King James VI of Scotland James was abducted by nobles led by William Ruthven, 1st Earl of Gowrie.

89 See F229, F151, F118, F74, F123, F307, F190, F198, F225, F153, F247.

90 See F194, F21, F158, F89, F130, F54, F26, F113, F34.

91 E.g., Robert Beale, see F125.

92 E.g., William Fowler (F123, F247), and Archibald Douglas (F151), both were indeed recruited by Walsingham.

93 Using the alias “le sieur de la Tour” (cf. Bossy 2001, 79).

94 F125, F58.

95 In our decryptions, we slightly modernize the French spelling to distinguish between u and v and between i and j. We capitalize and insert punctuation and apostrophes as in present practice. As for diacritics, we only add an acute accent on a stressed final e (e.g., passé, but: arrivee, apres, voyla). In all other respects we preserve the original spelling, except for obvious enciphering errors which we tacitly correct. When a person or location is referred to only by a cipher symbol, we rely on the spelling of the same entity in plaintext letters known to have been written by Nau, Mary’s secretary, e.g., Sheffeild (sic).

96 F42, F46, F58, F69, F78, F96, F221, some of those are also reproduced in Labanoff (1844).

97 We relied on various sources, such as the Calendars of State Papers, Leader (1880), and in particular on known letters from Mary to Beaton as reproduced by Labanoff (1844). In some cases, it turned out that such a letter to Beaton had the same dating and was apparently sent as an enclosure along with the enciphered letter to Castelnau. In the summaries, we indicate this in a footnote.

98 Such evidence is based, for example, on the mention of specific days of the week in some letters. Famous examples are Mary’s farewell letters to the Duke of Guise and to Beaton, whose date is given in each case as “ce jeudy vingt-quatriesme de novembre 1586” (24 November 1586 was a Thursday only according to the old style). We also find evidence in our newly deciphered letters, such as when Mary reports on 23 May 1584 (F42) to have been informed that Francis Throckmorton was put on trial “jeudy dernier”. The trial took place on 21 May (old style), indeed a Thursday, thus Mary’s date must be old style as well: otherwise, the dating of her letter would have preceded the trial (23 May 1584 new style is 13 May 1584 old style).

99 Those enclosed letters were likely written fully in cipher, like the main letter. The marking symbols were necessary to help Castelnau’s staff identify the recipients of the various enclosures.

100 Dated to 1578: A letter to Beaton of 10 April 1578 (Labanoff 1844, v. 5, 21) shows striking thematic overlap and even some of the same wording as the deciphered letter. The restoration of Atholl and Argyll to power also strongly suggests 1578.

101 The enclosed letter to Beaton is probably the one given in Labanoff (1844, v. 5, 31), dated 2 May 1578.

102 Not identified. Perhaps identical with a certain Singleton mentioned by Mary in letters to Beaton (see Section 6.4).

103 James VI (1566–1625), King of Scotland from 1567, who became James I, King of England from 1603.

104 Henry Hastings, 3rd Earl of Huntingdon (c. 1535–1595).

105 Colin Campbell, 6th Earl of Argyll (c. 1542–1584).

106 John Stewart, 4th Earl of Atholl (died 1579).

107 Not identified. The name also appears in a letter to Beaton of 7 November 1577 (Labanoff 1844, v. 5, 11).

108 Ninian Cockburn (died 1579).

109 James Douglas, 4th Earl of Morton (c. 1516–1581), regent of Scotland.

110 Dated to 1578: Mentioning Mar and Gondy, cf. Labanoff (1844, v. 5, 30, 38–39), and the reported reconciliation of Argyll and Athol with Morton.

111 Probably a secretary to Don John of Austria.

112 Jérôme de Gondi, Baron of Codun (c. 1550–1604), according to CSP Vatican (v. 2, 441 no. 840, footnote 27).

113 Thomas Wilson (1524–1581), privy councillor and Secretary of State (1577–81) to Queen Elizabeth I.

114 John Erskine, Earl of Mar (1558–1634).

115 Dated to 1579: Based on the timing of the transfer of office from Dolu to de Chaulnes, cf. Mary to Beaton, 24 June 1579: “J’espère, en bref, vous résouldre de celuy que je pourvoieray en son estat (i.e., Dolu’s office).” (Labanoff 1844, v. 5, 84). Also, this letter is most probably referred to in Nau’s postscript in the letter of 6 January 1580 (F179).

116 Antoine de Chaulnes, brother-in-law of Jean Arnault, Castelnau’s secretary.

117 Adam Blackwood (1539–1613), known for his later work Histoire de la martyre de la royne d’Escosse (1589).

118 René Dolu, Mary’s treasurer in France.

119 Dated to 1579 (tentatively): The contents, especially on the state of marriage negotiations, fit 1579 best, but there is no definitive evidence. Unfortunately, the mentioned letters from Mary to Queen Elizabeth supporting the marriage do not seem to have survived. John Hamilton fled to England in 1579 and from there to France. Castelnau may have assisted him (see the end of the summary).

120 John Hamilton, 1st Marquess of Hamilton and Commendator of Arbroath (1540–1604), usually referred to by Mary as “Arbroth”.

121 Dated to 1579: Mary’s non-encrypted letter to Castelnau from 21 November 1579 (Labanoff 1844, v. 5, 108–112) clearly deals with the same topic, the slander against Mary.

122 See letter from Castelnau to Mary, 7 December 1579 (Teulet 1862, v. 3, 57–61), which mentions Mary having allegedly spoken ill of the Duke of Anjou (“que aviez mal parlé de Monseigneur le Duc”). Leader (1880, 425) suggests that Mary may have spoken against the proposed marriage between the queen and the Duke of Anjou.

123 Dated to 1580: Mary continues the slander topic.

124 Thomas Fowler, former servant to the late Countess of Lennox, at that time serving the Earl of Leicester as a steward. It appears that Castelnau reported in his last letter to Mary that the slander was traced to Fowler.

125 See F233.

126 On 8 February 1582 (new calendar style), Castelnau reported similar contents to the king (Teulet, v. 3, p. 64), referring to a ciphered letter from Mary, which can be identified as this letter (possibly also referring to contents from other letters such as F181).

127 Lettice Knollys (1543–1634), who had a first marriage to Walter Devereux, 1st Earl of Essex (died 1576).

128 Dated to 1580: Mary writes that she is “sur le poinct de fermer mes premieres lettres,” which fits well with F179. Moreover, the letter opened in the Privy Council was likely her official response to the calumny from 21 November 1579 (Labanoff 1844, v. 5, 108).

129 Dated to 1580: Mentioning Leicester’s (temporary) fall from grace and John Hamilton’s stay in France.

130 The second location is indicated by an undeciphered symbol.

131 Dated to 1580: Leicester’s fall from grace, the expected arrival of the Duke of Anjou, and the latter’s plea for Leicester’s rehabilitation, all of which are also reported in a letter dated 13 January 1580 from Bernardino de Mendoza to King Philip II (CSP Spain (Simancas) v. 3, 1).

132 The duke’s request for Leicester’s rehabilitation.

133 Dated to 1580: Striking thematic overlap with a letter to Beaton dated 20 February 1580 reproduced in Labanoff (1844, v. 5, 120) and the mention of the marriage of Robert Sackville and Lady Margaret Howard, which took place in February 1580.

134 Most probably the letter to Beaton of the same date, reproduced in Labanoff (1844, v. 5, 120).

135 Sir Edward Stafford (1552–1605) had negotiated in France on the Elizabeth-Anjou marriage project.

136 Douglas Sheffield (1542/1543–1608), with whom Leicester had an affair for many years, had recently married Stafford.

137 Edward de Vere, 17th Earl of Oxford (1550–1604).

138 Oxford was the son-in-law of Burghley, who in turn had played a key role in the trial of the Duke of Norfolk.

139 Philip Howard, 13th Earl of Arundel (1557–1595), son of the Duke of Norfolk, who had been executed for his implication in the Ridolfi Plot. Although the title of Earl of Surrey was not granted at that time (his father having forfeited all his titles), Mary deliberately calls him Earl of Surrey.

140 Thomas Sackville, Baron Buckhurst (1536–1608) had married Lady Margaret Howard (1562–1591), a half-sister of Philip Howard.

141 Most probably Henry Middlemore (died 1592), gentleman of the privy chamber to Queen Elizabeth.

142 Claud Hamilton (1546–1621), brother of John Hamilton.

143 Dated to 1580: Mention of Robertson’s absconding and his plan to flee to France, which fits the context of other letters from this period, both to Castelnau in our collection and to Beaton (Labanoff 1844, v. 5, 179–82, 226–31; assuming Robertson and Singleton are one and the same, cf. Section 6.4).

144 Dated to 1580: The torn-out paper only shows "…tiesme septembre", probably “vingt-septiesme” because a letter to Beaton of 27 September 1580 (Labanoff 1844, v. 5, 179–82) has some strikingly similar wordings.

145 Jean de Simier, Baron de St Marc, the Duke of Anjou’s Master of the Wardrobe, negotiated in England on the terms of the proposed marriage of the duke with Queen Elizabeth.

146 And probably less favorable to France.

147 Dated to 1581: Walsingham left for France in July 1581 (cf. Wilson 2007, 148).

148 Unidentified.

149 Dated to 1581: George Douglas’s dispatch to France was in September of that year (cf. CSP Scotland v. 6, 47 no. 60).

150 Probably the letter to Beaton on the same date, reproduced in Labanoff (1844, v. 5, 253).

151 Archibald Douglas, 8th Earl of Angus (1555–1588), a nephew of Regent Morton.

152 George Douglas of Rungallie, gentleman of the king’s bedchamber to James VI. In 1568, had helped Mary to escape from Lochleven Castle where she had been under the guard of his brother William Douglas, 6th Earl of Morton (c. 1540–1606).

153 Gaspard d’Alègre (died 1610), Seigneur de Viverols.

154 Jean de Beaucaire, Seigneur de Puyguillon, Sénéchal de Poitou.

155 Dated to 1581: Mary is expecting Beale, and his first visit to Sheffield was in November 1581 (Basing 1994, 66). A mentioned letter to Queen Elizabeth must be the one of 10 October 1581 (Labanoff 1844, v. 5, 265), which also deals with Mary’s request to send an envoy to Scotland.

156 Unidentified.

157 Dated to 1581: Mary, still expecting Beale’s visit, is in poor health, which is consistent with Beale’s report of 14 November 1581 (CSP Scotland v. 6, 77 no. 80). Thomas Randolph had been dispatched to Scotland in 1581 (op. cit. v. 5, 571 no. 649).

158 Chrétien de Savigny, Seigneur de Rosne (died 1596), chamberlain to the Duke of Anjou.

159 Thomas Randolph (1523–1590), English diplomat.

160 Dated to 1582: The stay of the Duke of Anjou in London is mentioned, occupying all the attention of Queen Elizabeth.

161 Dated to 1582: The reference to her prior letter dated “le neufiesme de ce mois” can be assigned to F21. Moreover, it is no coincidence that this letter for the first time mentions Arbella Stuart: after the death of her mother on 16 January 1582, she became the ward of Bess of Hardwick, Countess of Shrewsbury.

162 More precisely, to either Robert Devereux, her son by her late husband, or Robert Dudley, Lord Denbigh, her son by Leicester, whom she had given birth to only the year before. In fact, the marriage plans for Arbella were about the latter, but he died as early as 1584. See also F64 and F89.

163 Dated to 1582: Mary mentions that the Duke of Anjou has recently departed, which he actually did on 10 February 1582.

164 Most probably, Christopher Hatton (1540–1591), Lord Chancellor of England.

165 Dated to 1582: Mary’s continued efforts to send Nau to Scotland, Leicester escorting the Duke of Anjou to Antwerp, and Mary’s desire to obtain a copy of the queen’s memorandum to Beale (cf. CSP Scotland v. 6, 74 no. 79 from November 1581).

166 Probably referring to the Burgundian treaty of 1548, also known as the Transaction of Augsburg, settling the status of the Habsburg Netherlands within the Holy Roman Empire.

167 Dated to 1582: Mentions the recent assassination attempt on William the Silent, which was on 18 March 1582.

168 Sir Fulke Greville (1554–1628), poet and statesman.

169 William the Silent, Prince of Orange (1533–1584), the leader of the Dutch Revolt.

170 Thomas Radclyffe, 3rd Earl of Sussex (c. 1525–1583), Lord Chamberlain to Queen Elizabeth.

171 Nicolas Errington (died 1593). In November 1581, he was refused entrance to Scotland (CSP Scotland v. 6, 84 no. 83).

172 Dated to 1582: In this letter, Mary refers to F125 from 7 April 1582 and mentions the letter she was requested to write to James about his refusal to receive the queen’s envoy Errington.

173 Unidentified.

174 Unidentified.

175 Dated to 1582: Refers to F165 from 16 April 1582 and mentions the letter Mary has written to James about his refusal to receive the queen’s envoy.

176 Most probably, the letter from 16 April 1582, reproduced in Labanoff (1844, v. 5, 293–294) and CSP Scotland v. 6, 115 no. 106.

177 Probably Jean Bodin (1530–1596), a French jurist and expert on sovereignty.

178 Dated to 1582: Refers to F165 from 16 April 1582, and apparently to a letter to Beale from 16 April 1582 (CSP Scotland v. 6, p. 113 no. 105). Also, Mary still hopes to be able to send Nau to Scotland.

179 Dated to 1582: The two physicians mentioned must be Dr. Smythe and Dr. Barsdale who, according to Leader (1880, 504), were in Sheffield in May 1582.

180 Paracelsus (c. 1493–1541) was a Swiss physician, alchemist, lay theologian, and philosopher of the German Renaissance. He was a pioneer in several aspects of the “medical revolution” of the Renaissance, emphasizing the value of observation in combination with received wisdom. He is credited as the "father of toxicology". Paracelsianism is the early modern medical movement inspired by the study of his works.

181 Dated to 1582 based on the mention of the Duke of Lennox. Esmé Stewart was not made Duke until August 5, 1581, and died on May 26, 1583, so this letter, written on a July 31 (“ce dernier juillet”), can only be from 1582.

182 The letter in question is most probably the one given in CSP Scotland (v. 6, 133 no. 126), dated June 17, 1582.

183 Probably a valet in the service of French king Henry III.

184 Dated to 1582: Obviously written shortly after the Ruthven Raid (August 22, 1582). Interestingly, there is a letter sent via the official channel to Castelnau from the same date which does not make any mention of the raid (Labanoff 1844, v. 5, 298–300).

185 Dated to 1582: Also refers to the Ruthven Raid.

186 Probably the letter to Beaton written on the same date, reproduced in Labanoff (1844, v. 5, 308).

187 On 28 September (new calendar style), Castelnau reported this request to the king (Teulet, v. 3, 151), referring to a ciphered letter from Mary, which can be identified as this letter with high confidence.

188 George Carey (1547–1603) was the eldest son of Henry Carey, 1st Baron Hunsdon, and later became 2nd Baron Hunsdon.

189 Archibald Douglas, Parson of Douglas (before 1540–1603) was, as Leader (1880, 535) puts it, albeit in the context of 1583, “playing the double part of betraying Mary Stuart to Walsingham, and pretending to betray Walsingham to Mary Stuart. “

190 In 1573, Sir William Kirkcaldy, Laird of Grange (born c. 1520) had been hanged at the conclusion of the siege of Edinburgh. The girl (“ceste bastard”) is most probably his illegitimate daughter of whom Ninian Cockburn had reported two days before the execution (CSP Scotland v. 4, 601, no. 710).

191 Dated to 1582: Mary reports on the aftermath of the Ruthven Raid of 22 August 1582 that she has reliable information that the Duke of Lennox has not yet left Scotland and is safe at Dumbarton Castle.

192 Douglas had been arrested by Henry Killigrew (CSP Scotland v. 6, 171 no. 171, Killigrew to Walsingham, 17 September 1582).

193 Dated to 1582: Du Ruisseau has just left Sheffield, on his way to France. According to Leader (1880, 512), “du Ruisseau was permitted to return to London on his way back to France in the middle of October [1582]”.

194 Renée de Lorraine-Guise (1522–1602), Abbess of St. Pierre, Reims.

195 Dated to 1582: See footnote for the previous letter, F245.

196 This letter to Ruisseau (enclosed with the enclosure for Beaton) could be the undated memorandum given in Labanoff (1844, v. 5, 301–308), which Labanoff tentatively dates to 2 September 1582.

197 Also, a new cipher symbol for Henry Howard is introduced.

198 William Parry, intelligencer, executed in 1585 for conspiring against the English queen.

199 Dated to 1582: Mary mentions that Queen Elizabeth’s answers to du Ruisseau’s requests have been presented to her by Shrewsbury; this was at the end of October 1582 (Leader 1880, 512), probably on the 31st (CSP Scotland v. 6, 194 no. 199).

200 Gilles du Verger, Mary’s counselor, involved in the affairs of her dowry.

201 Dated to 1582: This is the cover letter of a letter (F307) from the same date to de la Mothe-Fénelon, the French king’s envoy to Scotland.

202 John Mosman, son of James Mosman, who had been hanged after the siege of Edinburgh. In November 1582, he wrote to Mary asking for a reward as the son of her “grace’s master coiner and true subject” (CSP Scotland v. 6, 210 no. 211). Later, he apparently also acted as a courier (ibid. p. 327 no. 345, and p. 688 no. 724).

203 Unidentified.

204 William Fowler (c. 1560–1612), Scottish poet, “familiar of Castelnau, intelligencer for Walsingham” (Bossy 2001, xii). Not to be confused with Thomas Fowler (see footnotes to F179 and F64).

205 Dated to 1582: De la Mothe-Fénelon was in Scotland toward the beginning of 1583.

206 Thomas Kerr of Ferniehirst (died 1585), a Scottish landowner, Roman Catholic and supporter of Mary, Queen of Scots.

207 James Stewart (died 1595), Earl of Arran.

208 George Gordon, 6th Earl, later created 1st Marquess of Huntly (1562–1636).

209 George Carey (1547–1603), son of Baron Hunsdon, cf. footnote to F151 (10 September 1582).

210 Henry Stanley (1531–1593), 4th Earl of Derby.

211 Literally, “la princesse de”, followed by a symbol that appears only once in all our letters and therefore cannot be assigned with certainty. We suggest “Lorraine” because in 1580, in a letter to Beaton, Mary mentioned that Catherine de’ Medici had suggested the Princess of Lorraine as a match for James (Labanoff 1844, v. 5, 174). Another possibility would be the Princess of Navarre (Catherine de Bourbon), but it seems that Mary opposed this match, as a letter to Archibald Douglas in 1582 reveals: “concerning the marriage whereof some speaking has been of the daughter of [the king of] Navarre, his majesty is nowise inclined that way, because such is his mother’s direction and mind” (CSP Scotland v. 6, 236 no. 255).

212 Francis Stewart, 5th Earl of Bothwell (c. 1562–1612), was the son of Jean Hepburn (Mary’s sister-in-law by her third marriage with Lord Bothwell) with her first husband John Stewart, Mary’s half-brother (illegitimate son of James V).

213 Dated to 1582: By the end of 1582 – de la Mothe-Fénelon had not yet been authorized to go to Scotland.

214 Dated to 1582: Same date as F190, its cover letter.

215 Dated to 1582: End of 1582 – de la Mothe-Fénelon recently allowed to go to Scotland.

216 Dated to 1583: Mentions de la Mothe-Fénelon’s trip to Scotland in early 1583.

217 Dated to 1583: Mentions de la Mothe-Fénelon’s trip to Scotland.

218 William Davison (c. 1541–1608), secretary to Queen Elizabeth I.

219 On 4 January, William Fowler mentioned in a letter to Walsingham Lennox’s “conference with the Queen’s majesty, whose highness … first spoke hotly to him, yet courteously at the end,” but also that “Monsieur Mauvissière spoke with him two hours” (CSP Scotland v. 6, 244, no. 267).

220 Those “things” are not described in F247, but they may have been reported by the queen to Castelnau, then by Castelnau to Mary.

221 The mention of two different persons named Fowler in the same letter is immensely helpful in distinguishing the two. It seems that Labanoff (1844) equated both persons when he noted: “William Fowler, ancien serviteur de la comtesse Marguerite de Lennox” (v. 5, 361). Leader follows him (1880, 535).

222 “Je voy bien que c’est quelque coureuse qui ne demande qu’a passer pays et s’il est ainsi au lieu de luy faire jamais bien je la ferois plustost fouester au cul d’une charrette si elle estoit en lieu ou j’eusse puissence.”

223 Dated to 1583 – mentions the French Fury (17 January 1583).

224 John Dudley, who put Jane Grey to the throne in 1553.

225 Presumably, the reports of the arrested Frenchman and of the offer of marriage rejected by Mary both refer to the envoy of George John I, Count Palatine of Veldenz, who was apprehended on his way to Scotland where he wanted to propose the marriage of James with a daughter of the count, and was detained in York (CSP Scotland, v. 6, 299 no. 319). The proposal of the count, who hoped to escape poverty with such a marriage, surprised both the English and the French sides. Mary wrongly assumed the offer was made by de la Mothe-Fénelon.

226 Dated to 1583: Mentioning de la Mothe-Fénelon’s trip to Scotland.

227 Under the codename “le frere de Sa Majesté.”

228 This letter to de la Mothe-Fénelon is neither in our collections of deciphered letters, nor in Labanoff (1844).

229 This letter to the queen is not in Labanoff (1844); more specifically, Labanoff has nothing from Mary to anyone from February to June 1583. But apparently, the letter reached Elizabeth (CSP Scotland v. 6, 357, no. 383).

230 Sir William Stewart of Houston (c. 1540 – c. 1605), a.k.a. Colonel Stewart, was a Scottish soldier, politician and diplomat.

231 Elizabeth Cavendish, daughter of the Countess of Shrewsbury, was married to Charles Stuart, Earl of Lennox, and bore Arbella Stuart.

232 Also mentioned in F247.

233 Thomas Baldwin, the Earl of Shrewsbury’s agent in London (Bossy 2001, 92).

234 Dated to 1583: Mary mentions the death of George Buchanan (28 September 1582), while the letter must precede the arrest of Throckmorton in November 1583.

235 Francis Throckmorton.

236 Mentioned as a courier in several letters. See also Bossy (2001, 33, 66f, 70n, 79–81, 90, 100, 109, 123, 144, 149).

237 François de Roncherolles, Sieur de Mainneville was sent together with de la Mothe-Fénelon to Scotland in February 1583, cf. Section 2.

238 Lettice Knollys, Countess of Essex, then married to Leicester, had two daughters from her first marriage, the elder of whom had married in 1581. The younger was Dorothy Devereux (c. 1564–1619).

239 Nicolas d’Angennes (1533–1611), Seigneur de Rambouillet, chamberlain to French king Henry III.

240 George Buchanan (1506–1582), a Scottish historian and humanist scholar.

241 This letter to Castelnau is not in Labanoff (1844). However, Castelnau seems to refer to it in a letter to Walsingham of 6 April 1583 (CSP Scotland v. 6, 370, no. 388).

242 Unidentified. Might be the Edward Foster/Robert Bruce mentioned in CSP Scotland v. 10, 767 no. 746 (August 1592) as a servant of the Bishop of Glasgow. We assume “Fauster” in F54 and F26 to be the same person.

243 Probably Andrew Hamilton of Cochno, Governor of Dumbarton castle and Provost of Glasgow (died 1573).

244 Jerome Pasquier was master of the wardrobe of the Queen of Scotland. He was imprisoned at the time of the Babington plot, and he confessed having enciphered letters for her (CSP Scotland v. 9, 89, no. 80).

245 Probably William Holt (1545–1599), arrested in Leith, held for a while, but later taken by James to Scotland.

246 This seems to be contradicted by a letter from Bowes to Walsingham from March 28th: “one of the ciphers found with Holte is confessed to have been given forth by the Scottish Queen” (CSP Scotland v. 6, 347, no. 368).

247 Dated to 1583: Mary complains that Beale has no formal documents with him to establish his mandate by the queen, cf. Leader (1880, 521).

248 Under the codename “le sieur de la Tour”.

249 Throckmorton.

250 Dated to 1583: Beale had recently negotiated with Mary in person which fits with his visit to Sheffield from April 12 to 29, 1583 (cf. Basing 1994).

251 Under the codename “le frere de Sa Majesté”.

252 Literally: aye.

253 Literally: succedont.

254 He may be more willing to tell you verbally than me in writing.

255 Dated to 1583: Mary is awaiting Sir Walter Mildmay and Beale, who arrived on 1 June 1583 (Leader 1880, 527).

256 Probably, the duke’s servant John Smallett, sent to London in May 1583 (CSP Scotland, v. 6, 461, no. 474).

257 Sir John Norris, or Norreys (ca. 1547–1597), an acclaimed British soldier.

258 In May 1583, Norris left Flanders, temporarily (CSP Foreign: Elizabeth, v. 17, 353 no. 322).

259 Literally, “expedition”. Probably, a legal document.

260 Renée de Lorraine-Guise (1522–1602), Abbess of Saint-Pierre-les-Dames, Reims.

261 John Mowbray, Laird of Barnbougle was a supporter of Mary. His wife was a sister of William Kirkcaldy of Grange.

262 Leader (1880, 318, 332, 350) mentions Mademoiselle Rallay, an old French attendant who came to Sheffield in 1574. She might be her niece.

263 Dated to 1583: Mildmay and Beale are reported to have arrived this very evening, which is confirmed by Leader (1880, 527).

264 This ring, apparently a gift from Mary to Howard, was brought up against him in his interrogations in December 1583/January 1584, as the ring had also been mentioned in a letter from 3 July 1583 (F46) which was leaked to Walsingham, cf. Bossy (2001, 68–70).

265 Archibald Stewart (c. 1530–1584), Provost of Edinburgh.

266 Bernardino de Mendoza (c. 1540–1604), see also Section 6.1.

267 Dated to 1583: Mentions the arrest of Charretier in June 1583 (cf. Greengrass 1981, 335).

268 Mary asks Castelnau, if he receives a package either from Beaton or the Duke of Guise, to forward it to her, and also to send “him” the enclosure, but it is not clear to whom “him” refers.

269 Mathurin Charretier. He was once secretary of marechal d’Anville, a suitor of Mary. He was taken prisoner by the Spanish in 1583. See Greengrass (1981).

270 Valentin de Pardieu, sieur de la Motte, governor of Gravelines (Greengrass 1981, 335).

271 This summary is not included in the deciphered letter. It is also missing from the Harleian copy, and it is not reproduced in Labanoff (1844). We suggest that it was the basis for the summary that Castelnau forwarded to the king on 31 July 1583, a copy of which is in the d’Esneval papers (Teulet v. 3, 229–237).

272 The differences between the Harleian version and the d’Esneval version (reproduced in Labanoff 1844, v. 5, 349–352) are discussed at length by Bossy (2001, 67-70), who had hypothesized that either the mole in the French embassy had changed parts of the text, or Mary had sent the letter in two versions, leaning toward the latter option. We propose here a third hypothesis. The d’Esneval version of this letter is probably a copy of what Castelnau sent to the king, along with a summary of the status of Mary’s negotiations, as he writes to him on 31 July: “J’envoye à Vostre Majesté un petit sommère des articles de ce qui a esté traicté avec ladicte Royne d’Escosse par les députez de la Royne d’Angleterre, et ung extraict [de] ce qu’elle m’a escript en chiffre.” (I am sending to Your Majesty a short summary of the points which have been discussed with the said queen of Scotland by the deputies of the queen of England, and excerpts from what she (the queen of Scotland) wrote to me in cipher; Teulet 1862, v. 3, 229). This would account for a difference of 11 days between the date in the d’Esneval version and the date in the Harleian copy, which is only one day if one assumes that in the d’Esneval version, the date was written according to the new Calendar style. It would also account for the different contents in the d’Esneval version, Castelnau having probably removed or revised some parts irrelevant to the negotiations, also adding other material. Nevertheless, our analysis confirms Bossy’s assertion that the mole in the embassy did not tinker with the contents of the letter.

273 Dated to 1583 based on the year of the copy in Hatfield Calendar which is tentatively dated to September 1583. Bossy (2001, 184) estimates the letter to be from c. 10 September 1583, arguing that is must have been written after the letter from 3 September 1583 (F69, #6 in Bossy’s list), but without any further details. The date on our newly deciphered version is 17 August.

274 Not reproduced in Labanoff (1844). The Hatfield Calendar copy is available online: https://www.british-history.ac.uk/cal-cecil-papers/vol3/pp10-14 [Accessed December 10, 2022].

275 Probably Patrick Gray, 6th Lord Gray (died 1612), a.k.a. Patrick, Master of Gray, Scottish nobleman and politician.

276 Sir Charles Cavendish (1553–1617), son of Bess of Hardwick (later Countess of Shrewsbury) and William Cavendish.

277 Claude Courcelles, Castelnau’s secretary, who controlled a network of clandestine communications. See Section 6.4.

278 Reproduced in Labanoff (1844, v. 5, 361–369).

279 Mary Seton (1542–1615), one of the four attendants of Mary, Queen of Scots, known as the Four Marys. She and Mlle de Courcelles had been replaced by the waiting women coming from France (Bossy 2001, 69 fn 17; cf. F26).

280 Catherine-Marie de Castelnau de Mauvissière. See Mary to Castelnau, 12 November 1583, BnF 500 de Colbert 471/297, reproduced in Labanoff (1844 v. 5, 372, see 386), Mary to Catherine-Marie, 26 January 1584 (ibid. 406).

281 Unidentified. Literally, “Thoieson”, but might be due to one or more encryption errors.

282 Dated to 1583: The first lines refer to Walsingham’s visit to Holyrood, Edinburgh, in September 1583, mentioned in Labanoff (1844, v. 5, 360).

283 Free of her obligations toward France – sounds like a veiled threat.

284 Reproduced in Labanoff (1844, v. 5, 423–430).

285 “I beg you very humbly, Monsieur, to keep all this as secret as you possibly can, so that Monsieur the ambassador absolutely does not realise …” (Bossy 2001, 106; translation by Bossy). This postscript is considered evidence that the copies in British archives in the same hand (mostly in BL Harley MS 1582, see Appendix C) were made by a mole inside the embassy, and leaked to Walsingham.

286 Reproduced in Labanoff (1844, v. 5, 457–465).

287 Reproduced in Labanoff (1844, v. 5, 468–474), but without Nau’s postscript.

288 Reproduced in Basing (1994).

289 Reproduced in Labanoff (1844, v. 6, 45–49), but without the enciphered parts.

290 Taken from the transcript in Labanoff. We keep Labanoff’s spelling and punctuation unchanged, which may differ from our own conventions for the deciphered parts.

291 Being moved to another location, and another guardian.

292 Edward Manners (1549–1587), 3rd Earl of Rutland, 14th Baron de Ros of Helmsley.

293 John St. John, 2nd Baron St. John of Bletso (1596), an English peer, who was indeed appointed to be Mary’s keeper in January 1585.

294 To Scotland.

295 This letter is undated. It is held in the same archive collection (BnF 500 de Colbert 470) as the letter to de la Mothe-Fénelon from 12 December 1582 (F307), in an adjacent folio, but there is not enough evidence to establish temporal proximity between them.

296 Probably as Mary’s hostess.

297 Probably referring to her rights of succession.

298 E.g., Bossy (2001, 99).

299 Bossy (2001, 33).

300 E.g., “par la voye ordinaire”, Mary to Beaton, 4 July 1579, Labanoff (1844, v. 5, 85, see 87); “par la voye de Valsingham”, Mary to Castelnau, 22 and 25 July 1583, BL Harleian MS 1582/306, Labanoff (1844, v. 5, 357, see 358).

301 Also, CSP Scotland, v. 5, no. 225.

302 Mary to Castelnau, 8 October 1582, BnF Fr.3181/f.9, reproduced in Labanoff v. 5, 312; F118.

303 Bossy (2001, 32–33).

304 Mary to Castelnau, 3 September 1583, Hatfield Cecil Papers 133/31, Hatfield Calendar 13/234, F69, reproduced in Labanoff (1844, v. 5, 361, see 364).

305 One might expect that letters leaked by a mole in the French embassy hired by Walsingham (see Section 6.4) would have been secret letters because, obviously, there should have been no need to leak letters sent in clear via the official channel. Unexpectedly, however, Add MS 48049 includes copies in the hand of Feron of two letters of which the original in clear is preserved in French archives and which are thus considered to have been sent in clear via the official channel (Add MS 48049 f.272r-273r vs. Labanoff 1844, v. 5, 440, 22 March 1584 and Add MS 48049 ff. 280r-281v vs. Labanoff 1844, v. 5, 391, 2 January 1584). Probably, the mole did not bother, or did not have the time, to sort out which letters were official letters and were likely to be known to Walsingham anyway, and which were not.

306 When we only have the plaintext copy of a letter in archives, it is not always obvious that the copy was made from an original cleartext rather than from a deciphered plaintext, unless the copy is an original bearing an autograph signature, typically labeled “original” in Labanoff (1844).

307 A letter from 15 August 1585, the only one from Mary to Catherine de’ Medici from this period, is an autograph (Labanoff 1844, v. 6, 206).

308 Bossy (2001, 31, 108); Bossy (1991, 55); Mary to Beaton, 20 February 1576 (Labanoff 1844, v. 4, 284, see 285), 21 May and 1 June 1576 (ibid. 311, see 315); Pollen (1922, lii).

309 Mary to Beaton, 7 April 1582, reproduced in Labanoff (1844, v. 5, 280, see 281).

310 It is possible that Mary’s wording was intended to conceal from Castelnau the existence of her secret channel with the Spanish ambassador.

311 In March 1575, Shrewsbury reported to Lord Burghley, “What intellygens passeth for this Quene to and fro my house I doo not know.” (Collinson 1987, 24; Tu 2012, 169).

312 Mary to Beaton, (June 1574), reproduced in Labanoff (1844, v. 4, 175, see 181).

313 Strickland (1854, v. 7, 145, 160).

314 Mary to Beaton, 20 February 1576, reproduced in Labanoff (1844, v. 4, 284, see 294).

315 Strickland (1854, v. 7, 52), citing Shrewsbury and Huntingdon to Cecil, 19 December 1569 in State Paper Office MS.

316 Bossy (2001, 125); Pollen (1922, cxxix).

317 Strickland (1854, v. 7, 268).

318 Mary to André Beaton, 22 August 1577, reproduced in Labanoff (1844, v. 4, 377, see 378).

319 Strickland (1854, v. 7, 313, 315); Mary to Castelnau, 8 October 1582, BnF Fr.3181/f.9, reproduced in Labanoff (1844, v. 5, 312); Leader (1880, 511). See also a footnote for F74 in Section 5.3 regarding the letter to du Ruisseau.

320 Leader (1880, 477–481), citing Beale to Walsingham, 23 November 1581 (SP53/11/69).

321 Strickland (1854, v. 7, 335, 358); Leader (1880, 568–569).

322 Morris (1874, 50–51, 125–126).

323 F78. Also in Hatfield Calendar v. 3, 12 (Cecil Papers Volume 162, 23–24).

324 There are similarities between the person named ‘Robertson’ mentioned in our deciphered letters and ‘Singleton’ mentioned in letters to Beaton, reproduced in Labanoff (1844). (i) Mary thinks it best for Robertson to go to France because of the “cruel persecution” to Catholics (F237, 16 September 1580), whereas a letter to Beaton mentions the bearer is resolved to cross the sea because of the “cruel persecution” against Catholics and recommends him to Beaton (Labanoff 1844, v. 5, 179, 27 September 1580). Labanoff identifies this bearer with Singleton, which is logical because Mary asks Beaton in this letter to “deliver 500 ecus to this bearer,” while in another letter naming Singleton, Mary thanks Beaton for the 500 ecus delivered (ibid. 226). (ii) Both F239 (27 September 1580) and the above-mentioned letter to Beaton (Labanoff 1844, v. 5, 179, see 181) mention a pension and presentation to the king by the Duke of Guise. (iii) An enclosure for Singleton in a letter to Beaton was marked with the symbol “T” (ibid., 231), which is the same symbol as the one used for enclosures for Robertson in letters to Castelnau, F237 and F181. The same marking system of enclosures was used in letters from Mary to Beaton and Castelnau, e.g., the symbol for Morgan (TNA SP53/10, reproduced in Labanoff 1844, v. 5, 13, F21, and F185).

325 “There are (upon some letters of her majesty directed to my Lord Seton, to Flanders), some of her best servants in England discovered, and thereby now forced to come, as they say, to take part of my banishment” (Morgan to the Countess of Northumberland, CSP Scotland v. 5, 250). Also in F87, Robertson seems concerned about an interception of Mary’s letters to Seton in Flanders, but Mary states she has not written to him since he left.

326 Charles Paget to Mary, 17 July 1585 (TNA SP53/16/15).

327 Also spelled Fullgham or Foulgiam. Mary’s letter in cipher in TNA SP53/18/64 marked “Decifred 21 July 1586” is addressed “To Fulgeam.”

328 Bossy (2001, 31–35).

329 Labanoff (1844, v. 5, 430f); Bossy (2001, 14–28, 38, 43, 45–49).

330 Bossy (2001, 31–32).

331 Mary to Beaton, 20 May 1581, Labanoff (1844, v. 5, 225, see 227); Mary to Beaton, 27 September 1580 (ibid., 179, see 180).

332 The letter is in TNA SP53/12/92. On the capture, see Bossy (2001, 79, 184). On Feron possibly leaking the cipher table: In May 1583, Walsingham’s agent “Fagot” wrote to him that “I have made the ambassador’s secretary my friend that, if he is given a certain amount of money, he will let me know everything he does, including everything to do with the Queen of Scots and the cipher which is used with her” (Bossy 1991, 200; 2001, 16). Either way, we would have expected to find a copy of the reconstructed cipher key in British archives, but we have been unable to find such a copy. We found a very short undeciphered ciphertext fragment in the margin of a draft of a letter from Castelnau to Mary dated by Bossy (2001) to c. 14/24 December 1583 (BL Harley MS 1582, f.385). It deciphers as "monsieur … monsieur Nau trouvera", "madame … madame pour ja la." A seemingly corresponding plaintext is found on the second to last line on f.373r ("… madame. Monsieur Nau trouvera icy …") (March 1584). A possible interpretation is that the marginal ciphertext was scribbled in 1584 on the paper of the 1583 draft. However, it is not clear who wrote this scribble.

333 Bossy (2001, 175-176).

334 Bossy (2001, 115, 125, 174, 92–93).

335 Bossy (2001, 125, 116, 113, 117).

336 There are more than 10 letters from Mary to Castelnau from the spring of 1584 to November 1585 which were sent in clear and are now in French archives, which were presumably sent under the surveillance of Walsingham. However, there are no such letters to Châteauneuf (sent in clear) in French archives, from the time the latter was ambassador to England.

337 Walsingham to Phelippes, 3 September 1586 (TNA SP53/19/80).

338 Bossy (2001, 79); This alias is used in Castelnau to Mary from 1583 (ibid. 173), Mary’s letters to Castelnau from 1584 (Labanoff 1844, v. 5, 400, 428, 438, 472, 475), and in our newly deciphered letters, F125, F174, F130, F89, F34, F78, F58.

339 F58, also in Labanoff (1844, v. 5, 428–429). Bossy (2001, 108), in discussing this particular letter, considers the identification as probably made by Feron, who copied this letter, apparently assuming that Feron rendered the same symbol as “Sieur de la Tour” at one place and as “Throckmorton” at another, but our decipherment of F58 shows that is not the case. The enciphered text spells “de la Tour” and then “Trokmorton”. The tell-tale distinction was not made by Feron but it was in the original ciphertext.

340 Bossy (2001, 68, 108). One of the instances given by Bossy corresponds to F46. The words "mon frere" in F46 as well as its Harleian MS 1852 copy correspond to a blank in the d’Esneval version. See the footnote for F46 in Section 5.3.

341 Bossy (2001, 68).

342 Warnicke (2006, 215); F105.

343 Other uses are found in F30, F34, F46, F78, F89, and F113.

344 Bossy (2001, 124, 93) mentions the alias “Renous Banque”, based on the sentence “Le porteur s’appellera cy a present Renous Banque” (The bearer will now be called “Renous Banque”) in the plaintext copy of a deciphered letter to Courcelles (Labanoff 1844, v. 5, 479, see 480). This seems to be a decipherment error, as the sequences of deciphered letters would make better sense if split into words as “Le porteur s’appellera cy-apres entre nous Banque” (The bearer from now on will be called between us “Banque”), the alias being “Banque” rather than “Renous Banque”. This is a kind of difficulty one often faces when deciphering a ciphertext without word breaks.

345 While we first deciphered an initial set of letters (BnF fr. 2988), only after we recognized their value did we embark on a concerted effort to find other letters enciphered with the same graphical symbols. We started with BnF collections known to contain Castelnau’s papers. We found two additional encrypted letters in BnF Cinq Cents de Colbert 470 (F307, F308), and more importantly, a letter in BnF fr. 3158 (F57) from Mary Stuart to Castelnau mostly in clear, but with some passages encrypted. Labanoff was aware of this letter which he reproduced in Labanoff (1844), without the enciphered parts, and since he was also examining documents in clear from Mary Stuart in Castelnau files in BnF Cinq Cents de Colbert 470, in theory, he could have been able to deduce that the encrypted letters in that collection (F307, F308) were also from Mary Stuart. Such a conclusion might have also enabled a scholar examining the contents of BnF fr. 2988 to recognize they were also from Mary.

346 Megyesi et al. (2020) and Tomokiyo (2019–2022).

347 In particular, from the years 1578–1581 – see Section 5.1.

348 Bossy (1991) and Bossy (2001).

349 And via Castelnau, the channel with her associates and allies, mainly in France.

350 Jacques Nau, who wrote postscripts in several enciphered letters, Gilbert Curll, known to have enciphered Mary’s letters on several occasions (Pollen 1922, cxxix-cl) and also written postscripts to some of them, and Jerome Pasquier, who confessed having enciphered letters for Mary after he was arrested (CSP Scotland v. 9, 89, no. 80), are obvious candidates, but we could not reach any definitive conclusion.

351 A letter from Castelnau to Mary, see Section 6.4.

352 The letters mention more than one hundred names.

353 https://commons.wikimedia.org/wiki/File:Fran%C3%A7ois_Clouet_-_Mary,_Queen_of_Scots_(1542-87)_-_Google_Art_Project.jpg [Accessed November 29, 2022].

354 Precisely quantifying the number of possible key mappings for a homophonic cipher is beyond the scope of this paper. But for comparison, the number of possible keys for a monoalphabetic cipher with the 26 letters of the alphabet is 26! ≈ 4 × 10²⁶. Brute-force search for such a cipher is already beyond the power of all the world’s computers combined. The number of possible key mappings for homophonic ciphers is significantly larger than that.

355 Hill climbing - Wikipedia [Accessed December 9, 2022].

356 Several French texts from the Gutenberg project from the 16th and 17th centuries were used for that purpose.

357 Maxima and minima - Wikipedia [Accessed December 9, 2022].

358 Simulated annealing - Wikipedia [Accessed December 9, 2022].

359 TNA SP53/23/38. Although this bears the date "1577", it was used not only in 1577 (Cotton MS, Caligula, C III f.535-540), but also in a letter from Mary to Beaton, 10 September 1582. The letter, with a note from Walsingham to John Somer dated 24 October 1582, is reproduced in Labanoff (1844, v 5, 308). Despite the annotation in the catalog as "probably 1572", the content (the Ruthven Raid) matches 1582.

360 The symbol for R in the Mary-Beaton cipher is a slight variation of the symbol for C in the Mary-Castelnau.

361 Bossy (2001, 180–183).

Appendix A.

The codebreaking algorithm

The principle behind the codebreaking algorithm is mapping the problem of recovering the key of a cipher only from ciphertext into an optimization problem. To do so, a score is assigned to possible solutions (in this case, possible cipher keys) and a search is performed to find the optimal solution, i.e., the solution with the best score, which is likely to also be the correct solution, that is, the original key used to encipher letters of the alphabet.

One straightforward approach to search for an optimal solution is to test all possible solutions, i.e., a brute-force search, but this is not practical if there are too many options to check. In the case of a homophonic cipher, the number of possible solutions (or possible keys) is simply too large for a brute-force search.³⁵⁴

Stochastic search is a widely used family of techniques to search for the optimal solution more efficiently than with brute force. Hillclimbing is a simple and commonly-used stochastic search technique.³⁵⁵ To score candidate solutions – in this case, a possible key mapping of the homophones to alphabet letters, the ciphertext is first deciphered using the candidate key, and a score measuring the quality of the deciphered is computed as follows:

• Before starting the search:

  •    ○ Compute the relative frequencies F_g for every combination of five successive letters of the alphabet, a.k.a. 5-gram or pentagram, such as “ISION”, “EMENT”, “ETLES”, “OURLE”, that appear in plaintexts, after removing spaces and punctuation, based a reference corpus of texts in the language assumed to be the language of the original plaintext.³⁵⁶

• During the search, evaluate a tentative decipherment using a candidate key as follows:

  •    ○ Count the number of occurrences N_g of each 5-gram g in the decrypted text.

  •    ○ Count the number of occurrences N_c of each letter c in the decrypted text.

  •     ○ The score S for the tentative decipherment is computed as follows: $$S=\sum _g{N}_g\hspace{0.17em}\log \hspace{0.17em}{F}_g/\sum _c{N}_c^2$$

Hillclimbing starts by generating a random key, i.e., a random mapping of the homophones to the letters of the alphabet. It iteratively performs small changes in the key, each time deciphering the ciphertext with the modified key, scoring the resulting deciphered text, and if the score is better, the changed key is retained. If the score is worse, the change is discarded. There are two types of small changes performed during hillclimbing, swapping any two symbols, and reassigning any of the symbols, as illustrated in Figures A18 and A19.

Figure A18. Swapping two homophones.

Figure A19. Reassigning a homophone.

The algorithm continues to test various small changes in the key until it is no longer possible to obtain a better key (i.e., with a better score) via such small changes, in which case it has reached a maximum point. With most optimization problems, there might be multiple such maxima, only one of them being the correct solution (the global maximum) and the rest being wrong solutions, called local maxima.³⁵⁷ To overcome the problem of getting stuck in local maxima with hillclimbing, the algorithm could restart the whole process multiple times, each time with a different random starting solution. The hillclimbing process is illustrated in Figure A20.

Figure A20. Hillclimbing.

Instead, the algorithm employs a variation of hillclimbing, called simulated annealing,³⁵⁸ which addresses the problem of getting stuck in local maxima by allowing for some downslope changes that result in a worse score, as illustrated in Figure A21.

Figure A21. Simulated annealing.

The codebreaking algorithm only recovers the mapping of the homophone symbols, representing letters of the alphabets. Nomenclature symbols representing common words, parts of words, names, or places need to be recovered manually.

Appendix B.

Enciphering errors and cross-cipher contamination

In any historical ciphertext, random enciphering errors are likely to occur, such as a symbol omitted, an extra symbol added, or a wrong symbol used instead of the correct one. We were expecting to find such sporadic errors in the ciphertext documents we deciphered, and this was the case with some of them. But in other documents, the percentage of errors was relatively high (up to a few %), making their decipherment and interpretation more challenging. Furthermore, an in-depth examination showed that many of those errors were systematic, i.e., the same wrong symbol being consistently used to encipher the same letter of the alphabet.

We first considered errors due to visual similarity, like using a symbol without a dot instead of the symbol with the dot. An example of such an error is given in Section 4.6. Still, those types of errors accounted only for a small percentage of the systematic enciphering errors. While examining other ciphers employed by Mary, Queen of Scots, we observed that the reconstructed Mary-Castelnau cipher shares several features such as diacritics, homophone symbols, and most of its vocabulary with another homophonic cipher, which we call the Mary-Beaton cipher, used by Mary to communicate with James Beaton, the Archbishop of Glasgow and her ambassador to France.³⁵⁹ The cipher table for the Mary-Beaton cipher is shown in Figure B22.

Figure B22. The Mary-Beaton cipher – 1577 (Source: TNA SP53/23/38).

Knowing that letters from Mary to Castelnau were often sent together with enclosed letters addressed to Beaton, we hypothesized that systematic errors might have occurred when a secretary enciphered multiple letters using different ciphers, a phenomenon we call cross-cipher contamination. To illustrate the phenomenon, in Figure B23 we show counts of the most frequent errors, found in a subset of 17 deciphered documents. Some of those errors occur dozens of times, sometimes within the same document, supporting the hypothesis that those errors were systematic, rather than random. More importantly, several of those errors, highlighted in Figure B23, can be explained by the secretary erroneously using the graphical symbol representing the desired letter of the alphabet (c, i, e, and r) in the Mary-Beaton cipher instead of the correct symbol in the Mary-Castelnau cipher.³⁶⁰ With other frequent and systematic errors (like the first entry in Figure B23), other unidentified ciphers may have been involved.

Figure B23. Recurrent enciphering errors.

We formulated a hypothesis to explain why cross-cipher contamination errors occur while enciphering multiple letters using different ciphers, as was the case with the newly deciphered letters from Mary to Castelnau, which were often sent with enclosed letters to other recipients, encrypted with different ciphers.

When a cipher secretary starts using a new cipher he is not yet familiar with, it is likely that at first, he will be looking up the correct graphic symbol(s) for each letter of the alphabet he wants to encipher, in the cipher table. But after a while, the secretary is likely to start memorizing symbols representing the most frequent letters of the alphabet so that enciphering becomes faster, without always having to consult the cipher table. If the same secretary needs to encipher another letter on the same day with a different cipher, and if he continues to rely on his memory rather than on the cipher table to encipher the most frequent letters of the alphabet (e.g., i or e), he may subconsciously recall the symbol for that letter from the wrong cipher table, which he may have used recently to encipher another document. This theory is also consistent with such errors being systematic, and since the secretary might not be aware of them, he repeats them throughout the document he is currently enciphering. Only after we understood this interesting phenomenon could numerous obscure passages in the documents from Mary to Castelnau be finally be deciphered and correctly interpreted, by also looking at the meaning of some symbols in the Mary-Beaton code.

Interestingly, there are multiple places in the documents we deciphered, where the secretary was able to detect those types of errors while enciphering the letter and correct them on the spot. The secretary made those corrections either by using the deletion symbol , by crossing out the wrong symbol, or by overwriting it. An interesting example illustrating the three correction methods appears on the second line of F26, as shown in Figure B24, on an encipherment of the text: LARECEPTIONDEVOZDERN (the parts wrongly enciphered and later corrected are highlighted in bold). Such historical corrections constitute additional evidence in favor of the cross-cipher contamination hypothesis.

Figure B24. Cross-cipher contamination errors corrected by the secretary – Second line of F26 (Source for archive image: gallica.bnf.fr/BnF fr. 2988 f.26).

Appendix C.

Updated list of letters from Mary to Castelnau

In Figure C25 and Figure C26, we list all the previously known letters from Mary to Castelnau, together with the enciphered letters we discovered. Letters originally sent in cipher are highlighted (items #61, #68, and #72 are believed to have been sent in cipher, based on their contents, but a ciphertext copy was not found in archives). The last column is the item number in Bossy’s list.³⁶¹

Figure C25. Updated list of letters from Mary to Castelnau – Part 1.

Figure C26. Updated list of letters from Mary to Castelnau – Part 2.