Prey to a Privateer Two Portuguese Ciphertexts from 1649

HistoCrypt 2022 Proceedings of the 5th International Conference on Historical Cryptology 20-22 June 2022 Amsterdam Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 Editors Carola Dahlke and Beáta Megyesi Published by: NEALT Proceedings series 49 Linköping University Electronic Press, Sweden Linköping Electronic Conference Proceedings, No. 188 ISBN 978-91-7929-397-0 (PDF) URL https://doi.org/10.3384/ecp188 ISSN 1650-3686 (print) ISSN 1650-3740 (online) Contents Preface ………………………………………………………………………………………....v Eugen Antal and Jakub Mírka……………………………………………………………………………………..……......1 Wrong Design of Cipher Keys: Analysis of Historical Cipher Keys from the Hessisches Staatsarchiv Marburg Used in the Thirty Years’ War Richard Bean, Sarah Lang and Megan Piorko………………………….……………………………………….....12 Solving an alchemical cipher in a shared notebook of John and Arthur Dee Paolo Bonavoglia………………………………………………………………………………………...........................22 Venetian cryptanalysis treatises of the Renaissance Paolo Bonavoglia………………………………………………………………………………………...........................32 The Enigma of Franceschi’s Falso Scontro Mariusz Borowiak and Marek Grajek….………………………………….……………………………………….....42 Enigma History and an Unexpected Treasure Trove Jörgen Dinnissen and Hugo Araújo………………………………………….……………………………………..……50 Prey to a Privateer. Two Portuguese Ciphertexts from 1649 Magnus Ekhall and Fredrik Hallenberg……..…………………………….……………………………………..……72 Analysis of the US Navy Cryptanalytic Bombe Schematics and Simulation of Selected Circuits Nino Fürthauer, Vasily Mikhalev, Nils Kopal, Bernhard Esslinger, Harald Lampesberger, and Eckehard Hermann……………………………………………………………..….…………………………………….82 Evaluating Deep Learning Techniques for Known-Plaintext Attacks on the Complete Columnar Transposition Cipher Joachim von zur Gathen………………………..…………………………….………………………………………..……91 Unicity Distance of the Zodiac-340 Cipher Rian Hagebeuk and Katherine Mueller……………………..…………….……………………………………..…101 “The Subtelty of Witches”: A Reformation Era Cipher Mystery Mihály Héder and Beáta Megyesi…….…………………………………….……………………………………..…111 The DECODE Database of Historical Ciphers and Keys: Version 2 Florentijn van Kampen……………………………..…………………………….……………………………………..…115 Thomson’s Telegram Decrypting a Secret Message from Albania, 1914 Nils Kopál and Bernhard Esslinger…….…………………………………….……………………………………..…127 New Ciphers and Cryptanalysis Components in CrypTool 2 IX Benedek Láng…………………………………………..…………………………….……………………………………..….137 Colonnele Frank's Indecipherable Chiffre George Lasry and Paolo Bonavoglia………....…………………………….……………………………………..…142 Deciphering a Short Papal Cipher from 1721 George Lasry…………………………………………..…………………………….……………………………………..……147 Deciphering a letter from the French Wars of Religion Giacomo Magnifico, Beáta Megyesi, Mohamed Ali Souibgui, Jialuo Chen, and Alicia Fornés ……………………………………………………………………………………………………………………..153 Lost in Transcription of Graphic Signs in Ciphers Beáta Megyesi, Crina Tudor, Benedek Láng, Anna Lehofer, Nils Kopal, and Michelle Waldispühl………….……………………………….…..…………………………….……………………………………..…159 What Was Encoded in Historical Cipher Keys in the Early Modern Era? Samu Potka and Mikko Kiviharju……………...…………………………….……………………………………..…168 Telegrams on Negotiations on Capturing Petrograd, 191 Clément Poupard…………………………………..…………………………….……………………………………..……172 Mundane cryptography: Toward a cultural history of cryptography Winfried Stephan…………………………………..…………………………….……………………………………..……181 Use of T-310 Encryption During German Reunification 1990 Gerhard Straßer………………………………………………………………………………………………………………..191 Historical View of Signs and Sign Languages as a Potential for Secret Communication in Two Worlds: In Ottoman Courts and Catholic Religious Orders Ferenc Szigeti and Mihály Héder…………………..…..………………….……………………………………..……208 The TRANSCRIPT tool for Historical Ciphers by the DECRYPT project X Prey to a Privateer. Two Portuguese Ciphertexts from 1649 Jörgen Dinnissen Historian, The Netherlands [email protected] Abstract The deciphered ciphertexts from the former governor-general of Brazil, António Teles da Silva, reveal two letters from him to the king of Portugal from 1649, while he was a prisoner. The letters were written in Portuguese and found on a Luso-Brazilian ship captured by a Dutch privateer, from the Brasilse Directie Middelburg, before the coast of Brazil. These letters were already known as plaintext letters in the National Archives, The Hague. We were able to relate the ciphertexts to the plaintexts and translations. This paper shows that the revealed strategic military information and state secrets made no impact on the superiors of the Dutch company, the Gentlemen XIX, in The Netherlands. It also shows that ciphers were provided from 1638 onwards in Portuguese Brazil. Based on the surviving sources, ciphertexts were used by the Portuguese mainly after the Dutch structurally threatened their communication. They became frequent in short-distance communication (on land) after the revolt of the sugar planters in 1645. In long-distance communication, over the seas, after the Dutch privateers compromised the Portuguese shipping between 1647 and 1648. 1 Introduction The 167 ciphertext is a four manuscript pages of the Dutch West India Company (WIC) located at the National Archives, The Hague, The 1 The dates in both manuscripts are wrong in the archival record, probably a misread by the archivist. Hugo Araújo Federal University of Santa Maria, Brazil [email protected] Netherlands (Teles da Silva, 1649a1). The description of this manuscript reads: “Ciphercode (Dutch: cijfercode) from Antonio Telles da Silva, with key, 1649-3-28.” 1 The 168 ciphertext is also four manuscript pages found in the same archive (Teles da Silva, 1649b1). Its description reads: “Ciphercode from Antonio Telles da Silva, with key, 1649-12-3.” Figure 1: Part of the key from the 167 cipher. Figure 2: Part of the key from the 168 cipher. We made a digital transcription of all pages manually. This transcription was put into the software CrypTool 2 (CT2). Using the digitized nomenclature, CT2 generated the revealed plaintext used for this paper. Table 1 shows the very first word of the two documents in ciphertext, transcription, revealed plaintext in Portuguese, and its translation into English. Month of 167 must read 5 (not 3), and month of 168 must read 6 (not 12). Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 50 2 Table 1: First word of the two ciphertexts deciphered. Next to the ciphertexts, we found the Portuguese plaintexts and translations into Dutch. We have identified information in other sources about the ship that carried these ciphertexts and the frigate that seized it. At last, we found that a person with knowledge of ciphers (Dutch: persoon daer van kennijse hebbende) was able to decrypt the ciphertexts through cryptanalysis and, after that, translated them. Unfortunately, the source does not mention the cryptanalyst’s name. The ciphertexts, the revealed plaintexts, and the translations were sent in one package – together with other letters, reports, and an accompanying letterbook – by the government of Dutch Brazil, the Hoge Raad Brasil (Hoge Raad) to their company superiors, the Gentlemen XIX. The paper will analyze these ciphertexts, their cipher, relating their decrypting and translation into Dutch to the context of the war in Brazil. We will present how the obtaining and decryption of the ciphertexts was part of the administrative practices of the WIC. In particular, we will highlight the use of Jewish scholars who acted as decrypters for the Dutch. This paper is structured as follows: Section 2 contains a summary of the letters hidden in the ciphertexts and background information. Section 3 shows the complete trail of these letters within the Dutch administrative process. Section 4 discusses the importance of ciphertexts and how often they were used in Brazil. The technical matters of the used ciphers are discussed in Section 5. Finally, Section 6, presents five conclusions about these sources and their use. Content of the Two Letters There is a hidden letter in each of the two ciphertexts that relate to a decisive period for both the Portuguese and Dutch. At that time, they fought in the northeast of Brazil for the white gold of the 17th century: sugar. 2.1 Portuguese and Dutch Fighting for Sugar in Brazil In 1630 the WIC conquered Recife and Olinda from the Portuguese General Government of Brazil. The Dutch occupied the northeastern region of Brazil for 24 years (1630-1654), including the heart of the sugar plantations in South America, the captaincy of Pernambuco.2 Things went rather smoothly for the Dutch until the sugar planters of Pernambuco revolted in 1645. A part of the inhabitants and the Portuguese administration followed suit. The Dutch were unable to win the battle against the “rebels” decisively, and because of that, the military offensive actions and defensive measures were costly. Next to that, the backup from The Netherlands was inadequate. The revolt of 1645 made the Dutch very suspicious of the Portuguese diplomatic movements. In 1641 both nations signed a ten years truce, which both sides ignored in the following years. In 1647, the Portuguese king authorized his ambassador in the Netherlands to negotiate two proposals: an offer to buy the territories of Dutch Brazil or, in case they refused, a proposal to pardon all the rebels, granting them a safe retreat to Bahia. The negotiations did not advance because the Portuguese and the Dutch were suspicious of each other's actions. On the one hand, the Dutch suspected the meaning behind the offer was to avoid or delay sending of military aid to Dutch Brazil. On the other hand, the Portuguese were always on alert and suspected new Dutch invasions in the Atlantic. 2 The background information of Subsection 2.1 is based on Heijer den (2002), Knaap et al. (2015), and Roos (1992). Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 51 vessels on the anchorage of Portugal, Madeira, or the Azores. Article 7 indicates that all ships and goods seized should be delivered to the hands of the Hoge Raad. Three days after unloading, the Hoge Raad had to sell the prize ship and its cargo in cash and in public.4 In 1647 and 1648, the privateers (Dutch: kaapvaarders) captured, the vast majority by the Directie Middelburg, an estimated 220 Portuguese ships in the Atlantic. Half of these ships were bound for Portugal, negatively affecting the Portuguese maritime trade. Next to that, it also affected communication between the king of Portugal and the Brazilian part of his empire. As a countermeasure, from 1650 onwards, the Portuguese sugar ships sailed from Brazil in convoy and accompanied by 18 warships. The king also ordered Teles de Menezes (1647) that all relevant letters between the Crown and the general-government had to be encrypted. Figure 3: Map of the Atlantic ocean. Salvador, Bahia (1, Portuguese), Recife (2, Dutch), Alagoas (3) where Nossa Senhora de Remédio was captured. The hunting grounds of the Directie Middelburg is shaded (/).3 2.2 Brasilse Directie Middelburg The two ciphertexts taken on 1649-6-18 near Alagoas were carried by the ship Nossa Senhora de Remédio, who was not sailing under convoy. An accidental prey to a privateer on the watch. See Figure 3 for the location of Bahia (Portuguese), Recife (Dutch), and the hunting grounds of the Directie Middelburg. A new company was founded next to the WIC in December 1646, the Brasilse Directie Middelburg. The government in The Netherlands granted them the right, under strict conditions (Dutch: provisionele conditien), to capture enemy ships in time of war in the seas between Portugal, island Madeira, the Azores, and the coasts of Brazil and Africa. In return, the Directie Middelburg had to pay a share of their profits (Dutch: recognitiegelden) to the monopolist WIC. Article 1 of provisionele conditien defines that the privateers must make a report informing the name of the seized ship and its captain, a detailed list of guns, and crew of the prize ship. There is no mention of keeping and handing over letters and packing lists of the seized ships. Article 2 defines that the privateers should not attack Portuguese António Teles da Silva (1590-1650), governorgeneral of Brazil from 1642-1647, wrote the two ciphertexts while he was a prisoner. The king of Portugal ordered his imprisonment because he supported the revolters of Pernambuco in 1645. The king did not dare to openly support the rebels of Pernambuco because he was afraid of retaliation from the Dutch. But in secret, the Crown gave support to Teles da Silva, enforcing that officers had to follow his orders in this matter. When this became public in 1647, the king felt compelled to imprison Teles da Silva to appease 3 4 It is unclear how far towards North America this area ran. Provisionele Conditien (1646) does not give a definitive answer on this. That’s why question marks are put. 2.3 Prisoner and Former Governor-General of Brazil António Teles da Silva Provisionele Conditien, 1646. Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 52 the Dutch. 5 A few years later, in 1650, Teles da Silva died in a shipwreck caused by a storm near the coast of Portugal, dying without receiving a Royal pardon. 2.4 The Two Ciphertexts The ciphertexts contain two letters written in Portuguese. Teles da Silva wrote both letters from the capital of Brazil, the city of Salvador, in the captaincy of Bahia. The letters were addressed to the king of Portugal in Lisbon.6 See Table 2 for more details about the two letters: sender, receiver, and decrypter. Meneses, stated to Teles da Silva that he would be happy if the Dutch captured him and took revenge because it would be instrumental for peace with Portugal. This letter also accused Teles de Meneses of neglecting the defenses, mismanaging the finances, and only be concerned with his private business. In the cleartext, after the ciphertext, Teles da Silva complains to be forgotten and badly thanked by the king after being a prisoner for a year and a half, “becoming all grey and very sick”. 7 See Appendix 1, for the revealed plaintext of CT2 and Portuguese compared and a translation of the text into English. 2.6 Letter Dated 1649-6-3 from Teles da Silva to King Summary of the content (Teles da Silva, 1649b1): Teles da Silva informs that the soldiers were rioting and threatened to kill the Portuguese commanders Francisco de Figueiroa and Francisco Barreto because they did not receive proper rations and clothes. Pressured by the soldiers, Barreto paid the wages and provided clothes, becoming in debt to preserve the command of the troops. Table 2: The two ciphertexts. We don’t know who encrypted these two letters for Teles da Silva. A good candidate could be Bernardo Vieira Ravasco, who was the secretary of state of Brazil in 1649 and the right hand of the governor-general Teles de Meneses. On 1650-7-10, Teles da Silva (1650a) wrote a letter of recommendation for Ravasco to the king of Portugal. In this letter, he emphasizes the ability of Ravasco to decipher the king’s ciphertexts. 2.5 Also, Teles da Silva declared to receive information about the feeble state of Dutch defenses, indicating that they could leave the forts in three months if they did not receive help from The Netherlands. He claimed to know that four hundred “useless Dutch” – injured soldiers and people coming from Angola – sailed to their homeland; he also received news that many Jews returned to The Netherlands with their wives. See Appendix 2, for the revealed plaintext of CT2 and Portuguese compared and a translation of the text into English. Letter Dated 1649-5-28 from Teles da Silva to King Summary of the content (Teles da Silva, 1649a1): The new governor-general, António Teles de 5 In 1647 two pamphlets revealed that the king and his army supported the rebels of Pernambuco: Kort Discours Rebellye (1647) and Claar Vertooch (1647). 6 Notice, that Teles da Silva writes his own name ‘Telles’ with ‘ll’ and that ‘Silvas’ ends with ‘s’. 7 Teles da Silva (1649b2) mentions: “twenty months”. Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 53 3 the dispatched to The Netherlands, with a copy kept at the office (Dutch: ten contoire alhier) in Brazil. This section shows the flow of information within the administrative process of the WIC. The daily minutes of the Hoge Raad (1649d) registered on 1649-6-29 that the WIC bought, from the captured Nossa Senhora de Remédio, 20 rolls of tobacco. There is no mention of the remainder, 180 rolls of tobacco and 150 crates of sugar. Probably it was sold, together with the ship, in Middelburg, The Netherlands. 3.1 The Two Letters within Administrative Process Flow of Information in Dutch Brazil On 1649-6-18, captain Bastiaen Thuijnman interrogated the skipper Pasquael Coelho and navigator Francisco Anes of the captured caravelle Nossa Senhora de Remédio (Thuijnman, 1649).8 They declared that they left from Bahia with 18 crew members (Dutch: koppen) with 150 crates of sugar and about 200 rolls of tobacco. They delivered all the letters they had on board to captain Thuijnman, and claimed not to be carrying letters from the governor-general to the king. On the next day, 1649-6-19, captain Thuijnman from frigate (Dutch: fregat) De Brack, from the Directie Middelburg, reported to the Hoge Raad that he seized the Nossa Senhora de Remédio near Alagoas (Hoge Raad, 1649a). The Hoge Raad gave orders to interrogate the captured skipper and navigator and to examine and translate the letters. After that, the Hoge Raad ordered to dispatch these letters with the first ship sailing to The Netherlands since its content was considered worthy of attention. On 1649-6-21, the Hoge Raad (1649b) registered in the daily minutes that, after reading the translations of the letters captured by Thuijnman, the content was considered relevant enough to be copied by the registry (Dutch: griffie) and had to send with other reports to The Netherlands in the first ship that departs. The daily minutes of the Hoge Raad (1649c) registered on 1649-6-23 that they had read a translated report (Dutch: missive) found on the latest captured ship by captain Thuijnman. This report presented news from Angola since the Portuguese reoccupied it. This report also was Dutch spelling. ‘Pasquael’ is probably the misheard Portuguese name ‘Pascoal’. The ship has the name of 8 On 1649-7-23, the Hoge Raad (1649e) wrote to the Gentlemen XIX their periodic report (Dutch: generale missieve). In this report, under letter C, they listed the letters captured on the ship Nossa Senhora de Remédio. Some were in ciphertext (Dutch: in cijffer geschreven) and were deciphered (Dutch: ontcijfferen) by an unnamed person described as skilled in this task. The WIC officials translated the deciphered letters and the interrogation of the Portuguese prisoners into Dutch. The Letterbook (1649) from the same day, 1649-7-23, registered under letter C, an original report found on a captured ship by captain Thuijnman and its translation. This report was written by the former governor-general Teles da Silva to the king of Portugal. The letters from Teles da Silva dated from 1649-5-28 and 1649-63 are in ciphertext (Dutch: ciffer letters). Next to these, the Portuguese revealed plaintexts, and their translations into Dutch are mentioned with their folio numbers. We do not know whether the Hoge Raad themselves did anything – next to sending the reports to their superiors in The Netherlands and keeping copies of them – with the information contained in the two letters from Teles da Silva. There is no trace of this in the daily minutes nor their periodic reports. There is also no indication whether they passed on the information to Admiral Witte de With or whether it affected his military policy (With de, 1649). the catholic saint ‘Our Lady of Remedies’. In Portuguese it should read: Nossa Senhora dos Remédios. Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 54 3.2 Flow of Information between Brazil and The Netherlands received in 1638 a cipher to use in transatlantic communication.11 The package with letters, including ciphertexts 167 and 168, arrived in The Netherlands on 16499-20 (Letterbook, 1649). What did the Gentlemen XIX do with this information? Did it change their policy? We don’t know if they talked about it in detail because there are no preserved (secret) minutes of this period. Another known example of Portuguese ciphers in this context is the report presented by Abraham de Pina (1646) to the Hoge Raad. De Pina explains the process he used to decipher several letters apprehended by the Dutch. João Vieira de Alagoas made the ciphertexts while he was a spy in Recife, pretending loyalty to the Dutch. 12 He ordered a "mullato" to deliver the ciphertexts to a rebel contact in the outskirts of Recife, but instead, he handed the letters directly over to the Hoge Raad. This event caused the imprisonment and torture of Vieira, who refused to deliver the key. Faced with the revealed content of the four deciphered letters, Vieira confessed that he had written the ciphers addressed to the rebels. He was tried and sentenced to the death penalty in a public execution. On 1649-10-1, the Gentlemen XIX (1649) sent orders (Dutch: missiven) to the Hoge Raad, replying to their report of 1649-7-23. There is no mention of Teles da Silva or his ciphertexts. The Gentlemen XIX informed to be raising money from the government (Dutch: Regeringe) of The Netherlands to continue the conquest. They also declared the sending of ships, warships, and victuals to Brazil. There is a complete silence about the information received about the Portuguese. They must have regarded it as daily operational frictions among the Portuguese. This information had no impact on their strategy. 4 How Often Were Ciphers Used in the Portuguese Empire? This section indicates when and where the use of ciphers occurred in the Portuguese Empire. 4.1 Use of Ciphers in Portuguese History There is evidence of encrypted correspondences in Portuguese relations with the Ottoman Empire, dating from the 1550s.9 4.2 Use of Ciphers in Portuguese Brazil 1645-1654 The archives of Portugal and Brazil do not have many ciphers concerning the wars in Brazil, specifically between 1645 and 1654. The only ones found so far are in the Dutch archives: De Pina’s report (1646) and António Teles da Silva’s (1649a1 and 1649b1) ciphertexts. See Table 3 for the ciphertexts that we found and records mentioning them. In Appendix 4, we provide more details about these sources. In a diplomatic context, the Portuguese ambassador in France exchanged ciphertexts with the king of Portugal between 1646 and 1649, the same period of the war in Brazil.10 Regarding the wars with the Dutch in Brazil, the governor-general Conde da Torre, who was in charge of an Armada sent to recover Pernambuco, 9 Couto, 2012. Coelho, 1942. 11 Torre, 1638. 12 The daily minutes from 1646-5-16 (Hoge Raad, 1646) do not mention De Pina directly. They talk about 10 Table 3: Ciphertexts in Dutch and Portuguese sources. “a certain person from the jewish nation” (Dutch: seker persoon van de joodse natie). In future works, we will analyze the João Vieira de Alagoas’ ciphertexts described in De Pina’s report (1646). Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 55 We analyzed all 52 letterbooks (Dutch: overgezonden lijste der brieven of papieren) from the Hoge Raad between 1645 and 1654 to the Gentlemen XIX, chamber Zeeland, government, or simply to The Netherlands. In nineteen letterbooks (37%), we found mentions of intercepted letters, and only a couple of them mention ciphertexts. See Appendix 3 for the complete list of letterbooks and their content. Based on the ciphertexts found and mentioned, we draw the following conclusions for Portuguese Brazil. The use of ciphers in long-distance communication became frequent after the heavy losses that Dutch privateers inflicted in 16471648 to the Portuguese shipping on the route from Brazil to Portugal. The use of ciphers inside Brazil became frequent after the revolt of 1645. Luso-Brazilian spies sent ciphertexts to inform the rebel army about the state of fortresses and the WIC army. Therefore, these ciphertexts circulated in shortdistance communication. In this sense, the use of cryptography by the Portuguese started after the Dutch structurally threatened their strategic military information and state secrets. On the other hand, this is also why the Dutch did not use cryptography in Brazil because they exerted control of the coastal towns and naval superiority in the South Atlantic. The Dutch control of the seas – they were king of the seas – assured the continuity of long-distance communication. 5 Technical Analysis Cipher 167 is a monoalphabetic substitution cipher 13 that replaces every plaintext letter with the same symbol of the ciphertext alphabet. Figure 1, in Section 1, shows the first four letters. Of the 26 letters of the Latin alphabet, the cipher used 20. The letters J, K, V, W, X, and Y are missing, with 13 Cipher 168 is a homophone substitution cipher 14. In this ciphertext, every plaintext letter is replaced with the same two to ten symbols of the ciphertext alphabet. Figure 2, in Section 1, shows the first four letters. Of the 26 letters of the Latin alphabet, the cipher used 21. The letters J, K, V, W, and Y are missing, with no use of other symbols. For the reconstruction of 15 codes, see Subsection 5.4. Table 4: Symbols used in cipher 167 and 168. Ciphertexts and Keys This section shows technical details about the used ciphertexts and their keys. 5.1 no use of other symbols. See Table 4 for more details. Kopal, 2018: 30. Table 5: Encrypting strategies not in key. Of both ciphertexts, we have the original ciphertext (Teles da Silva, 1649a1 and 1649b1), revealed plaintext in Portuguese (Teles da Silva, 1649a2 and 1649b2) and translation into Dutch (Teles da Silva, 1649a3 and 1649b3). We compared the revealed plaintext from CT2 with the revealed plaintext in Portuguese. We also compared the Portuguese and Dutch texts with 14 Kopal, 2019: 108. Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 56 English as their bridge language. See Appendix 1 and 2 for the details. codes (in different colors) used over the letters. This strategy made cryptanalysis harder. Striking is that the key of cipher 168 is more complex than that of cipher 167, once they were made only six days apart. What conclusion should we draw from this? 5.2 Table 6: The number-codes (in different colors) of cipher 168 are evenly distributed over the letters to make cryptanalysis harder. Our analysis shows that the translations from Portuguese into Dutch are done correctly, with minor differences. The translator did not miss anything essential. Our analysis also shows the use of a couple of encrypting strategies to make cryptanalysis more difficult. See the Table 5 for the complete list. For example: sometimes they deliberately do not use spaces; in general, they do not use accents and often use different spelling and abbreviations. On the other hand, one strategy made cryptanalysis easier: using bigger written numbers for capitals letters. For making ciphertext 168, the encrypter probably used a separate piece of paper on which he kept track of how often he used codes. If we look at Table 6, we see the even distribution of the Jewish Cryptology Specialists in Dutch Brazil Who did the cryptanalysis and was thus able to decipher the two ciphertexts? Based on the found and mentioned ciphertexts, see Table 3, we have two suitable candidates: Moyses Raphael Dagilaer and Abraham de Pina. Both were Jews of Iberian origin who lived in Recife after the Dutch conquest and acted as rabbis. They served as cryptanalysts for the Hoge Raad on previous occasions. Jütte (2015) demonstrates how Jews served as cryptanalysts in Europe, especially emphasizing the role of scholars in the use of cryptography. Therefore, we suggest that these particular Jews acquired the knowledge of cryptanalysis before they came to Dutch Brazil. We haven't found out yet how these men learned how to decipher. Some factors could help us get closer to understanding this situation. The proximity they had with the Portuguese through the shared language and culture could be one of the elements. Also, the social function of secrecy15 was used by these particular Jews. They employed the knowledge of cryptography as a way to improve their relations and their business with the Dutch in Brazil. Figure 4: CT2 substitution component deciphering the 168 ciphertext using the digitized key. 15 Jütte, 2015. Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 57 5.3 Has there been a Loss of Knowledge and Skills since 1640? Also striking is that neither the cipher 167 nor 168 contain nomenclature elements as the ones used by Conde da Torre (1638). See Figure 5. Figure 5: Part of the cipherkey that Torre (1638) used with nomenclature elements. A question emerges: did the cryptographic skills diminish from Portugal since it broke with Spain in 1640? Could this be the reason why the ciphers of 167 and 168 are simpler than the ones used by Conde da Torre? Future works should develop this question about the impact of the end of the Iberian Union on Portuguese cryptography. 5.4 Generating Plaintext with CrypTool 2 We have the keys of both ciphertexts, which appears in the manuscripts, and assumed that both were substitution ciphers. Therefore, we entered the key manually into CT2. With the help of the substitution component, we were able to decipher most parts of the ciphertexts correctly. The output of cipher 167 was almost flawless from the start. The cipher 168 was challenging since we had to reconstruct fifteen codes unaccounted in its key. Five of them were code mistakes made by the encrypter, and these mistakes were not repeated throughout the ciphertext. A mistake, for example, is the code 215 instead of 205 for letter E. We reconstructed ten of the codes because they fit into the used key-strategy ‘first number, +1, +2, and +5’. For example, the key of the letter M reads ‘3, 35’ but the reconstructed code reads ‘3, 31, 32 and 35’. The unaccounted code 32 fits into this strategy. See Appendix 2 for the complete list of unaccounted for codes of cipher 168. Based on this, we suggest that the mistakes indicate that this deciphering process was made without the original key, through cryptanalysis. After that, the decoder was able to fill the remaining codes. This conclusion probably was drawn from the obtained Portuguese plaintext and not through the final version of the key. CT2 provided great help with reconstruction step-by-step of a key with codes unaccounted. The software facilitated the recurrent uploading of the nomenclature and the visualization of the results in plaintext. Are all codes resolved? Are the words still meaningful? See Figure 4 for the ciphertext 168 in CT2, the nomenclature, and the revealed plaintext. 6 Conclusions Our main findings have been: 1) The plaintexts of the deciphered ciphertexts reveal two letters that, in retrospect, were already known in the National Archives as plaintext Portuguese letters. We relate them and their translation into Dutch held in the same archive. 2) After cryptanalysis, the decrypter revealed both ciphertexts. In the case of cipher 168, the key obtained was sufficient to read enough of the plaintext. Ten codes were unaccounted for in this key. The decrypter reconstructed those codes, as is shown in the revealed plaintext but didn’t process them in the final version of the key. 3) The encryption process used the key and additional steps to make cryptanalysis harder. For example, the code sometimes does not leave spaces between words and the use of abbreviations. But there was also one element that made cryptanalysis easier: using bigger written numbers for capitals letters. 4) It’s a fact that cipherkeys were provided in Portuguese Brazil, at least from 1638 onwards, for communicating about strategic military information or state secrets. Ciphers were used by the Portuguese mainly after the Dutch structurally threatened their communication. In short-distance communication, they became frequent after the revolt of the sugar planters in 1645. The use of ciphers in long-distance communication increased after the Dutch privateers compromised the Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 58 Portuguese shipping between 1647 and 1648, notably from 1649 onwards. 5) The Gentlemen XIX did not do anything with the strategic military information and state secrets found in the two deciphered letters of Teles da Silva. The orders sent by the Gentlemen XIX after receiving the information from Dutch Brazil show that they did not use the seized information. References Claar Vertooch. 1647. Anonymously, Claar Vertooch van de Verradersche en Vyantlijcke Acten en Proceduren van Poortugal, In 't Verwecken ende Stijven van de Rebellie ende Oorloghe in Brasil. Corea de Saa, 1649. 1649-7-4. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-171, under number 6: unfoiled. 1649-11-29. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-154, under letter T: unfoiled. Dagilaer, 1649. 1649-12-31 and 1650-1-18. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 74: unfoiled, scans 109 and 174. Daniel Jütte. 2015. The Age of Secrecy. Jews, Christians, and the Economy of Secrets, 1400-1800: 8-11, 26-27, 56-60. Dejanirah Couto. 2012. ‘Spying in the Ottoman Empire: sixteenth-century encrypted correspondence’. Correspondence and Cultural Exchange in Europe, 1400-1700. Cultural exchange in early modern Europe: 274-312. Dias, 1649. 1649-11-28 and 1649-12-03. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 73: unfoiled, scans 1260-1262 and 1288-1291. Doeke Roos. 1992. Zeeuwen en de Westindische Compagnie: 69-77. Gerrit Knaap, Henk den Heijer and Michiel de Jong. 2015. Oorlogen Overzee: Militair optreden door compagnie en staat buiten Europa 1595-1814: 289295. Henk den Heijer. 2002. De geschiedenis van de WIC: 65-68. Hoge Raad, 1646. 1646-5-16, 1646-5-19, and 1646-528. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 71: unfoiled, scans 420, 421, 426 and 437. Hoge Raad, 1649a. 1649-6-19. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 73: unfoiled, scans 680-681. Hoge Raad, 1649b. 1649-6-21. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 73: unfoiled, scan 681. Hoge Raad, 1649c. 1649-6-23. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 73: unfoiled, scans 688-689. Hoge Raad, 1649d. 1649-6-29. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 73: unfoiled, scans 712 and 771. Hoge Raad, 1649e. 1649-7-23. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-105: ff240-241. Kort Discours Rebellye. 1647. Journael ofte kort Discours nopende de Rebellye ende verradelijcke Desseynen der Portugesen, alhier in Brasil voorgenomen, 't welck in Junio 1645 is ontdeckt. Gentlemen XIX, 1649. 1649-10-1. NL-HaNA, OWIC, 1.05.01.01, kopieboeken van uitgaande stukken, 10 1646 juli 21 - 1657 okt. 10: ff64-68. Letterbook, 1649. 1649-7-23. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-107: unfoiled. Nils Kopal. 2018 ‘Solving classical ciphers with CrypTool 2’. Proceedings of the 1st International Conference on Historical Cryptology HistoCrypt 2018. No. 149: 29-38. Nils Kopal. 2019 ‘Cryptanalysis of homophonic substitution ciphers using simulated annealing with fixed temperature’. Proceedings of the 2nd International Conference on Historical Cryptology HistoCrypt 2019. No. 158: 107-116. Pina de, 1646. 1646-5-??. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 62-44: unfoiled. P.M. Laranjo Coelho. 1942. Cartas de El-Rei D. João IV Ao Conde da Vidigueira (Marquês de Niza) Embaixador em França. Provisionele Conditien, 1646. Zeeuws Archief, Middelburg (MdbZA), 1.4.10.1., 2118. Letter from 1646-12-06, unfoiled, scan 366-370. Teles da Silva, 1649a1. 1649-5-28. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-167: ff146-147. Teles da Silva, 1649a2. 1649-5-28. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-88: f140. Teles da Silva, 1649a3. 1649-5-28. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-59: f139. Teles da Silva, 1649b1. 1649-6-3. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-168: ff142-143. Teles da Silva, 1649b2. 1649-6-3. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-89: f145. Teles da Silva, 1649b3. 1649-6-3. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-90: ff141. Teles da Silva, 1650a. 1650-7-10. AHU_ACL_CU_005-02, Cx. 14, D. 1702. Teles da Silva, 1650b. 1650-7-19. Virgínia Rau. 1984. ‘Fortunas Ultramarinas e a nobreza portuguesa no século XVII’. In: José Manuel Garcia. (Ed.) Estudos sobre história econômica e social do Antigo Regime: 36-46. Teles de Menezes, 1647. 1647-8-10. BNRJ-SM, 08,01,016. No 002. Thuijnman, 1649. 1649-6-18. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-97: ff154-155. Torre, 1638. João Paulo Salvado, Susana Münch Miranda. Cartas do Primeiro Conde da Torre (volume II). 2001: 32-33. Vieira, 1646. 1645-05-29. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 62-43: unfoiled. With de, 1649. 1649-4-24 to 1649-7-31. NL-HaNA, OWIC, 1.05.01.01, inv.nr. 65-69: ff209-226. Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 59 APPENDIX 1: Ciphertext 167 from 1649-5-28 RESOLVED PLAINTEXT CT2 (CIPHERTEXT) AND PORTUGUESE PLAINTEXT COMPARED INDEX COLORS. Grey = in ciphertext not in Portuguese; Yellow = in Portuguese not in ciphertext; Blue = differences in used letters e.g. ‘i’ versus ‘e’ (different written letters) or ’S’ versus ’s’ (capitals). Space is ‘_’. {l.iii} Sabendo os Olandezes por zer_selim16 {l.} que estaua prizioneiro nesta prasa e_o C {l.v} onde o soltou por seu gosto contradi {l.} zendo_lho todos por_quanto17 eu estaua {l.} para ir em huas naos ingrezas; se resol {l.} ueram a uir_me buscar com a sua Capi {l.} tania e seis naos dos estados, a esta ba {l.x} rra, onde andam ha dias. e dizendo_o ao C {l.} onde q nam conuinha dar este capitam o {l.} landes por muitas razoins apontandolhe que {l.} estaua eu para partir, e que o iria18 dizer e {l.} me ueriam esp_erar como uieram; respon {l.xv} deo o conde que se me tomasem seria muit[o] {l.} bom porque se uingariam os olandezes de {l.} mim, e logo fariam pazes com portugal. o {l.} ……………………………………………… {l.} tomarem_me nam quis deus dar esse gosto {l.xx} ao conde, mas tememos q uenha o conde de {l.} castel_melhor dar em suas maos, epodendo_ha {l.} uer preuensois dos nauios que aqui estam {l.} para o dito conde castel_melhor ser so {l.} corrido; o governador se lhe nam da nada, antes {l.xxv} folgara. porque so atende a que este estado se {l.} perca como he uender tudo, dar lic {l.} enca a_todos que se uam, leuar todos os ar {l.} telheiros; tirar a_artelharia que estaua {l.} para as forteficasois; e todo o_dinhei {l.xxx} RIJKSARCHIEF’S-GRAVENHAGE {l.} ro que esta por uenser o_gasta de maneira {l.} que o nouo governador que uem, he forsa que tenha c {l.} om elle grandes desauensas, por mais sof {l.} rido que seia. acuda sua magde a_isto co {l.xxxv} mo lhe pareser melhor tomando as infor19 {l.} masois todas q lhe pareser- aarmada esta {l.} incapas20 para o que se quizer fazer delas. por21 {l.} que como sua magd_e mandou fazer pres ‘Zerselim’ or ‘Zer selim’ is not a Dutch word. It’s written with a Capital in the Portuguese plaintext. Probably a person's name, in this case of some prisoner of the Portuguese. 17 Ciphertext: 131. In Portuguese text: q. Hence code should have been: 13. 18 Ciphertext: ‘o iria’. Portuguese: ‘oiria’. Probably ‘oiria’ refers to a Spanish word that means ‘he has heard’, like Portuguese ‘ouvira’. 19 Ciphertext: ? (unreadable). In Portuguese text: r. 20 This sentence is not in the Portuguese plaintext. Mistake by the decrypter. 21 Ciphertext: ? (unreadable). In Portuguese text: r. 16 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 60 {l.} tes capitania e almirante, so delas22 {l.xxxx} [f.146v] {l.xxxxi} se trata, as mais nam dam cuida_do, e so o {l.} tem em meter muita fazenda na capi {l.} tania de pau de iacaranda misturado {l.} com pau de brazil e outras muitas {l.xxxxv} couzas que sam publicas se as quize {l.} rem saberO Seguinte estava por letra23 {l.} e_quando não com a_diser satisfaso. anno {l.} e meio ha, que estou prezo vendo todos as dias entregar me {l.} hum sargento a_outro, com_as ordens que_tenho escrito e_mandado {l.l} sertidois, e_quando me_veio tão matratado por caia causa {l.} estou todo branco e_muy_doente, não sinto meos tra_valhos {l.} senão perder sua Magde hum vasalo que tanto amava {l.} seu serviso, e_tão zeloso era dele, bem curo24 que isto he_lo {l.} cura mas esta durara em mim sempre, ainda que veija {l.lv} quam mal se_me agradese, pois vendo sua magde o_estado em_que {l.} ficava tão mal o_remediou. e_do_conde não tenho queixa, {l.} pois obrou com_forme seo ioiso, mas sua magde vendo {l.} tão bombaria25 com_sem_tila26 he_muito para a sen {l.} tir. {l.lx} muito tempo ha que_não tenho carta vossa, vindo algu {l.} mas en_barcasen_is27. de_presento veio, sebastião {l.} vas, e não tive carta vossa nem de_ninguem: todos nesta {l.} ocasião me largaram, assim o_em_xergei28 nas_poucas {l.} delinnsias29 que_se_fizeram com sua magde e_tenho eu {l.lxv} resposta do dito señor a_não tive vossa, nem de_mui {l.} tos que_escrevi que_sei que_são chegadas. como vos {l.} tenhais saúde, acostumado sou eu, certas_for {l.} tunas. de 11 de_abril temos novas de_essa sidade {l.} e,_como não avia novas da_par_tida de_castelo melhor, {l.lxx} ficaram as naos para o_anno que_vem, com_agla30 {l.} ficarei fasendo, oi[_]tres annos de_prizam, {l.}_147_[f.147r] {l.lxxiii} e_de_estar no_brazil, oito, e_o_mais_serto he ficar nello que he, {l.} o_que os_ministros pertendem, as_novas maes de_esta {l.lxxv} baia, não faltara quem_as_de, guarde vos_deos {l.} como_dezeio. baia em 28 de maio de [1]649 Ciphertext ‘des?’ (partly unreadable). Portuguese ‘delas’. This sentence is not in the ciphertext. Added by the decrypter. After this line the text is in Portuguese cleartext. 24 Ciphertext ‘curo’ (English: I heal). Portuguese ‘veijo’ which is the orthographically incorrect form of ‘vejo’ (English: ‘I see’). ‘I heal’ doesn’t make sense in this context. 25 Ciphertext ‘bombaria’ (English: a batch of bombs). Portuguese ‘zombaria’ (English: mocking). ‘A batch of bombs’ doesn’t make sense in this context. 26 Ciphertext ‘com sem tila’. Portuguese ‘consetila’. Both words are in English: to consent. 27 Ciphertext ‘en barcasen is’. Portuguese ‘embarcasois’. Orthographically incorrect form of ‘embarcações’ (English: vessels). 28 Ciphertext ‘emxergei’. Portuguese ‘enxerguei’. Both words are in English: ‘I can see’. 29 Ciphertext ‘delinnsias’. Portuguese ‘deligencias’ (English: proceedings or preparations for an event). Probably the encrypter made a mistake when he wrote ‘delinnsias’. 30 Ciphertext ‘agla’. This could be an abbrevation for Portuguese ‘aquela’ (English: that one). Portuguese ‘que’ (English: what) makes more sense because it’s about how much time Teles da Silva already has been in Brazil. 22 23 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 61 {l.} estava asinado Antonio Telles da Silvas PORTUGUESE AND DUTCH TEXTS COMPARED WITH ENGLISH AS BRIDGE LANGUAGE INDEX COLORS. Blue = differences in used words; Grey = in Dutch not in Portuguese; Yellow = in Portuguese not in Dutch. So the Dutch know by zerselim that I was a prisoner in this square (by hostage-taking and the Count released him for his pleasure contradicting everyone because I was here ready to leave with the English ships); they decided to come for me, to look for me, with his admiral and six ships of the States, to here, where they have been for days. And saying to the Count beforehand that it was not convenient to hand over me to this Dutch captain for many reasons, pointing out to him that I was about to leave, and that he himself would reveal and that they would observe my arrival just as they did; the Count replied to this that if the Dutch took me it would be very good because they would avenge the Dutch on me, and would hit peace with Portugal. But taking me, God did not want nor to give this pleasure to the Count, but we fear that the Count of Castelo Melhor will come into their hands, and there could be preventions of the ships, that are here for repair, are here to help the said Count of Castelo Melhor in case of emergency; the Count does not mention it, he rather neglect it. Because he only cares that this city loses everything, as one can perceive, like selling everything, giving permission to all who are going, taking all the constapels; take away the artillery that was for the fortifications, and all the money that is still to be due, he spends in such a way that the new Count who is being expected, is bound to have big disagreements with him, however meek he may be. Assist His Majesty in this as he see fit by taking the information for what is to be done with it after thinking carefully himself. As His Majesty ordered to make ready the admiral ship and the vice-admiral ship, he concerns only with these two, the others are not mentioned and he only takes care of by putting many goods in the admiral ship of jacaranda wood and brazilwood and many other things that are public if one wishes to know them. I comply with this What comes next was in writing. And when I don't say it I satisfy myself. I have been in prison for a year and a half, seeing one sergeant hand me over to another every day, with such orders that I have written and next to the sent certificates, and when I find myself so ill-treated, because of which I am all grey and very sick, anyway I feel mischief but to lose His Majesty a vassal who loved his service so much, and was so zealous of it. I clearly see that this is madness, but still the same year, it will last in me forever, even if I see how badly I am thanked. Seeing his Majesty the state in which I was, so very badly remedied it. And I have no complaint about the Count, for he acted according to his intellect, but His Majesty, seeing such a mockery31, consenting to it is too much is very difficult for me. It's been a long time since I've had a letters from you. As yet some ships had come, presently Sebastian Vas came, and I have had no letters from no one. All have left me on this occasion, as I have noticed seeing in the few proceedings that have been taken for me to His Majesty, and I have an answer from the aforementioned Lord and I have not had a letter from you, nor many that I have written that I know that they were received. If only you are in good health, accustomed am I, to these certain fortunes. From April 11 we have news from over there, and as there was no news of Castelo Melhor's departure, so the ships will have to stay moored until next year, and will I have to remain in prison for more years after I’ve been already three years of imprisonment, and of being eight in Brazil, and they think that most certain I’ll stay forever, which is, what the ministers intend, the more news from Bahia, there will be no lack of those who will have written, may God preserve you as I wish. Bahia on May 28, [1]649. It was signed Antonio Telles da Silva KEY 167 IN MANUSCRIPT Dutch reads ‘burteneij’. Unknown word in Dutch. We follow the Portuguese translation and assume that it must signify ‘mockery’ (Dutch: bespotting, aanfluiting). 31 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 62 APPENDIX 2: Ciphertext 168 from 1649-6-3 RESOLVED PLAINTEXT CT2 (CIPHERTEXT) AND PORTUGUESE PLAINTEXT COMPARED INDEX COLORS. Grey = in ciphertext not in Portuguese; Yellow = in Portuguese not in ciphertext; Blue = differences in used letters e.g. ‘i’ versus ‘e’ (different written letters) or ’S’ versus ’s’ (capitals). Letters with * are reconstructed letters. Letters with ! are improved code errors. Space is ‘_’. Señor {l.iv} francisco_barreto mestre_de {l.v} campo_general_de_este_estado {l.} me escreueo,_que_uindo a_nazare, {l.} a_recolhe_r*_den_tro, as_m_quatro {l.} naos i*ngrezas dei*xara_no arra {l.} i*al*gouern_ando_os mestres de {l.x} cam_po francis_co_de fi_guei*ro {l.} a e_i*oam fr’z*. e_que ioam*fr’z. uiei* {l.} ra, se_fora para_sua_caza_que {l.} ixozo de_nam fi*ca r_so gouerna {l.} ndo, e_que_logo socedera,_amoti* {l.xv} narense os soldados q!uerend_o {l.} matar_a_francisco de_fig!uei*roa {l.} e_que_o_mesm!o! hauiam de_fazer ao {l.} mestre de_campo_general, pois_l* {l.} hes na’ daua a_racam e_fardas, {l.xx} tendo_oi*tenta mi*l*_cruzadom e {l.} si*nco mil* fardas, _q’ue uossa magestade {l.} nas di*tas naos i*ngrezas manda {l.} ra. e_q‘ue_nam_queriam_outro_gou {l.} ernador sen*a’ i*oam fr’z. ui*eira, {l.xxv} o_qual de_sua caza_ti*nha ac {l.} odido_a_aqui*etar os sol_dad {l.} os prometendo l*hes tudo_quan {l.} to elles pedi*am, av!i*zando_a_na {l.} zare ao_mestre_de_cam_po_genera {l.xxx} l*_q’uez_fosse logo para_o_arraial; o {l. RIJKSARCHIEF ‘S-GRAVENHAGE {l.} qual teue outro aui*zo ne_o_u {l.} tras pesoas_que_na’_fosse porque o_qx {l.} uer_i*am32 de matar no_no_cami*nho, e_por {l.xxxv} esta_ca_uza_na’ foi*, mas mando1n {l.} ao_mestr*e de_campo andre_ui_d {l.} [f142v] {l.xxxviii} al, q’ue_estau a em napare q’ue fosse a {l.} qui*etar os_soldados l*euondo_os ca {l.xxxx} rregasoi*s do_q’ue uinha_nas naos in {l.} grezas para_q’ue_uis*em os_solda_dos {l.} q’ tudo o_q’ue_cuida_uam ui*nha_nas n 32 Ciphertext: qx uer iam (or without spaces: qxueriam). In Portuguese text: avião de? Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 63 {l.} aos, hera_aleuantado pel*os e {l.} mulos do di*to_mestre de cam*po {l.xxxxv} gener_al. mas nada_bastou_pa {l.} ra_os aqui*etar, e_ui*ol enta {l.} men*te fi*z_eram com_q’z se_l*hes de {l.} sem a_qu_atro_mi*l rei*s, e_oi*to ce {l.} ntos reis a ca_da_solda_do?33, q’ue dis {l.l} francisco barr_eto_q’ue nam sabe don_de {l.} os ha de_ti*rar mas i*ase34cont {l.} i*nuando em lhos dar. eq’ue hua’ {l.} co_m_panhi_a_de i*oam fr’z uiei*r {l.} as_q’ue estaua em naz_ar_e, se_l*eua {l.lv} ntara para_i*r para_o_ma1to; mas {l.} o mestre de_campo_general*os a {l.} quietara dando_l_hes loguo su {l.} as fardas. di*smq’ue o_ini*mi*go_e {l.} sta em tal esta_dos_q’ue namz tem {l.lx} com_q’ue possa sahi*r fora. e_q’ue pa {l.} ra_olanda parti_?am35 quatr {l.} o_centos olandez_es i*nutei*s, {l.} assi*_dos_q’ue uieram de angola, c {l.} omo os_q’ue estaua’o estropea_d {l.lv} os, das sahi*das_q’ue haui*am fe {l.} i*to. e_q’ue foram mui*tos iudeus c {l.} om suas mol*heres di*zendo_os r {l.} endi*dos_q’ue sahi_ao’q’ue hera p {l.} rat_i*ca entre os soldados_q’ue {l.lxx} se dentro em tres mezes lhes {l.} nao’ uier socorro de_ol*anda_ha {l.} ui*ao’ de_largar as_forsas- e_q’ue {l.} i*ndo os ol*andezes ao si*ara por {l.} terem noti*sia q’ue haui*a_mi*n_as {l.lxxv} de ouro, os indi*os der_ao’ nelles {l.} e_os matarao’ todos e na’o escap {l.} 143[f143r] {l.} an_do36 mai*s_q’ue dous q’ue trouxe {l.} rao’_as nouas- e_q’ue as naos dos {l.lxxx} estados com o_poder_q’ue tinh {l.} ao’, andauam nesta_barra d {l.} esta bahi*a esperandome a {l.} mi*m por noti*s_i*as_q’ue ti*rham_q’ue {l.} eu hia, como_tambem ao conde {l.lxxxv} de_castel melhor- esta_he {l.} senhor a_sustansi*a_da_carta {l.} de francisco_barret_o_a_qual*_t {l.} anto q’ue_a recebi*_a mandei* por {l.} luis da_si*lua_telles, ao_cond {l.xc} e de_ui*l*a_pouca, por elle nao’t Code ‘1’ in ciphertext. No letter needed in Portuguese plaintext, hence ‘?’. Ciphertext: iase. In Portuguese tekst: chase. 35 Unreadable in ciphertext, hence ‘?’. Letter ‘r’ needed in Portuguese plaintext. This is code ‘8’ or ‘81’. 36 Ciphertext: escapando. In Portuguese tekst: escaparão. 33 34 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 64 {l.} er, notisi*as tao_mi*udas do s {l.} uceso, e_se para a_rezo_lus_am {l.} q’ue se tomasse, fosse necessario_d {l.} ar eu_algu_ma’_info_rmasao’_ou {l.xcv} pareser, me! _oferesi*, por q’ue_so {l.} me lembraua o_ser_uiso_de_u {l.} ossa magestade, e_me esquesi*a {l.} estar prezo_ha uinte_mezes, {l.} entregue_a_hu’m sargento- de {l.c} mi*m_nao’_qui*zeram saber nada, {l.} nem_q’ue eu_soubese a_rezolusam {l.} q’ue_se_tom_aua, quei*ra deus_q’ue {l.} sei*a a_q’ue conuem ao_serui*s_o {l.} de_uossa_magestade_q’ue heso37 he o_qu {l.cv} e_dezeio_q’ue se acerte- guarde de nosso {l.} Srde38 a Catolica real pessoa de vossa magestade como {l.} a Cristandade e Seos vassalos avemos {l.} mister. baia em 3 deuiho39 de [1]649 {l.} [Watermark: three-leaf or trefoil with letter GP or CB] {l.cx} Antonio Telles {l.} da Silvas PORTUGUESE AND DUTCH TEXTS COMPARED WITH ENGLISH AS BRIDGE LANGUAGE INDEX COLORS. Blue = differences in used words; Grey = in Dutch not in Portuguese; Yellow = in Portuguese not in Dutch. Lord Francisco Barreto, General Field Marshal of this city wrote me, that when he came to Nazare, to collect the four English ships, he left governing the in the Arrajal, the Field Marshals Francisco de Figueiroa and Joan Fernandes Vieira and that Joan Fernandes Vieira went to his house complaining that he would not be the only one governing and that soon followed by the rioting of the soldiers wanting to kill Francisco de Figueiroa and that they would do the same to the General Field Marshal, because he did not give them the rations and cloths, having eighty thousand crusados and five thousand cloths, that Your Majesty had sent in the English ships. And saying that they wanted no other governor but Joan Fernandes Vieira, who came from his house had helped to calm the soldiers by promising them everything they asked for, writing to the General Field Marshal in Nazare to go to the Arraijal; who had another warning from other people not to go because they would kill him on the way, and for this reason he didn't go but sent to the Field Marshal Andre Vidal, who was also in Nasare, to calm the soldiers by taking the delivery notes with him of that what was coming in the English ships to show, so that the soldiers could see everything that they cared about that was coming in the ships, it was only a decoration, someone-who-begrudges40 of the General Field Marshal. But nothing was enough on the contrary and they violently made them give four thousand réis, and eight hundred réis to each soldier, which Francisco Barreto says he doesn't know where to get that money from but he still keeps on continuing to give to them. And that a company of João Fernandes Vieira that was in Nazaré muitinied and wanted to go into the woods, but the General Field Marshal had calmed them down by giving them some cloths. They says that the enemy is in such a state that they have no power of getting out of the field. And that four hundred useless Dutchmen left for Hollanda, both those who came from Angola and those who were crippled from the exits they had made. And that many jews with their wives went, the defectors 37 Ciphertext: heso. In Portuguese text: isso. Both in English: that. Ciphertext: Srde. In Portuguese text: S.or. Both are an abbreviation of ‘Señor’ (English: Lord or Sir). 39 Ciphertext: uiho. In Portuguese text: Junho. Both refer to the month ‘June’. 40 Dutch 'misgunnen'. Translation: begrudge, thinking that someone shouldn't get something. Dutch 'aemuleren' (English: emulation). A 'rival' is an 'aemulus' in Latin. 38 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 65 declare that among the soldiers is said that if no help came from Holland within three months, they would leave the fortresses41. And that when the Dutch came to siara because they had news that there were gold mines, the brasilians attacked them and killed them all and only two escaped, who brought the news. And that the ships of the States with the power they had were near this harbour waiting for me, for they heard the news that I was going to leave thither, as well as to the Count of Castelo Melhor. Lord this is the substance of Fransisco Barreto's letter which as soon as I received it I sent (by Luis da Silva Telles) to the Count of Vila Pouca, because he had no news so small of the success and if for the resolution that was taken, it was necessary for me to give some information or opinion, I offered myself, because I only remembered Your Majesty's service, and I forgot myself that I had been in prison for twenty months, in the care of a sargeant. They did not want to know anything about me, nor did they want me to know the resolution that was to be taken; may God grant that it be the one that is convenient for Your Majesty's service, for that is what I wish to be done, may God take care the royal Catholic person of Your Majesty, alike Christendom and His vassals must. Bahia on June 3, [1]649. And it was signed Antonio Telles da Silva. KEY 168 IN MANUSCRIPT CODES IN CIPHER 168 NOT ACCOUNTED FOR IN ITS KEY Nr 1 2 3 4 5 41 Code (count) mistake by encrypter 215 (1) instead of 205 for E 219 (1) instead of 292 for G 726 (1) instead of 7261 (or 7262 or 7265) for Q 1405 (1) instead of 405 for V 3211 (1) instead of 32+11 for M+O Portuguese reads ‘forças’ (English: forces). Dutch reads ‘fortressen’ (English: fortresses). Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 66 Nr 1 2 3 4 Reconstructed code (count) 32 (2) for M fits in used key-strategy +1, +2, and +5 45 (2) for N fits in used key-strategy +1, +2, and +5 52 (1) for Z fits in used key-strategy +1, +2, and +5 61 (11) for L fits in used key-strategy +1, +2, and +5 5 6 7 8 9 10 62 (5) for L fits in used key-strategy +1, +2, and +5 72 (1) for S fits in used key-strategy +1, +2, and +5 82 (2) for R fits in used key-strategy +1, +2, and +5 311 (27) for I fits in used key-strategy +1, +2, and +5 312 (28) for I fits in used key-strategy +1, +2, and +5 315 (26) for I fits in used key-strategy +1, +2, and +5 COMPLETE RECONSTRUCTED KEY OF CIPHER 168 INDEX COLORS. Yellow is a code that fits in used key-strategy +1, +2, and +5 but that’s not used. Green is a reconstructed code. Letter a b c d e f g h i (ij) j k l m n o p q r s t u (v) v w x y z Code 24 9 10 20 30 16 19 22 31 6 3 4 11 25 26 8 7 13 40 2 5 +1 +2 241 242 91 92 101 102 201 202 301 302 161 162 191 192 221 222 311 312 letter not used letter not used 61 62 31 32 41 42 111 112 251 252 261 262 3261 3262 7261 7262 81 82 71 72 131 132 401 402 letter not used letter not used 21 22 letter not used 51 52 +5 Notes 245 95 105 205 305 Before reconstruction: 30, 301, 302, 305 165 195 225 315 Before reconstruction: 31, 301, 302, 305 65 35 45 115 255 265 3265 7265 85 75 135 405 Before reconstruction: 6, 65 Before reconstruction: 3, 35 Before reconstruction: 4, 41 Before reconstruction: 8, 81 Before reconstruction: 7, 71 25 55 Before reconstruction: 5, 55 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 67 NUMBER-CODES OVER LETTERS OF CIPHER 168 APPENDIX 3: Letterbooks WIC between 1645-1654 List of WIC sources with letterbooks and their archive number in Nationaal Archief, Den Haag (NL-HaNA), Oude West-Indische Compagnie (OWIC), inv.nr. 1.05.01.01: Nr WIC Year Month No Nothing Examinaties Intercepted Cipher(tekst) Letter archive content relevant prisoners or letters mentioned King number (only defectors Portugal cover not in page) cipher 1 60-74 1645 6 1 2 60-100 1645 6 1 3 60-125 1645 8 1 4 60-189 1645 8 1 5 60-192 1645 9 1 1 6 60-146 1645 9 1 1 7 60-12 1645 ? 1 8 60-26 1645 ? 1 9 60-48 1645 ? 1 10 60-103 1645 ? 1 11 61-56 1646 3 1 1 1 12 62-74 1646 9 1 13 61-60 1646 ? 1 1 14 62-5 1646 ? 1 15 62-16 1646 ? 1 16 62-43a 1646 ? 1 17 62-45 1646 ? 1 1 1 18 62-89 1646 ? 1 1 19 62-111 1646 ? 1 1 20 63-38 1647 5 1 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 68 Nr 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 WIC Year Month archive number 63-75 63-101 63-104 63-99 63-14 63-67 63-102 64-30 64-63 64-76 64-86 64-103 64-104 64-131 64-1 65-40 65-65 65-107 65-120 65-154 65-186 661-1 661-17 661-34 661-59 661-72 661-57 661-77 662-1 662-35 662-34 662-89 662-59 662-70 662-80 662-69 662-79 662-87 662-94 662-2 67-1 67-40 67-60 67-94 1647 1647 1647 1647 1647 1647 1647 1648 1648 1648 1648 1648 1648 1648 1648 1649 1649 1649 1649 1649 1649 1649 1650 1650 1650 1650 1650 1650 1651 1651 1651 1651 1651 1651 1651 1651 1651 1651 1651 1651 1652 1652 1653 1654 8 8 9 12 ? ? ? 7 8 8 9 10 10 12 ? 3 4 7 9 11 12 ? 3 5 6 6 6 11 1 6 6 7 8 9 9 9 9 10 12 ? 2 10 3 2 No content (only cover page) Nothing Examinaties Intercepted Cipher(tekst) Letter relevant prisoners or letters mentioned King defectors Portugal not in cipher 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 69 Nr WIC Year Month archive number Total No content (only cover page) 12 Nothing Examinaties Intercepted Cipher(tekst) Letter relevant prisoners or letters mentioned King defectors Portugal not in cipher 17 27 19 3 2 Summary WIC sources with letterbooks. Year and relevance: Year 1645 1646 1647 1648 1649 1650 1651 1652 1653 No content Nothing (only cover relevant. page). Count Count 1 3 1 4 2 1 Intercepted Cipher(tekst) Letter King letters. mentioned. Portugal not in Count Count cipher. Count 5 2 2 3 5 2 4 5 1 2 1 1 3 4 4 5 4 4 3 1 1 27 19 1 2 2 1 1654 Total Examinaties prisoners or defectors. Count 12 17 3 Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 70 2 APPENDIX 4: Ciphertexts Mentioned between 1645-1654 Sources in Dutch and Portuguese archives concerning Brazil, between 1645-1654, with ciphertexts or mentioning them: Original Jewish Nr Archive Source Sorting year Where What connection letter 1650-7-19: in his testament Antonio Teles da Silva referring Teles da Silva, mentions states that he had several 1 Portuguese to 1642- Seas none 1650b cipher ciphered letters that his 1647 brother had to hand over to the king of Portugal. Hoge Raad, 1646-5-8: intercepted letters 1646 mentions deciphered in cipher from Jan Viera 2 Dutch 1646 Land Vieira, 1646 cipher by "a jew" d'Allegro [or João Vieira Kort Discours de Alagoas]. Rebellye, 1647 1646-5-??: letter from Abraham de Pina [or decipherer Aarão Sarfati to Dutch] De Pina is 3 Dutch Pina de, 1646 ciphertext 1646 Land with key for ciphers and a jew explanation. Decripted letters. 1647-08-10: governorgeneral Antonio Teles de Teles de mentions 4 Portuguese 1647 Seas none Menezes received a cipher to Menezes, 1647 cipher be used in "matters that ask for secrecy". 1649-5-28: letter in cipher from prisoner Antonio Teles da Silva, deciphered Telles da Silvas to king 1649a1 by "a Portugal. 5 Dutch ciphertext 1649 Seas Teles da Silva, skilled 1649-6-3: letter in cipher 1649b1 person" from prisoner Antonio Telles da Silvas to king Portugal. 1649-11-29: letterbook mentions a deciphered and translated letter from admiral Corea da Saa, mentions Salvador Correa de Saa to 6 Dutch 1649 Seas none 1649 cipher the king of Portugal. Other source: letter is dated 16497-4. 1649-11-28: prisoner Steven (Stevão) Dias has to explain mentions 7 Dutch Dias, 1649 1649 Land none the caracters, lines, points secret and words in a letter that his communication brother sent to him. 1649-12-3 and 1650-1-18: jew Moyses Raphael Dagilaer [or Moisés Rafael decipherer de Aguilar] is promised 30 mentions 8 Dutch Dagilaer, 1649 1649 Seas Dagilaer is guldens for find out content cipher of encrypted letters for a jew governor in Angola. Found in a captured ship by captain Hartman. 1650-7-10: Teles da Silva emphasizes the ability of Teles da Silva, mentions 9 Portuguese 1649 Seas none Ravasco to decipher the 1650a cipher king’s ciphertexts. Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022 71 Published by: NEALT Proceedings series 49 Linköping University Electronic Press, Sweden Linköping Electronic Conference Proceedings, No. 188 ISSN: 1650-3686 eISSN: 1650-3740 ISBN: 978-91-7929-397-0 (PDF) URL: https://doi.org/10.3384/ecp188 Cover photo: Ecolex I, Crypto Museum Eindhoven