At a glanceKey from adjacent plaintext · partial (84% of the cipher)
- The telegrams
- Twelve telegrams, ten with cipher, sent from Rio de Janeiro to Fortaleza between 10 March and 20 May 1914 by the Minister of War, Gen. Vespasiano de Albuquerque, his aide-de-camp 1st Lt Rego Barros (Setembrino’s son-in-law) and E. Pamplona. Most were sent twice, the second copy as a service check.
- Where
- CPDOC/FGV, Rio de Janeiro, Arquivo Fernando Setembrino de Carvalho, FSC pce 1914.03.10/4 (DocReader FSC_PACEARA, pagfis 1317–1332, 16 images). Catalogue entry 1470; not on DECODE.
- The cipher
- Vigenère over the standard 26-letter alphabet, plain = cipher − key, a key of eight letters restarted at every cipher group (one word, sometimes two short words run together). MFTDBSJV for 14 March to 17 April; a near variant, LETCARIV, for 10 March; another key, OVHXZIRF, for 20 May.
- How it was read
- The key MFTDBSJV was rebuilt from the contemporary pencil decipherment written on the sheet of 14 March. The pre-registered test of the key on 731 held-out letters failed (0.733 against 0.85); a larger lexicon chosen after that result gives 0.865. The telegrams read as continuous Portuguese either way.
- What it says
- That the government palace of Fortaleza was mined, with an engineer’s account of the works under it; Ceará faction politics; telegraph staff; and a reserved warning that the Minister was displeased by the cipher part of a telegraph conversation.
- Still open
- Two groups (11 and 7 letters) stay unread, the keys of 10 March and 20 May are not validated, and the transcription was not compared with the images. Counted in cipher groups, the unit of the read bar, the main key’s seven telegrams are 92.3% read and the file 83.8%: under 95%.
- Credit
- Daniel Bourdeau found the system, rebuilt MFTDBSJV and made trial readings (vetting of 28 October and 3 November 2026). Brice Dudout (with AI language-model agents) made the blind transcription of the 16 images, ran the pre-registered test, read the telegrams and found the 20 May key.
01 The telegrams
The folder holds sixteen images: telegram forms with the address and the stamp of the Fortaleza telegraph office, the cipher groups copied on paper tape, and for most telegrams a second strip, marked “Ct”, repeating the groups. They went from Rio to Col. (from April Gen.) Fernando Setembrino de Carvalho, inspector of the military region and federal commander at Fortaleza during the Juazeiro sedition (Padre Cícero’s rising against Governor Franco Rabelo). Some telegrams mix clear words and cipher groups; the sheet of 14 March carries a contemporary pencil decipherment of its telegram, and a pencil attempt on the correction telegram (pagfis 1326) stops at “complot amli…”.
No decipherment of the other telegrams is on the sheets. The catalogue entry (key_lookup in the project’s own vetting) noted that a key rebuilt from the pencil decipherment read the March and April telegrams; this page is that reading, made with a blind transcription and a pre-registered test.
02 The cipher and the keys
Each cipher group is one Portuguese word (two short words are sometimes run together, the key restarting at the second), enciphered with the same eight key letters from the first letter of the group. Three keys are involved:
- MFTDBSJV, for the telegrams of 14 and 18 March, the correction telegram, 19, 24 and 29 March and 17 April (Vespasiano, Rego Barros, Pamplona). Rebuilt from the pencil decipherment of 14 March: the contributor fixed it on that sheet alone, where every key position agrees with the 17 groups except one conflict at a letter already marked doubtful in the transcription.
- LETCARIV, for 10 March (Pamplona), Daniel Bourdeau’s proposal; it differs from MFTDBSJV by one step in six of eight positions, and on 11 March Pamplona wrote in clear that the key he used was the Minister’s. The clerk seems to have mixed the two, which gives many one-step slips. Provisional.
- OVHXZIRF, for 20 May (Rego Barros), found here from the ciphertext alone by hill climbing on the even groups of one copy (period 8; periods 8 to 14 all give the same first eight letters). Provisional.
03 The test, which failed
By the contributor’s record, the transcription and pre-registration were frozen before any key trial (a blind transcription of all 16 images, made by an agent that saw no other file, and a pre-registration; SHA-256 values in the files). The key MFTDBSJV was to be fixed on the sheet of 14 March alone and measured on two other telegrams: 731 held-out letters (18 March and 17 April). The measure is the share of cipher letters whose group deciphers to a word of a Portuguese lexicon built from six literary books, or to two such words, against 1,000 random keys and 1,000 copies with the letters shuffled inside each group; the criterion is 0.85 and a margin over every control.
The pre-registered test failed: S = 0.733, under 0.85 (the best of the controls is 0.097). The groups it did not count are ordinary words missing from that small lexicon (informada, latrina, dispositivos). The contributor then enlarged the lexicon by a rule, with the measure and thresholds unchanged (“amendment A1”, declared as written after the result), and got S = 0.865 against at most 0.129 for the controls. That is a changed rule and not a pass of the test; both results are in the files. The key is credible for other reasons: the pencil decipherment agrees at every position, the held-out telegrams read as continuous Portuguese, and the controls stay under 0.13 either way. LETCARIV (93 letters) and OVHXZIRF (held-out half of 152 letters, S = 0.546) did not reach the criterion either.
The scripts and the lexicon are not in the repository, so a third party cannot re-run the measure. Our own check was a decoder written from the key file: it gives the same mechanical decryption as the contributor’s for 242 of 243 groups (the one difference, OJTUB, is taken from a service note on the sheet). The order of events (hashes frozen before the trials, single commits) is by the contributor’s record.
04 The reading
Pessoa que me parece bem informada communica-me saber por parente Rabello que o palacio acha-se cuidadosamente minado. (14 March)
In English: a person who seems to me well informed tells me he knows from a relative of Rabelo that the palace is carefully mined. This is the telegram with the pencil decipherment, so it is a verification, not a new reading.
O Dr. Marques Porto declarou-me o seguinte: foi feito movimento de terras nos baixos do edificio do palacio do governo para construcção de uma fossa septica e sumidouros da latrina da secretaria do interior. […] Uma escada que fica no corredor fronteiro do gabinete da secretaria do estado do interior conduz á sala que fica aos fundos do corpo da guarda e sob o grande salão do palacio do governo. (18 March, three sheets, 571 cipher letters)
Dr Marques Porto told the Minister that earth had been moved under the palace for a septic pit and drains, that a tiled floor and a room for the Interior Secretariat’s archive were made over them, and that a staircase leads from the corridor to a room under the great hall; he was replaced by Lieutenant Bizerril, who went to Crato as engineer officer of the forces under Alípio Barros. The 19 March telegram from Rego Barros says that the General censured Pinheiro for publishing a telegram, and tells of Thomaz Cavalcanti and Liberato; 24 March (Pamplona) refuses a telegraphist and says why Andrade Fortuna should not stay; 17 April answers the Pernambuco judge pending the Supreme Court. The telegram of 20 May, in the provisional key, is Rego Barros’s reserved warning to his “querido sogro” that the Minister was displeased by the cipher part of a telegraph conversation.
The sentences read in the 1914 spelling and telegraphic style (telegramma, elle, commando); a small Portuguese lexicon marks only 28% of its sentences clean for that reason, and the translation is in the full texts.
Counts. The contributor counts by cipher letter: the seven telegrams in MFTDBSJV have 1,043 letters, 1,013 sure, 1 probable, 11 restored and 18 unresolved (97.2%), and the whole file 1,440 letters (95.8%). The read bar of the README counts words, and a word with a letter restored by the sense, or unresolved, is not read. By cipher group: the seven telegrams in MFTDBSJV have 182 groups, 168 with no restored letter and none unresolved (92.3%); the file has 240 groups, 201 such (83.8%). Both are under 95%, so the file is read in part.
05 What remains uncertain
- MRELDTKRIMV (pagfis 1326, after “complot”, 11 letters) and SUQHSJU (pagfis 1328, after “teme ataque”, 7 letters) give no word under any rotation of MFTDBSJV, LETCARIV, OVHXZIRF or LBQPOHVJ (the key of a 7 October 1914 item, catalogue 1471). The contemporary pencil attempt on 1326 stops at “complot amli…”. Too short to test a transmission slip.
- OJTUB (“Ceará”) is read from a service note on the sheet and counted probable.
- The keys of 10 March (the contributor counts 82.8% of its 93 letters as sure; 16 restored) and 20 May (the contributor counts 95.1% as sure plus probable letters, 289 of 304; 15 restored) are not validated; both texts are provisional.
- The transcription was not compared with the images: the DocReader pages load their images by JavaScript, so a browser session is needed. The typed clear copies FSC pce 1914.02.19/2 of Vespasiano’s telegrams, a crib in the same fonds, were not opened.
- The pre-registered test failed and the scripts are not in the repository (see section 03).
Full texts
The transcription and the decryption of every document on this page, so that the reading can be used and checked. Short files are shown here; all are in the repository.
Cipher telegrams Rio de Janeiro -> Fortaleza, 10 March - 20 May 1914 (FSC pce 1914.03.10/4) read in part
Telegrams to Col./Gen. Setembrino de Carvalho, federal commander at Fortaleza during the Juazeiro sedition, from the Minister of War Vespasiano de Albuquerque, his aide-de-camp and son-in-law Rego Barros, and E. Pamplona: reports that Governor Franco Rabelo's palace was mined and the engineer's account of the works under it, Ceará faction politics (Thomaz Cavalcanti, Liberato), telegraph staff, the Pernambuco judge and the Supreme Court, and a reserved warning on 20 May that the Minister was displeased by a cipher telegraph conversation.
Builds coherent sentences, but under the read bar of 95% of words read as sense (84% here); the holes are marked in the reading.
Transcription: transcription.txt
Decryption
Setembrino 1914 (catalogue 1470): mechanical decryption. Vigenere, standard 26-letter alphabet, plain = cipher - key, key of 8 restarted at every cipher group. Columns: cipher group, mechanical plaintext. Readings and corrections are in reading.md. == p1317_10mar key LETCARIV # 10 Mar 1914, E. Pamplona (pagfis 1317). Key LETCARIV (Daniel Bourdeau, vetting 28 Oct 2026); many one-step keying slips. OSNVOJ doutos QEUQ fabo NLXHE chefe OMLVRTCJ distrcuo OIN deu BEKVE qarte OSXPTV doente OIOG deve EYTUTACPTPH tuastjuuilo LGMXAD actvam PRUCRJMBLHH enbarsegado PPEG elle YIKGCVAU nerecesz EYS tuz NSGHIRVXL confianca == p1319_14mar key MFTDBSJV # 14 Mar 1914, Rio no. 40914, Vespasiano de Albuquerque (pagfis 1319; 1320 = ampliacao, same text). Contemporary pencil decipherment on 1319. BJLVPS pessoa CZX que YJ me BFKHDW parece NJF bem USYRSEJYM informada OTFPVFRXM communica YJ me EFUHS saber BTK por BFKHOLN parente DFUHMDX rabello BFEDDAX palacio MHAD acha EJ se OZBGBTXNMRXQUW cuidabosamente YNGDEG minado == m52414_18mar key MFTDBSJV # 18 Mar 1914, Rio no. 52414, Vespasiano de Albuquerque (pagfis 1321-1323). PONTO = clear word. PW dr YFKTVWB marques BTKWP porto PJVOBJXP declarou YJ me EJZXJFCZ seguinte RTB foi RJBWP feito YTOFNWWOA movcmento PJ de FJKUBK terras ZTL nos NFBAPK baixos PT do QIBIJURJ edificio BFEDDAX palacio STOHSFX governo BFKD para OTGVUJDXOFH construccao GRT uma RTLVB fossa EJIWYUJ septxca Q e EZFLEGDMAX sumidouros PF da XFMUJFJ latrina PF da EJVUFLJMUF secretaria PT do USMHSAXM interior PONTO djuqn FTUUF tobre QXLHT esses PNLSPKROUAHV dispositivos PJ de QXZRUG esgoto RTB foi RJBWP feito XFWUJDQVYJGWP ladrilhamento QTBWXSBJX eoitwasol PJ de GRT uma EFED sala BFKD para A o MWVKJNX archivo PF da DJYHSAMV referida EJVUFLJMUF secretaria PONTO djuqn GRT uma QXVDES escada CZX que RNVD fica ZT no OTKUFVXM corredor RWHQUWRMA fronteiro PT do SFULOWCZ gabinete PF da EJVUFLJMUF secretaria PT do QXMDEG estado PT do USMHSAXM interior OTGGVR conduz M a EFED sala CZX que RNVD fica MTL aos RZGGPK fundos PT do OTKSP corpo PF da SZTUES guarda Q e ETU sob A o SWTQEW grande EFEDP salao PT do BFEDDAX palacio PT do STOHSFX governo A o PW dr YFKTVWB marques BTKWP porto RTB foi EZUVUACPUIH substituido BJER pelo FJGHOLN tenente NNSHSJRG bizerril CZX que EJZXJG seguio QR em PNKHDUJJ direccao OWTWP crato OTFR como QSVDSJNBMIH encarregado EJKYCUX servbco QSZHOZJMUF engenharia RTKFBK forcas CZX que ETU sob OTFPWFMJ commvndo MQBSJG alipio NFKUPK barros QCIHEALDASTUBE expedicionaram BFKD para MQB ali == p1324_19mar key MFTDBSJV # 19 Mar 1914, Rio no. 54314, 1o Ten. Rego Barros (pagfis 1324; 1325 = service repeat). Mixed clear/cipher; only the cipher groups here. SJGHSSU general QCIUPTXP exprobou BWHFFVRHQOMR procedimejto BNGKFAAN pinheirs BZUOJUJIPT publicando FJEHHJJHYF telegramma EJN seu PNLVF disse EW sr NJF bem PJOLB devia FJEHHJJKTFK telegraphar OTFR como MRBJP amigo FMHPBR thomaz XNUHSSCJ liberato == p1326_23mai key MFTDBSJV # correction telegram, Rio no. 62314, Rego Barros (pagfis 1326): "houve engano palavras sao" + 2 groups. OTFSMGC complot MRELDTKRIMV amlicbbwwhc == p1327_24mar key MFTDBSJV # 24 Mar 1914, E. Pamplona (pagfis 1327). Mixed; cipher tail only. DJOHMSLVA revelacao FJEHHJJHYF telegramma MKKDOUX afranco DFUHMDX rabello PJVXK decuj BTELUALV politica QWTDEWYOA eraadepto == p1328_29mar key MFTDBSJV # 29 Mar 1914, Quartel General Rio no. 89914, Rego Barros (pagfis 1328). Mixed. OUTUB: service note on the sheet gives OJTUB. SUQHSJU gpxerrl YRLLDS mmsica YFGLGWBOMHTR manifestacao FMHPBR thomaz OUTUB cpara DJIUFKNIFFK representar == p1329_17abr key MFTDBSJV # 17 Apr 1914, Rio no. 62814, Vespasiano de Albuquerque (pagfis 1329, with Ct repeat; repeat readings used: BFKHDW, MUXVBJ, ONMDEGB). BFKHDW parece OTGYFFRZZYX conveniente DJLSPFMZD responder VZEC julz BJKGBEKPOT perdambuco PNSHOVX dizendo CZX que MUXVBJ apesar ZFH nao VZEJBJ julgar UIXQUALJE identicos OFLRT casos ONMDEGB citados OTGWVVX contudo MHTWBFMJ acatando EZT sua PJVLTSX decisao MLNDSVJ aguarda ETEXDSX solucao PJYLOACDHF definitiva EZIUFEX supremo FWBEVFJG tribunal == p1331_20mai key OVHXZIRF # 20 May 1914, Rego Barros, Urgentissimo reservado (pagfis 1330-1332, three copies; p1331 used). Key OVHXZIRF found here ciphertext-only (provisional, see validation.txt). OHPDM amign GDUGDZF sinjero DVYBMBV parente RZCBQ dever ZZHICIUJ lealdade OQPPZZ avisar UZUBQIC general RDGBQ dizer RZTLMAKWJP demonstrvu QJUQQIIMSYHAD contrarhedade ODMOZLR aifrada AJUSDZJF monversa VJUGDU honjem OKYBBQRW apreciar HZSBFZRRAVZ telegrammas SIAETAZFGHHANA enthusiasmados SSAOZVYTI extranhou VJUGDU honjem OQPPN aviso OHPDM amign OHPWZIV amizaae RZKLBW dedoco DZUL peno JZY ver HZSBFZRRAV telegramma DMVSZ prova GDUZOZZIOYL sincpridade UZUBQIC general OAMFQUF affirmo GJI sob SGSB elle GZABLIIMBJ setemarhno HRT twm OHPWZLV amizade DZJL peco EPLOHLF querido GJNON sogro BJZPN nosso ADAFRBIT mitistro SIAETAZFGHHKCW enthusiasmando GZTMQM sempre QJYODKKF correcta OOAFSCUJ attitude == copy variants OJTUB ceara (service-note variant of OUTUB) RZKFBW dedico (p1332 copy variant of RZKLBW) DZJL peco (p1332 copy variant of DZUL)
Reading
# Setembrino 1914 (catalogue 1470): texto Clean Portuguese text of the cipher parts (with the clear words around them), 1914 spelling. Marked reading and count: reading_count.md. ## Chave MFTDBSJV 1. 14 de março de 1914, Vespasiano de Albuquerque. Pessoa que me parece bem informada communica-me saber por parente Rabello que o palacio acha-se cuidadosamente minado. Saudações. 2. 18 de março de 1914, Vespasiano de Albuquerque. O Dr. Marques Porto declarou-me o seguinte: foi feito movimento de terras nos baixos do edificio do palacio do governo para construcção de uma fossa septica e sumidouros da latrina da secretaria do interior. Sobre esses dispositivos de esgoto foi feito ladrilhamento e o preparo de uma sala para o archivo da referida secretaria. Uma escada que fica no corredor fronteiro do gabinete da secretaria do estado do interior conduz á sala que fica aos fundos do corpo da guarda e sob o grande salão do palacio do governo. O Dr. Marques Porto foi substituido pelo tenente Bizerril, que seguio em direcção ao Crato como encarregado do serviço de engenharia das forças que sob o commando de Alipio Barros expedicionaram para ali. Saudações. 3. 19 de março de 1914, Rego Barros. Abraços. O general exprobou o procedimento de Pinheiro publicando telegramma seu; disse que o senhor bem devia telegraphar como amigo. Thomaz, segundo me consta, anda dizendo que o senhor demittirá individuos indicados pelo partido delle; garante sob palavra de honra a integridade do commandante demittindo A ou B sem distincção, conforme julgar. Liberato manda dizer que a combinação marcha ás mil maravilhas. Penso que a promoção do general sahirá nestes oito dias. 4. Março de 1914, Rego Barros. Houve engano; as palavras são complot [...]. Indaga o commandante se não tem recebido telegrammas meus. Todos bons. Saudações. 5. 24 de março de 1914, E. Pamplona. Quanto a Andrade Fortuna, não me parece conveniente a permanencia delle ahi, porque sua remoção foi motivada pela revelação de telegramma a Franco Rabello, de cuja politica era adepto. 6. 29 de março de 1914, Rego Barros. Segue no paquete Ceará o ex-deputado Dr. Gonçalo Souto; peço garantil-o, teme ataque [...]. Negou musica á manifestação Thomaz, declarando não poder metter-se na politica do Ceará, nem se fez representar. Todos bons. 7. 17 de abril de 1914, Vespasiano de Albuquerque. Parece conveniente responder ao juiz de Pernambuco dizendo que, apesar de não julgar identicos os casos citados, contudo, acatando sua decisão, aguarda a solução definitiva do Supremo Tribunal. Saudações. ## Provisório (chaves não validadas) 8. 10 de março de 1914, E. Pamplona, chave LETCARIV. O doutor Fabio, chefe do districto, deu parte de doente; deve ser substituido pelo actual encarregado; elle merecerá tua confiança. Responda amanhã, que só resolverei depois da sua opinião. 9. 20 de maio de 1914, Rego Barros, chave OVHXZIRF. Como amigo sincero e parente, julgo dever de lealdade avisar que o general, apesar de nada dizer, demonstrou contrariedade pela parte cifrada da conversa pelo telegrapho hontem. Acostumado a apreciar vossos telegrammas enthusiasmados, extranhou o de hontem. Aviso meu distincto amigo exclusivamente pela grande amizade que lhe dedico. Por isso peço não ver neste telegramma senão mais uma prova da minha sinceridade. Quanto ao general, affirmo sob palavra que para elle tudo é Setembrino; lhe tem verdadeira amizade. Assim peço ao meu querido sogro telegraphar ao nosso ministro, enthusiasmando com a vossa sempre correcta attitude.
Translation
# Setembrino 1914 (catalogue 1470): English translation Telegrams to Colonel (from April General) Fernando Setembrino de Carvalho, inspector of the 5th (later 4th) military region, federal commander at Fortaleza during the Juazeiro sedition (Padre Cícero's rising against Governor Franco Rabelo), March-May 1914. "The General" / "the Minister" = Gen. Vespasiano de Albuquerque, Minister of War. Words in [ ] are supplied for the sense. ## Key MFTDBSJV (validated) 1. 14 Mar 1914, Vespasiano de Albuquerque: A person who seems to me well informed tells me he knows from a relative of Rabelo that the palace is carefully mined. Greetings. 2. 18 Mar 1914, Vespasiano de Albuquerque: Dr Marques Porto told me the following: earth was moved in the basement of the government palace building to build a septic pit and soakaways for the latrine of the Secretariat of the Interior. Stop. Over these drainage works a tiled floor was laid and a room prepared for the archive of the said secretariat. Stop. A staircase in the corridor facing the office of the Secretary of State for the Interior leads to the room at the back of the guardroom, under the great hall of the government palace. Dr Marques Porto was replaced by Lieutenant Bizerril, who left for Crato as engineer officer of the forces sent there under the command of Alípio Barros. Greetings. 3. 19 Mar 1914, 1st Lt Rego Barros, aide-de-camp: Embraces. The General censured Pinheiro's conduct in publishing your telegram; he said you would have done well to telegraph as a friend. Thomaz [Cavalcanti], I am told, goes about saying you will dismiss individuals named by his party; he pledges his word of honour on the commander's integrity in dismissing A or B without distinction as he sees fit. Liberato sends word that the arrangement is going marvellously. I think the General's promotion will come out within these eight days. 4. [23 Mar?], Rego Barros to 1st Lt Lafayette: There was a mistake; the words are "complot [unresolved]". The commander asks whether he has not received my telegrams. All well. Regards. 5. 24 Mar 1914, E. Pamplona: In reply to your esteemed telegram, I am sorry to say that the shortage of staff at Pojuca prevents me from granting what you ask for telegraphist Brígido. As to Andrade Fortuna, I do not think it advisable that he stay there, because his removal was caused by his revealing a telegram to Franco Rabelo, whose politics he supported. 6. 29 Mar 1914, Army Headquarters, Rego Barros: Ex-deputy Dr Gonçalo Souto is coming on the steamer Ceará. Please protect him; he fears an attack [unresolved word]. [The Minister] refused a band for the Thomaz demonstration, declaring he could not meddle in Ceará politics; nor did he send a representative. All well. 7. 17 Apr 1914, Vespasiano de Albuquerque: It seems advisable to answer the Pernambuco judge saying that, although not regarding the cases cited as identical, nevertheless, respecting his decision, [you] await the final ruling of the Supreme Court. Greetings. ## Provisional (keys not validated) 8. 10 Mar 1914, E. Pamplona: Doctor Fábio, head of the [telegraph] district, has reported sick; he should be replaced by the present officer in charge, who will deserve your trust. Answer tomorrow, for I will decide only after your opinion. 9. 20 May 1914, Rego Barros, "Most urgent - Reserved": As a sincere friend and relative I think it my duty in loyalty to warn you that the General, although he says nothing, showed displeasure at the cipher part of yesterday's telegraph conversation. Used to enjoying your enthusiastic telegrams, he found yesterday's strange. I tell you this, my distinguished friend, solely out of the great friendship I bear you, so I ask you to see in this telegram nothing but one more proof of my sincerity. As to the General, I affirm on my word that for him everything is Setembrino; he has true friendship for you. So I ask my dear father-in-law to telegraph our Minister, with your always correct attitude, warmly.
Key: key.tsv
06 Sources
- CPDOC/FGV, Arquivo Fernando Setembrino de Carvalho, FSC pce 1914.03.10/4, DocReader bib FSC_PACEARA, pagfis 1317–1332; the images are not reproduced here and were not fetched by the integration (the viewer needs a browser).
- Vetting, system and trial readings by Daniel Bourdeau, 28 October and 3 November 2026 (
research/latam_vetting/1470_sample.txt). - Blind transcription, pre-registration, test, key OVHXZIRF and readings by Brice Dudout with AI language-model agents (pull request #441), in
targets/setembrino1914.
The key, the transcription, the reading, the pre-registration and the checks are in targets/setembrino1914.
Cite this page
Brice Dudout and Daniel Bourdeau, “Cipher telegrams to Setembrino de Carvalho, Fortaleza, 1914”, Cyphersolver, posted 12 November 2026, https://cyphersolver.com/setembrino1914.html.