https://orcid.org/0000-0003-3207-8323
![Sample our Computer Science journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5928/TOC-Subject-Sample-2021-Computer-Science.jpg"})
Abstract
In July 1939, at the dawn of the Second World War, Polish cryptologists revealed their mastery of the Enigma encryption machine to their French and British counterparts. The Enigma machine was the main encryption method used by the German armed forces at the time. An alliance was born: XYZ (the letters are codes identifying each of the allies: X represents the French, Y represents the English, and Z represents the Polish). The Poles transmitted their methods to their allies, starting with the sheets invented by Zygalski. These sheets were made by the English, who then sent copies to the French and the Polish. These joined the French after the Polish defeat. With these sheets, from January 1940 onward, the allies were able to read the messages of the enemy. Unfortunately, in May 1940, the Germans changed their system, and the Zygalski sheets became obsolete. Only the English managed to continue breaking the Enigma keys via manual methods. In all, thousands of messages were read. In June, France capitulated, and consequently, the triple alliance ended. This story has been told before, but recently declassified French archives have allowed us to clarify several points, particularly the central place of the Cyclometer.
Acknowledgments
I would like to thank my daughter Alice for her help and George Lasry who criticized the draft of this article and who shared his software regarding throw-on indicators. I would also like to thank the reviewer for informing me about important technical and historical points.
Notes
1 Thanks to the presence of the keys in the documents provided by Bertrand, the Steckers are, therefore, known. In this case, there were no longer any fundamental difference between the commercial Enigma and the military Enigma.
2 Without the documents provided by Bertrand, Rejewski never would have been able to discover the wiring of the Enigma's rotors. Indeed, Lt R.E. Greenwood tells us that, to find the entry permutation (ETW or Steckers), we would have had many (about 20) long messages (80 letters each) in-depth to be successful (Weierud Citation2016). A priori, this situation never arose.
3 Knox was a former colleague of Denniston at the navy’s famous Room 40, which notably deciphered the Zimmerman telegram during the First World War.
4 Messages that have been encrypted with the same key are said to be "in-depth." This practice should be prohibited. Indeed, it allows the enemy to reconstruct the plain message regardless of the encryption method, including current methods. For some encryption systems, two "in-depth" messages are sufficient, but for Enigma, at least ten superimposed messages are needed.
5 The attachment is the plug-board with Steckers.
6 Given the extent of the information transmitted by the French, it is said that Denniston was very impressed and did not know what to give in return (Batey Citation2009).
7 Bertrand probably did not take full advantage of this meeting. Indeed, his cryptographic knowledge was weak. At subsequent meetings, he would ask the EMA to lend him an expert.
8 A priori the Bertrand Cribs.
9 The use of the Kenngruppen dates back to the 1937 indicator method, which was known to the allies through Asché. In this method, the location of the Kenngrupe was indicated by the daily key, but from the beginning of 1939, the Kenngruppe would always appear as the first group. The appendix to the book Enigma and its Achilles Heel (Skillen Citation1992) provides Enigma's 1937 encryption notice. In this, one can see manual additions which make it conform to the procedure of September 1938.
10 For the first three rotors, the key tables provided by Asché made it possible to know their names (I, II, and III). In contrast, I do not understand how the Poles associated names like IV and V with the cabling they reconstructed. Do not forget that Asché, as of September 1938, was no longer providing any information.
11 I think the Germans feared that several messages could be superimposed with a significant traffic ("in-depth"), such as that which allowed Knox to break the Italian Enigma.
12 It is odd that the Poles did not rebel against the English proposal to ask a spy to inspect the rotors. We know that, for example, the American Friedman had opposed espionage carried out by the navy to reconstruct the functioning of a Japanese encryption machine, because if it had been discovered by the Japanese, it could have called into question the results of pure cryptanalysis.
13 Mrs. Batey (Citation2008) reminds us that, in his work on the Enigma (National Archives HW 25/3 Citation1940), Turing wondered about the best way for a spy to allow the wiring of the rotors to be covered. The solution he gives is to type at least nine alphabets on the machine. In fact, Asché did it on his own but only typed the letters on the keyboard (Bouchaudy Citation2020a).
14 This questionnaire is created by the English, intended for the French and for Asché.
15 This is probably related to depths. This is indirect evidence of English breakthroughs.
16 Another indirect proof that the British played the game thoroughly and clearly indicates their stumbling block in their attempt to reconstruct the rotor cables.
17 For this point, it is obvious that Bertrand wished to benefit from the advances of the two other countries.
18 Denniston offered a tour of the Chatham listening center facilities during the meeting.
19 (SHD - FROM 2016 ZB 25/5 [file 268]; Gallehawk Citation2006)
20 Dolinoff is the code name for Dunderdale, a representative of SIS in France.
21 The device is, without question, the Cyclometer (cf. Subsection 5.1.2).
22 We will see later (see Subsection 5.7.4) that from May 1937 onwards, the German navy changed its methods, and consequently, the Poles could not continue to read their messages.
23 Denniston rubbed salt in the wound. Bertrand had only Braquenié as a cryptologist.
24 Figures 2, 3, and 8 correspond to blueprints made by the French (see Subsection 5.8). In all likelihood, these are copies of the Polish plans.
25 None of the Rejewski reports indicate this use: use the Cyclometer to directly break the keys of the day. The following description comes from Hut 6. Perhaps it is inspired by the Rejewski’s report brought by Langer? (cf. Subsection 5.1.1)
26 The probability is (1 - (1 / sqrt (e)) (SHD – DE 2016 ZB 25/6 Citation1949 [file 281]).
27 The set of Grundstellung to be tested for a Walzenlage was called "menu" by the Polish (SHD – DE 2016 ZB 25/6 Citation1949 [file 281]).
28 A document (SHD – DE 2016 ZB 25/5 Citation1949 [file 278]) indicates the unsuccessful tests against the Enigma TGD network. It is specified that the traffic was very low and that, at most, they found only eight females, which was insufficient to find the key, because some had to be eliminated due to the turnovers.
29 The example given by Carter does not use Steckers. Additionally, I do not use the Grundstellung TES, which causes a turnover for the Walzenlage III–II–I.
30 We can deduce the alphabets at p1 and P4: (AC)(BP)(DR)(EK)(FV)(GW)(HT)(IO)(JZ)(LM)(NY)(QX)(SU) (AN)(BK)(CF)(DR)(EO)(GS)(HZ)(IM)(JU)(LW)(PV)(QT)(XY) hence the couple (DR) is also clear to see.
31 The period mid-November is not compatible with the letter of 1 November 1939. The Cyclometer must have been completed much earlier, perhaps at the end of September.
32 Among these 137 people, there were about 30 cryptanalysts, including a third dedicated to breaking the Enigma.
33 Later, a multi-part Enigma message must use the same Kenngruppe for each part.
34 The excitement around the creation of the sheets in the cottage probably did not allow for the concentration necessary for Turing to invent his Bombe. It is more likely that he already invented it in the tranquility of Cambridge University.
35 It appears that Welchman's original intention was to provide entry for a secondary chain of constatations, not simultaneous scanning. Joan Clarke remembers Turing jumping up with the remark that "the diagonal board will give us simultaneous scanning" and rushing across to Hut 6 to tell Welchman. This must have been after 17 June 1940 when Clarke joined Hut 8.
36 My personal opinion is that the French–Polish team was unaware of the British Bombe. If that was the case, I am sure Rejewski’s report would have mentioned it (it mentions Herivel tip).
37 We see several similarities with the old system: the location of the dumb letters and the use of a book to find a trigram.
38 In reports from Hut 8, this machine was called Enigma O Bar. This designation does not appear in the Polish SHD report.
39 Maybe the Poles even broke the 29-key Enigma after the Army Enigma.
40 This was no doubt through Knox (cf. Subsection 7.5).
41 Bigram tables were published in sets of 9 and renewed from time to time. Only one was in effect per day.
42 Statistics: a term probably referring to the Zygalski sheets.
43 Perhaps, Denniston attempted one last thing: to refuse to send the rest of the Zygalski sheets to the French to force them to accept that the Polish experts needed to travel to England. It was then that Knox threatened Denniston to resign if he did not send these sheets to the French (Batey Citation2009).
44 In the French archives, the first key broken by the English was not completely dated. Only the month (January) is indicated (Figure 10).
45 The History of Hut 6 gives the red key of January 6 as a BP break. Perhaps both teams were successful. A slide on display at BP in 2010 suggests that this came out on January 22.
46 I think we can interpret Denniston's message as follows: a copy of the method for one day, perhaps, corresponds to the report of important information from the messages transmitted to HQ.
47 Instead of the “new machine,” you must understand the “new procedure.”
48 It does not appear that Herivel had access to the Enigma manuals (Lord Citation2019). If this had been the case, he would have had confirmation that the Walzenlage was configured before the Ringstellung, which was essential for his method to work.
49 Cillie is short for the German female given name Cäcilia. It is said that a German operator used the corresponding message keys (CIL and LIE) when thinking of his girlfriend.
50 Maybe W corresponds to the United States?
51 Arrow was rammed by Schiff 37 earlier that morning and directed Griffin to investigate Polares.
52 Wright (Citation2017) concluded that the pairs of dates April 23, 24 and April 25, 26 had been reversed so that the forms were for April 23, 24 which is why the Z serials start on April 23, 24.
53 Asché had warned the French SR that German troops would pass through the Ardennes (Paillole Citation1985).
Additional information
Notes on contributors
Jean-François Bouchaudy
Jean-François Bouchaudy is a Computer Scientist and Linux Specialist. His primary interest in cryptographic research is the study of cipher machines. He discovered them during his military service in French Army(CX-52, CD-57, Tarec, Myosotis).