English manual methods (May 1940)

• Cryptanalysis of the Enigma

Introduction

On 1 May 1940, a dramatic event occurred: the Enigma German encryption manual published in January came into force. A complete blackout ensued: no more deciphering of the Enigma was obtained (except in Norway, where the Germans kept the old system). This change in procedure completely masked the invasion of France starting on May 10. The Zygalski method developed by the Polish and based on double encryption of the message key was now obsolete. The Turing and Welchman Bombe would not be operational until several months later (August 1940).

Fortunately, German operators made errors that the brilliant cryptanalysts at Hut 6 exploited. They shared their methods with their counterparts at PC Bruno (French and Polish). Then, descriptions of these methods can be found in the French archives (see Rejewski’s report).

Manual methods consist essentially of the Herivel method and Knox's Cillies. Key rules reduce the number of tests to be performed.

German procedure of May 1940

Here is the beginning of the example message from the manual of 13 January 1940:

	1755 129 WEP HFI ULZNU SGEXU NFOPR SALME YDRJG
	QARZU BHFEM…

• 1755 The hour (17:55, 5:55 PM).
• 129 The number of characters of the message (the group containing the Kenngruppe included).
• WEP Grundstellung chosen by the operator.
• HFI The message key (XFR), encrypted only once at the WEP position.
• ULNZU The group containing the Kenngruppe (NUZ) and the two letters (UL) which complete the group.
• SGEXU… The message encrypted using the XFR key.

Periods of use for manual methods

From the moment they were discovered, manual methods were used to optimize the use of Zygalski sheets.

As early as May 1, 1940, cryptanalysts planned to use them exclusively to break daily keys. By the end of May, the methods were effective enough to break keys for the Air Force network. A Hut 6 report indicates that all keys for June were broken using these methods. In fact, they were used exclusively from May to August.

During August, the first Bombe (Agnus) came into operation. By June 1941, there were still only four Bombes. Consequently, manual methods continued to be used —either alone or to supplement the Bombes—depending on the networks being targeted (fewer than ten).

Note: Manual methods were the only ones available to the French team (comprising French, Polish, and Spanish members) that had taken refuge at PC Cadix. Thanks to these methods, they managed to break a few keys.

Herivel method (or Herivelismus)

Herivel, one day (in March 1940?), had an idea of genius. He put himself in the place of a German cipher clerk who was to deliver the first message of the day. He updates the Walzenlage and then the Ringstellung. Then, without turning the rotors, he encrypts his first message. If many cipher clerks do the same, the Grundstellung of the first messages would correspond either exactly to the Ringstellung of the daily key or to a very close value. For example, if the Grundstellungs of the first six messages on the same network are XNF, XNF, XMG, WME, YOF, and XNE, the Ringstellung is probably XNF (Herivel 2008).

Note: At first, I struggled to fully grasp the Herivel method. The videos I watched regarding Enigma setup described configuring the ring settings (Ringstellung) while the rotors were outside the machine. Indeed, if the rotors are inserted afterwards, there would seem to be no inherent link between the Ringstellung and the rotor positions. However, operators—as the manual confirms—actually perform this configuration AFTER the rotors have been placed in the machine. The operator rotates each ring until the letter which specifies the ring from the day's key is positioned roughly at the rotors'starting point.

Note: Extract of the German manual (1938)
13. In order to set the correct ring position (Ringstellung) (cf. 10b), the holding springs (IV, 22) of the number or letter rings (IV, 13) are lifted at the button (IV, 23) using the right hand, and the rings rotated with the left hand until the spring-loaded pivot (IV, 24) (which has a red mark starting with machine number 1253) snaps into the hole next to the number as specified by the codebook. The metal cover (II/II, 17) is now closed and fastened using the holding bolts (II, 18).

When Herivel explained his find to Welchman, it was tested to no avail. However, as soon as the cipher clerks were under the pressure of war operations, it ultimately did correspond to their practices.

Knox’s Cillies

The Cillies were discovered by Knox at the end of January 1940. Immediately, they reduced the processing times associated with Zygalski’s sheets by reducing the number of Walzenlages to be tested and were essential when the latter could no longer be used.

Note : Cillie is short for the German female given name Caecilia (Cäcilia). It is said that a German operator used the corresponding message keys (CIL and LIE) when thinking of his girlfriend.

Cillies are the result of a combination of two errors by German cipher clerks when encrypting a multi-part message :

1. After encrypting a message, the manual specifies that the operator must turn the rotors before encrypting the next message. Often, especially when the operator had to follow a frantic pace, he used the position where the machine stopped for the Grundstellung of the next part to be encrypted. This Grundstellung is clearly indicated in the header of the message. Therefore, we can deduce that it was the message key of the previous message. This position is obviously sensitive to the turnovers of the rotors, which on the one hand, complicates things, but on the other hand, makes it possible to make assumptions of the rotors on the right and in the middle.
2. It is the cipher clerk who chooses the message key at the position of the Grundstellung he chooses. Again, the manual clearly states that this key must correspond to a random trigram. Often, the cipher clerk did not follow these injunctions and used small words (UHR, DIE,…), pieces of words (HIT, LER, CIL, LIE…) or letters that follow each other on the keyboard of the Enigma (QWE, ASD,….). Only a few dozen values were common and could be guessed (and tested) by the cryptologist. For a given key, based on the Ringstellung deduced from the Herivel method and several Cillies (Cillie story), the cryptologist could restrict the Walzenlage to sometimes only three possibilities, if he could find the right rotor and the middle one, for example, V–III–I or IV–III–I or II–III–I.

The key rules

As early as the spring of 1940, cryptologists at Hut 6 assumed, without any certainty, the existence of key rules. They began to search the broken keys for any building rules followed by the enemy. However, to be certain, they had to accumulate many keys. The rules were used as soon as they were discovered. They made it possible to significantly reduce the tests carried out in conjunction with manual methods and when using the Bombe.

Red network rules

• a. The non-clashing rule: the same wheel could not be used in the same location for two consecutive days in the same month. This rule reduced the Walzenlages to be tested from 60 to 32.
• b. The rule of non-repeating, which was never absolute, simply provided that, within one month, the same Walzenlage was to be used only once.
• c. If during the first 26 days of a month the Ringstellung was written in three columns, each column contained all the letters of the alphabet only once. After the 26th day, the Ringstellung is arbitrary.
• d. The Consecutive Stecker Knock-Out (CSKO) rule: a Stecker cannot be done with two consecutive letters (A/B, C/D, E/F, etc.), but A/Z is allowed. e. Over a period of one month, the same Stecker-pairing should not be reused. The consequence is that the possible Steckers are limited to 300. In the case of a month of 31 days, repetitions are authorized during the first five days of the month.

Challenge

In my Enigma challenge, the final problem (Problem 24) can be solved using the manual methods just described (link).

References

Books & Articles

• Herivelismus and The German Military Enigma, by John Herivel, Editor: M&M Baldwin (2008).
• Solving Enigma's Secrets. The official History of Bletchley Park's Hut 6. Edited by John Jackson. Book Tower Publishing (2014).

Web sites

• Rejewski's report from the French secret service archives (in French). SHD (Service des archives de l’armée française à Vincenes) – DE 2016 ZB 25/6. 1949. Fond Bertrand – dossiers 280 à 285. The German version, file 281, and the French version, file 282.
  This report describes the Polish methods and The Herivelismus and Knox's Cillies method.
• Enigma manuals (1938 and 1940) - Bob Lord's Crypto Museum (link).