GERMANY'S FIRST CRYPTANALYSIS ON THE WESTERN FRONT: DECRYPTING BRITISH AND FRENCH NAVAL CIPHERS IN WORLD WAR I

Cryptologia, Jan 2005 by Brückner, Hilmar-Detlef

The Royal Navy continued to use Gronsfeld. This naturally facilitated the work of Roubaix and led to a considerable increase in decrypts. In May 1915 W/T Command 6 reported the continuous solution of the telegrams of

* all coastal patrols, which used the a-p (auxiliary patrol) cipher, a 7-digit numerical key (in May 1915 it was 7412563');

* the Channel patrols, which used the a-p cipher for all reports concerning the fleet - which were easily identifiable, because they contained the keyword "lavan" ( "naval" written in reverse) - and the "transposed code" for everything else; and

* all telegrams in the ABMV(All British Merchant Vessel) Gronsfeld cipher, which used a 6-digit numerical key.18

In June 1915 the W/T Command 6 reported another success: the new three-digit Allied Fleet Code had been broken.19

By then Roubaix had a mechanical device that facilitated the finding of Gronsfeld keys and the decrypting of its messages: a rack that resembled a multiple slide-rule. It held 20 slides, on each of which were written one full and one incomplete alphabet, starting with the letter q, continuing with a full alphabet after z, then starting anew and going on to the letter j. Thus it was possible to visualize finding Gronsfeld keys: laborious counting of letters up and down the alphabet was replaced by moving the slides up and down till the new key had been found. Then a stencil was cut with square openings in the keying positions and inserted in fittings on both sides of its stand. Afterwards reading a Gronsfeld telegram was easy: once the key was known it sufficed to set the ciphertext at the baseline immediately beneath the stencil. The plaintext appeared in the square openings of the stencil.

A second mechanical decryption device at Roubaix was designed by lieutenant Hoffmann. A wheel carried vertically arranged bigrams, turning under a transverse-mounted strip carrying a vertically arranged alphabet. No information could be located in the archive as to how the wheel was operated and what results were obtained by it, if any. But its existence proves that Roubaix developed mechanisms to find unknown keys.

Another factor eased the difficult task of Foppl and his colleagues: involuntary help by the British.

At the end of April 1915 the decryption of a newly introduced key was made easy, because one of the British stations had asked: "please repeat your message in old code", and another one had cabled: "your message is undecipherable", whereupon the telegrams were sent once more in the old code. All Roubaix had to do was to compare the two versions to reconstruct the new key.20

In June 1915 Roubaix broke into the Allied Fleet Code, because close attention to the intercepted cables revealed that a weather report had been sent in its usual stereotyped form. A comparison with a preceeding one revealed that every word used corresponded to a three-letter codegroup, that both the plaintext words and the codegroups were arranged alphabetically, and thus both ran parallel to each other. It was not too difficult to compile the complete dictionary.21