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

Surveillance Command's monthly reports for April, May, June, and August 1917 have survived.37 Their structure and content resemble to that of the August 1916 report of the Observation Office: low-level information on the activities of Allied warships, German submarines and their sinkings, Allied and German minefields, navigational obstacles and aids, nets, shipping routes, and Allied and neutral shipping. The reports were supplemented by maps indicating the positions of German submarines in the restricted area around England and France, in the Channel and in the Mediterranean; the April report, by additional maps on German mines in the Thames estuary and in the Straits of Dover, all as indicated by the Allies.

All monthly summaries provide information on the decryption of British ciphers and codes:

* In April, the British Admirality introduced a new code and used it for 135 messages on German minefields that it has dissolved and swept. More than 100 were solved. One of the ciphers formerly used by the British submarine base Rosyth (call sign: 83 u) was cracked. Retroactively decrypted cables revealed that British submarines patrolled the North Sea extensively.

* In May, a new edition of the Fleet Signals Book appeared. It was more complicated than any before. Nevertheless about half of the 150 British Admirality messages describing discovered German minefields and their sweeping were solved.

* In June, Surveillance Command noted that the two British naval signal books most frequently used had not been changed. It commented: "Apparently the enemy held the new extensively enlarged unalphabetic a.-f. code [auxiliary force] and the patrol code, which was changed daily, to be sufficiently secure".

* In August, the British 5-letter Fleet Code and the British 3-letter Fleet Signals Book were solved. Both had been introduced simultaneously at the beginning of August. This had been intended to make it impossible to start decryption by comparing messages. The command explained that therefore it had taken some time to crack them.

British signals in a. f. 6 used as a submarine code had been decrypted, too, partly retroactively. This permitted compiling a list of all British submarines of the Yarmouth and Harwich stations used in operations in the southern part of the North Sea since April 1917.

In that period French naval ciphers had been decrypted, too:

* In June, the newly introduced threefold superencipherment of the French Naval Signal Book had been solved.

* In August at the latest, the French universal system (chgt) had been deciphered. Surveillance Command reported at the end of that month that it had not been changed, because it was apparently thought to be sufficiently secure.38

However, decrypting was becoming more and more difficult. Surveillance Command warned in June that the British knew about the constant surveillance of their W/T traffic and tried to escape it by using techniques which were becoming more skillful all the time. In August it signaled that the British increasingly limited the forwarding of information by telegram, were much more careful when doing so, perfected their ciphers, and had changed all naval call-signs, which impaired the tracking of the British warships considerably.