Chapter VI
Compromise of Allied Codes and Ciphers by German Naval Communication Intelligence

  1. German naval intelligence: "Convoy Expected;" its importance.

  2. Its effectiveness.

  3. Source acknowledged in U-boat traffic. "B-Dienst."

  4. Suspicions of cipher compromise.

  5. Difficulty in proving cipher compromise.

  6. Compromise of Naval Cipher #3 demonstrated, May 1943. Naval Cipher #5 introduced 10 June 1943.

  7. Compromise of Naval Cipher #5 demonstrated, October and November 1943.

  8. Extent of German success with Naval Cipher #3.

  9. The delay in Allied counter reaction.

  10. Other German Cryptanalytic Successes.

  11. Failure with US Navy Ciphers and the British Type X.

  12. Conclusion: Effectiveness of German naval communication intelligence.


Compromise of Allied Codes and Ciphers by German Naval Communication Intelligence

  1. German naval intelligence: "Convoy Expected;" its importance.

    From the beginning of the Atlantic Section's reading of current U-boat traffic it was apparent that the enemy made good use of intelligence on Allied ship and convoy movements. The "convoy expected" messages soon became a familiar part of recurring patterns of U-boat groups, not only in the messages which formed the groups and ordered their patrol lines or courses for sweeping. It was not surprising that group tactics should appear in traffic to be dictated throughout by knowledge of convoy movements, for U-boat warfare in open sea was by its nature peculiarly dependent upon convoy intelligence. From their experiences in the First World War and certain German naval authorities had long since come to the conclusion that the combination of group attack with convoy intelligence would provide an ambush in mid-ocean that could defeat the confoy system. Admiral Bauer, World War I Commander of U-boats until the summer of 1917, wrote in 1931 that:

    the success of the convoys was owing essentially to the fact that they succeeded in reaching their destination for the most part unnoticed, or damaged by single submarines only moderately in proportion to the large number of ships. Here the convoy's unexpected arrival was just as important as its guard by warships, which had been the decisive factor for the convoys of earlier times. If it had been possible to determine the presence of convoys and report them early enough so that a fairly large number of submarines could join the attack, even a strong guard could not have protected the group from heavy losses.

    Reconnaissance and the assembling of as large a number of U-boats as possible for the attack were the only suitable means of defeating the convoy system with respect to


    both the element of surprise and the military strength of the convoys.

    As one method of preventing a convoy from entering the "combat area" unnoticed, Admiral Bauer suggested U-cruisers which could lie in wait outside ports of departure and shadow the convoy on its ocean route. He also recognized, in part, the role radio intelligence might play in convoy locations, although

    The whole conduct of the war so far as the use of the radio message is concerned was still primitive and harmless.

    Germany had no difficulty, he says, in reading the simple cipher which England used when she began "... to guide commercial and escort vessels by radio, in accordance with the intelligence she gained about German U-boats."

  1. Its effectiveness.

    It was obvious that the U-boat groups were being moved with intelligence during the winter of 1942-43. The only questions were:

    1. How accurate and timely was their information, and
    2. What was its extent and source?

    These questions could not at first be answered here with any confidence, and if the British could have done so their information was not available to the Atlantic Section. Preliminary surveys undertaken shortly after the Atlantic Section began to read traffic did not permit and specific conclusions which could explain the individual "convoy expected" messages. There were cases in which U-boat Command seemed well informed up to the time of convoy diversion. The day of expected meeting with the convoy would arrive--but no sighting reports. Command would stir his boats and urge sharper lookout while ordering a change of course.


    Then, again there were cases in which everything worked out according to plan.

  1. Source acknowledged in U-boat traffic. "B-Dienst."

    Sources of intelligence disseminated to U-boats were sometimes acknowledged: agents, submarine and aircraft reconnaissance, prisoners, and the intercept net. Very little was known about the German communicatino intelligence organization, except that its mark appeared occasionally in U-boat traffic as "B-Dienst," tagging a repetition of plain language reports which were relevant to U-boats at sea. It was assumed that the Germans were working on our ciphers systems, but nothing further was known. Had the Atlantic Section made no attempt to correlate the traffic it processed, the suspicion of compromise might have been aroused with difficulty, for the "convoy expected" messages were routine. The U-boat naturally "expected" convoys, and general information such as convoy departures, cycles, and other related matters could scarcely be kept from an industrious enemy. Furthermore, Allied North Atlantic convoys observed no such thing as radio silence. On the other hand, U-boat "expectations" were often disappointed, precisely when we began to get a good first decryption hold on U-boat traffic. If our radio intelligence was effective in convoy diversions, then it seemed that the enemy's was not effective in countering these diversions.

  2. Suspicions of cipher compromise.

    The sudden shifting of lines for U-boat groups Neptun-Ritter and the formation of Knappen, all ordered within the space of 30 minutes on the night of 18 February, led not only to a very successful operation against convoy ON 166 but to very definite fears here of Allied cipher compromise. It was another matter, however, to prove the insecurity of the cipher.

  3. Difficulty in proving cipher compromise.

    It may be difficult to demonstrate with certainty that ones own ciphers are being broken by the enemy--even when one is in turn reading part of the enemy's traffic. For, when properly used, special intelligence is sufficiently disguised so sources other than actual decryption of a specific message can


    not be eliminated. The enemy is presumed to possess information whose possible sources are so varied as to come under the cognizance not of one but several different counter-agencies. Without complete knowledge of ones communications systems (including ciphers), communications h abits, actual movements, and plans, it may not be possible to judge the source of the intelligence appearing in enemy traffic, unless the enemy fails to observe elementary precautions. By such failure he can yield a self-evident proof, as would be furnished by a direct quotation whose origin can be traced solely to one dispatch. Without such indubitable proof that a major cipher is compromised there is a natural reluctance to instigate and authorize the complicated changes and expenditures which would be necessary if a reading enemy is to be forestalled.

  1. Compromise of Naval Cipher #3 demonstrated, May 1943. Naval Cipher #5 introduced 10 June 1943.

    While the weakness of combined Naval Cipher #3 was appreciated in some circles it was not until May 1943 that its compromise was so thoroughly demonstrated that action could be taken. In the middle of May 1943 the Germans offered the necessary proof of combined cipher compromise in the form of three Offizier messages to groups Rhein, Elbe, and Drossel during the course of their patrol and action against convoys HX 237 and SC 129. Convinced that cipher compromise was involved, the Atlantic Section gained access to Commander 10th Fleet Convoy and Routing files. The compromised Allied dispatches were identified and the evidence was submitted to COMINCH. The conclusions were accepted. An exchange of signals followed between COMINCH and Admiralty involving proposals for additional security precautions, but the marked increase in similar compromise messages in German traffic toward the end of May led to the introduction of Naval Cipher #5 on 10 June.

  2. Compromise of Naval Cipher #5 demonstrated, October and November 1943.

    Beginning in September with convoy operations, the familiar signs of convoy intelligence were once more in evidence. After having repeatedly expressed its fears, the Atlantic Section conclusively demonstrated the compromise of


    Naval Cipher #5 on one occasion in October and one in November. COMINCH and the British concurred.

  1. Extent of German success with Naval Cipher #3.

    1. Tables "M" and "S".

      The demonstration of compromise in May led to a thorough examination of all previously suspected cases in German traffic, and examination which was now possible with access to the convoy files. From these studies, together with conclusionns which GC&smp;CS made available at this time, it was evident that during the first half of 1943, German cryptanalysts had read extensively and with usable currency in the world-wide ("M") and the North Atlantic area ("S") tables of the combined cipher (Naval Cipher #3). From the interrogation of Italian communication intelligence officers following the capitulation of Italy, it is known that the German and Italians cooperated in their attack on the combined cipher. Copies of the basic book had been obtained and were known to the enemy under the code word "Frankfurt" and "Francoforto." The reciphering tables were identified. With the aid of the basic book, messages were usually readable in part on a depth of three, sometimes even on a depth of two. After the introduction of a new basic book, five to seven months were necessary before good progress could be made, and this depended on luck and Allied mistakes.

    2. Weakness of the combined cipher.

      The combined cipher was in practice a weak system. In the first place, it was overburdened, as might be expected, for it was the main link in all dispatches concerning convys, wherein, of course, the emphasis of the Atlantic lay. The North Atlantic had been turned into a vast freighting area in which the movement and portection of convoys involved a constant


      exchange of dispatches by radio. Running through the whole thing was a regular pattern of standard announcements: sailing telegrams, ocean route, ocean meeting point, escort relief, stragglers rendezvous, convoy position reports, etc. If any U.S. Navy command was on the address for information, as was norall the case, the combined cipher was used. In the second place, combined cipher involved use of combined call signs from which the addressees could be determined and the general nature of the content deduced.

  1. The delay in Allied counter reaction.

    Although the British had recorded cases of combined cipher compromise prior to May 1943, no action was taken until the demonstrations of May. It is worthwhile to list here some of the problems and difficulties which prevented an earlier demonstration on the part of the Atlantic Section.

    1. Cryptanalytic: The difficulties experienced with the ENIGMA general traffic during the first part of 1943 were even greater in the case of Offiziers, or messages sent in a double encipherment. Offizier messages contained the only identifiable cases of compromise.

    2. Lack of convoy dispatches: Although the Atlantic Section id receive copies of routine convoy dispatches for information, it had had to fight even for daily convoy position estimates. It did not receive the significant diversion and routing dispatches. Even with access to the convoy files, the problem was not solved, for it appeared doubtful that there was anywhere a complete file of Allied transmissions.

    3. General:: A more fundamental difficulty was the system of combined communications itself. U.S. Navy-British naval communications were so complex, and often repetitious. that no one seemed to know how many times a thing might not be sent and by whome--and in what systems. it is possible that the question of cipher compromise might have been settled earlier than May had the Combined communications systems been less obscure and had there been closer cooperation between the British and the US in such matters.


  1. Other German Cryptanalytic Successes.

    In addition to Naval Cipher #3 and #5, the following systems are known to have been read by the German naval Communication intelligence organization:

    1. Various British Naval and Air codes (low grade): including COFOX, MEDOX, FOXO, LOXO, SYKO, Air Force code and Aircraft Movement code.

    2. U.S. Hagelin: A Hagelin reencryption of 26 September 1943 was read by the Germans. The compromise was noted by the British from their reading of German Mediterranean traffic.

    3. Anglo-French.

    4. Merchant signals: Supplied with the books from merchant shipping abandoned or sunk, the enemy had no difficulty with Mersigs.

    5. Russian: aircraft reporting system, weather system, and low grade naval traffic.0

  2. Failure with US Navy Ciphers and the British Type X.

    There is no evidence that the Germans ever had any success with high grade machine ciphers. If the Japanese Naval Attaché's reports can be believed on this point, the German Navy knew very little about U.S. Navy ciphers and needed assistance from the Japanese as late as summer of 1944 in identifying U.S> Navy traffic. Italian communications intelligence officers stated that their organization did little work on U.S. Navy ciphers since they believed that anything of importance to them in the Mediterranean area would be found in the combined cipher. According to Commander Porta, neither the Germans nor the Italians had attacked British cipher Type X to any extent up to the time of the Italian armistice. The


    Germans at that time, however, decided to work on it, but as far as is known they had no success.

  1. Conclusion: Effectiveness of German naval communication intelligence.

    Reliable information on the German Navy's successes in decryption as at present confined for the most part to what has been discovered through reading German naval traffic itself, i.e. to those messages which contain statements traceable to Allied dispatches. Some good information has been obtained from Italian captured documents and from Italian naval communication intelligence officers. The references in Japanese Naval Attaché traffic to this subject are too general to be of much assistance. It is hoped that captured German documents will eventually offer material for a complete and detailed account of German cryptanalytic successes. The importance of a full treatment cannot be over estimated as a future safeguard against complacency and ineffective organization, whether at home or among Allies.

    According to present information, the German naval communication intelligence organization achieved only one major success against "high" grade Allied naval ciphers, namely the combined British-U.S. Navy cipihers using tables "M" (world-wide_ and "S" (North Atlantic). A successful attack on combined cipher gave the German Navy the kind of information they most needed for offensive operations: convoy intelligence for use of U-boats. On the other hand, their success, such as it was, in this kind of intelligence could not compensate for their lack of insight into Allied naval operational traffic such as was carried by machine cipher. In consequence, there is no indication that the enemy ever had reliable advance information on any major Allied amphibious undertaking. Apparently dependent on analysis of alow-grade traffic, such hints as the German Navy had of these major operational movements came too late--before adequate counteraction could be taken. Nevertheless, within the realm in which it concentrated, its successes undoubtedly increased merchant shipping losses, and thereby contributed to a lengthening of the war.


Table of Contents ** Previous Chapter (5)

Transcribed and formatted for HTML by Chuck Roberts for the HyperWar Foundation