by Nigel West
The Yugoslav double-transpositions used by Draža Mihailović and his Chetniks were solved by Corporal Herzfeld of Inspektorate 7/VI, usually within one to three days. He had found that in each message the same key was used for the two transposition matrices, and the matrices were usually incompletely filled. The width of the transposition matrices were assumed by Corporal Herzfeld each time with some degree of accuracy, based on his previous experience with messages of the same length and on the same networks. This permitted a marking off in the cipher text of the approximate columns of the second transposition matrix. The words ‘General Draža Mihailović’, which nearly always appeared as a signature, made an excellent crib and often provided access to the rest of the text.
Cryptanalytic successes against American strip ciphers were obtained by at least three German agencies. Dr Rohrbach, the Pers Z S cryptanalyst, claimed that his group of six cryptanalysts solved the 0-2 US State Department strip cipher in 1942 without any previous knowledge concerning the general system, but required over a year for solution. Rohrbach’s methods were recorded in a detailed paper written by him at the request of his TICOM interrogators and these were similar to the statistical methods suggested as a general solution by US Army cryptanalysts in 1944 when messages had been punched on IBM cards to identify the repetition so valued by cryptographers. Although Rohrbach indicated in his paper that he had had no previous knowledge of the general system when he started his solution activities in November 1942, it was known that Pers Z S had received photographic copies in 1941 from the OKW/Chi of a set of instructions for the earlier 0-1 American strip cipher and ‘four series of strips by means of which a number of messages could be deciphered’. Pers Z S files captured in 1945 also contained photographs of two Tables of Numerical Keys for the same earlier cipher. The advance knowledge of the general system that Dr Rohrbach indicated in his paper was present in 0-2 State Department traffic (such as ‘Strictly confidential from Murphy …’) together with the State Department’s reuse of each daily key an average of nine times throughout the year (only forty different strip arrangements were provided for 365 days’ traffic) should have made the solution much easier.
Dr Hüttenhain stated that the OKW/Chi worked on American diplomatic strip systems, but had not been helped by any ‘pinches’, and their successes had been achieved by sheer cryptanalysis. Accordingly, the OKW/Chi had built a special rapid analytic machine (the statistical depth-increaser) for facilitating statistical solutions. According to Voegele:
It was quite evident from the cipher text that there was a break after each 15 letters … Accordingly an analysis was made on the basis of groups of 15 letters with the assistance of IBM machines. A depth of 80 passages of parallel construction was needed to reconstruct the 100 strips, 30 of which were valid in any one day … The system was read as long as it was used.
In 1943 a new difficulty presented itself. While thirty strips were still valid on any one day, the encipherer could arbitrarily remove any five of the strips to encipher any one message … After about six weeks, some of these messages were also deciphered. However, at the same time the volume of this type of traffic began to decline, so that finally the analysis work had to be discontinued.
The techniques employed by Voegele and his assistants are not known. Decipherment after about six weeks of some of the later messages, when strip elimination was employed, may have been accomplished by the skilful use of cribs. It is interesting to note that soon after strip elimination had been introduced, ‘the analysis work had to be discontinued’.
Another device, the automaton, was also developed by Pers Z S for rapid deciphering of a large backlog of strip traffic. Major Rudolph Henze, head of the cryptanalysis group of the OKH/GdNA, reported the solution of an American ‘strip’ cipher, the intelligence of which was ‘mixed military and diplomatic’ and which, from the description of the system and its indicators, was actually enciphered by the US Army Type M-9 device rather than by any of the strip cipher devices. Another interviewee, Werner Graup, claimed that he had cryptanalysed an American ‘strip’ cipher (actually the M-94) carrying Iceland and Caribbean area traffic, while he was working for Inspektorate 7/VI in Berlin. He used cribs, with the help of synoptic tables, to determine the ‘strip’ (disk) orders, and stated that he later believed IBM methods were developed to eliminate impossible keys. According to the interrogator, Graup knew of what he called a thirty-strip system, but stated very definitely that it had never been solved. Colonel Mettig, who commanded Inspektorate 7/VI between November 1941 and June 1943, recalled ‘… it was eventually recognised that the main cipher procedure used by the Americans was the strip method whereby twenty-five variously arranged alphabets were vertically laid out one alongside the other. In the workshop of In 7/VI mechanical aids were constructed and with the help of the IBM section and by noting the addresses and signatures, the various alphabets were recreated.’
TICOM judged that German cryptanalysts had been very successful on grade systems, partly because they were sufficiently skilled to take advantage of the presence of low-security traffic, but also because the Allies had failed to realise how insecure their procedures really were. However, it was also evident that German cryptanalysis was unsuccessful when deployed against high-security systems. Reassuringly, TICOM’s assessment of the evidence suggested that the Germans had been unable to read any US Army or Navy high-level systems. According to the accumulated data, the Army Converter M134C, known as the Sigaba; the Army Teletypewriter Cipher Attachment, known as the Converter M-228; the Army Teletypewriter Privacy Set; the Army High Security Teletypewriter Cipher System; the Army Speech Equipment RC-220-T1; the Combined Cipher Machine (CCM); and the Navy’s Electric Cipher Machine Mk III had proved to be completely secure. One Army strip system (System Ho. 47 or 67) and one Navy strip system (probably C. S. P. 1404) had been read for short intervals until the principle of strip elimination had been introduced. The low-grade telephony device (Speech Equipment AM/GSQ-1, or SIGJtP) was not read, although theoretical solutions were worked out. However, both of the unenciphered War Department Telegraph Codes (SIGRIM and SIGARM) were read by German cryptanalysts. Hungary received photostat copies of War Department Confidential Code Number 2, probably from the Bulgarians, together with at least one set of cipher tables, and the Italians reconstructed subsequent editions of the enciphering tables. The compromise appears to have been shared with other Axis powers, including Germany, Finland and Japan. Additionally, the US Military Intelligence Code No. 11 had been physically compromised in a burglary, but was still used by Colonel Bonner F. Fellers between October 1940 and August 1942 to report on current and future British military operations, after he had attended General Claude Auchinleck’s daily staff conferences at GHQ in Cairo. The Germans also read messages in several versions of the US Army Division Field Codes.
TICOM estimated that German cryptanalysts had solved from 10 per cent to 30 per cent of intercepted US Army M-209 messages. Except where the keys had been captured, they were usually read too late to be of any tactical value. Almost 100 per cent of messages sent by the US Army in Slidex, Codex, Bomber Code, Assault Code, Aircraft Movement Code, Map Co-ordinate Codes, and Cipher Device M-94, were read regularly.
Chillingly, the researchers found that virtually all the messages sent with the Combined Naval Cypher No. 3, used by the US Navy and the Royal Navy for the Atlantic convoys, were read from the end of 1941 until the middle of 1942. During these months of compromise, Allied convoy shipping losses were six times greater than in any other comparable period. In summary, German T/A and cryptanalysis provided a comprehensive order of battle for the US Army and Army Air Forces in the United Kingdom, in the Mediterranean and, after D-Day, on the Continent, and, according to a captured Luftwaffe officer, ‘no attack of the Eighth Air Force came as a surprise’.
Having captured the Balkan Archive at the Schloss Bergsheidingen, near Leipzig, TICOM concluded that one focus of German attention had been on Soviet military commun
ications. Some success had also been achieved in the solution of medium and low-grade British military and naval communications. In addition, the diplomatic traffic of Italy, Japan, France, Turkey, Bulgaria, Greece, Portugal, Spain, Switzerland, and other smaller nations had been read routinely.
‡
The Wehrmacht’s skill in assessing the Allied order of battle had been recognised in the aftermath of the disastrous Dieppe raid in August 1941, when one of the ill-fated operation’s priority objectives had been the recovery of useful intelligence from a command post manned by the 371st Infantry Regiment. Specially briefed commandos had seized the position and retrieved a quantity of Wehrmacht paperwork, including a large volume, the 457-page Das Britische Kriegsheer, which listed every component of the British Army and its Empire units, complete with large fold-out maps. Intelligence analysts in London studied the book, dated 10 April 1942 and issued by the OKW in Berlin, and invited MI5 to make an assessment of the content, which was judged to be quite accurate. Where had the Germans acquired their detailed knowledge? If SIGINT had been responsible, then transmitting procedures needed to be tightened up. If there was an uncontrolled spy at liberty, the leak had to be plugged urgently.
Since July 1941 the Abwehr had been fed a large amount of misleading material by a double agent in Cairo, code-named CHEESE. Known to his Abwehr handler Kurt Knabe as ROBERTO, Renato Levi was an Italian Jew from Rapallo and the legitimate holder of a British passport who had developed a taste for espionage while working for the British Secret Intelligence Service (SIS) and the French Service de Renseignements in Paris during the weeks before the Nazi occupation. Having duped the Abwehr and gained their trust, Levi had been sent on a mission to Egypt to rebuild the Axis spy rings that had been rounded up and interned when Italy joined the war. Guided by an imaginative staff officer, Dudley Clarke, and a case officer from Security Intelligence Middle East, Evan Simpson, Levi had acted as a conduit channeling bogus information to the Abwehr wireless station in Bari. Not only did he transmit messages with a fabricated content, he created a non-existent network of sub-agents willing to act as mercenaries. To sustain the charade, Simpson, who had been a theatrical producer and Spectator reviewer, introduced CHEESE’s fiery Cretan girlfriend, and a Syrian deputy, Paul Nicossof. Much to everyone’s surprise, ISOS intercepts proved that ROBERTO was held in high regard by his German controllers, and Knabe’s status within the organisation rose accordingly.
Using CHEESE as a guinea pig, Clarke followed his commander-in-chief Archie Wavell’s request to exaggerate Allied strengths across the Middle East, and this he accomplished by inventing entire divisions, such as the 2nd Indian Infantry Division, 10th and 15th Armoured Divisions, the 74th Armoured Brigade, the 7th Infantry Division, and the 27th, 38th, 39th and 101st Battalions, Royal Tank Regiment. Other units were simply misrepresented so as to construct a thoroughly misleading order of battle. Thus the 9th Army, which was no more than at corps strength, deployed against the Vichy French, became a full army, as did the 10th Army, which supposedly was sent to the Caucasus to support the southern Soviet flank in August 1942. Also invented was the 12th Army, components of which were drawn from GHQ’s substantial presence in Cairo and the Delta. The illusion was completed by Clarke’s deputy, Noel Wild, who created a range of unit insignia, including a seal balancing a globe on its nose, which was duly reported by CHEESE. Another of his inspired ideas was the distinctive unicorn as the sign of an equally fictitious 12th Army Division.
The significance of Das Britische Kriegsheer, being the standard reference work for the Reich’s armed forces, and classified as secret, is that it dutifully listed Dudley Clarke’s fake military formations, including his favourite, the 1st Special Air Service Brigade, which he had dreamed up to support the British offensive in Libya by pretending this elite parachute brigade was to lead an attack in Somaliland. The scheme, code-named CAMILLA, failed in January 1941 when the Italians unexpectedly withdrew instead of assembling reinforcements, but the 1st SAS had definitely been established in the enemies’ eyes. Study of German documents captured in North Africa and Italy proved that the OKW had overestimated British Middle East forces by 30 per cent.
The OKW’s handbook was typically detailed, methodically cataloguing every regiment, division and corps, and naming their last confirmed location. The mistakes were few, but one was to prove useful. According to Das Britische Kriegsheer, the 76th (Norfolk) Infantry Division, which consisted of 213, 220 and 222 Brigades, included the 18th Battalion of the Welch Regiment, whereas in reality the brigade was composed of the 11th Royal Scots Fusiliers and the 18th Battalion, the Royal Lincolnshire Regiment. So how had the error occurred? MI5 learned that in fact this particular battalion, with no connection to the locally garrisoned 76th Division, had been engaged in airfield perimeter security duties near Norwich during 1941. This meant that either radio interception or the personal observation of a spy could have shown the unit’s deployment in Norfolk and the issue was potentially grave because by March 1942 MI5 was confident that not a single German agent was at liberty in England. Those enemy spies that did exist were all under MI5’s control, so the possibility that there might be an undiscovered agent who was free to roam the country and check on other spies had a chilling impact.
After a lengthy investigation, MI5 blamed poor signals security, which had allowed the enemy’s direction-finding and traffic analysis experts to reconstruct the British order of battle and the deployment of its forces. The fear of an uncontrolled source was removed, and new measures were introduced to improve the army’s lax communications security.
In retrospect, the capture of Das Britische Kriegsheer was an important turning point for the OVERLORD planners as it provided a sound foundation of what Fremde Heere West knew and understood about the British order of battle. It also demonstrated that the enemy could be duped, as had happened so vividly in the Middle East through CHEESE’s mischief. Another reasonable conclusion was that the Germans appeared to have no grasp of the concept of strategic deception. ISOS revealed that Abwehr I personnel were always alive to the dangers of double agents, but never contemplated a sophisticated co-ordinated campaign.
The OVERLORD plan considered by Freddie Morgan in May and June 1943 was quite breathtaking because of its lack of orthodoxy in every respect. Yet Morgan himself was quintessentially a conformer. Born in Kent, and educated at Clifton, he graduated from the Royal Military College, Woolwich. He had served in the First World War as an artillery battery commander and then in staff positions, and remained in the army during the inter-war period, spending a long period in India. He was not promoted to the rank of major for fifteen years, but returned in England in 1934 to command an anti-aircraft battery. When posted to the 3rd Infantry Division as GSO1 he served alongside Bernard Montgomery, and then fought in France with the 1st Armoured Division before being evacuated in May 1940. There was nothing in Morgan’s career that suggested he was an outstanding leader or intellectual giant, but his steady soldiering, with action in two conflicts in France, seems to have prepared him for the monumental political and military challenges represented by OVERLORD, an Anglo-American cross-Channel adventure that was set be executed in May 1944.
It was Morgan’s plan that was presented to Churchill as he lay in bed in his spacious stateroom aboard RMS Queen Mary moored in the Clyde on 4 August 1943 preparing for the five-day voyage to Halifax, Nova Scotia, to confer with President Roosevelt at the QUADRANT meeting in Quebec.
Morgan’s experience as a planner, apart from major exercises conducted in peacetime in India, was as commander of Force 125, a contingency plan to land troops on Gibraltar from Algiers to defend the Rock in the event of a German attack through Spain, and another plan, for an Allied landing on Sardinia. Thus, through chance, Morgan came to know the two men who eventually would take control of OVERLORD, Dwight D. Eisenhower as supreme commander, who had commanded the US forces in Algiers, and Monty as commander of the ground troops. It would be Monty that escalated the inva
sion plan by demanding an assault by five divisions, to be followed quickly by two more, and not the three originally contemplated by Morgan. Furthermore, he widened the front from 26 miles to 40, citing the fear that the Allied troops might be trapped on beaches with narrow exits, blocked by wrecked vehicles.
Despite these changes, Morgan’s basic scheme was accepted, and he graciously stepped aside and offered to serve as Walter Bedell Smith’s deputy when his role as a planner appeared to come to an end, and Monty vetoed his promotion to command an army corps. Instead of taking on a combat role, as he wished, Morgan was posted to Norfolk House, now designated as SHAEF’s headquarters in St James’s Square, where he would eschew the limelight and liaise with the deception planners and the intelligence services. Although Dudley Clarke was reluctant to move from Egypt where his group of deceivers, grandiosely self-styled ‘A’ Force, was making a growing contribution across the Middle East, his deputy Noel Wild went to London in December 1943 to take over from John Jervis Read a mirror-imaged organisation, designated Ops (B), which would carry out the schemes approved at a strategic level by TWIST, a dedicated sub-committee headed by Johnny Bevan, then also in charge of the London Controlling Section (LCS).
Experimental, and located in two residential flats directly below a brothel at 6 Sharia Kasr-el-Nil in central Cairo, ‘A’ Force would never employ more than five officers but nonetheless had pioneered the concept of strategic deception and given General Archie Wavell a much-needed advantage against the Afrika Korps. Whereas CHEESE had been regarded initially as something of a joke, his strategic impact was undeniable, and hence his value.