Double Cross in Cairo

by Nigel West

  In another TRIANGLE text, on 23 February, Zähringer was told that CHEESE was worried about his funds, and was directed to deal with the matter. On 18 March a message to Istanbul demanded to know whether Zähringer had solved the problem and, if not, why not? On 29 March Athens sent an urgent request to Rossetti in Istanbul, where he had arrived on 22 January, to provide an address in Syria where CHEESE’S money could be made available. This development fascinated SIME, which speculated that the Abwehr might be contemplating the use of PESSIMISTS or even QUICKSILVER as an alternative to the Aleppo route proposed by CHEESE the previous September. Analysis of TRIANGLE later revealed that PESSIMISTS were also controlled by Rossetti.

  The loss of U-372, a Type-VIIC of the 23rd U-boat Flotilla, off Haifa in August 1942, which was carrying his courier, Jawad Hamadi, codenamed HAMLET, also prevented further funds from reaching CHEESE. The entire crew of forty-eight, and the sole passenger, Hamadi, were captured by the Royal Navy, one of fourteen U-boats sunk in the Mediterranean in 1942. Under interrogation Hamadi, a Druze student, admitted that he had been recruited by the Sensburg organisation in Athens to deliver money to CHEESE. When captured, Hamadi had been wearing a money-belt containing $3,500 and £500 in other currencies.

  According to Walter Sensburg, who was questioned about HAMLET after the war,

  HAMLET was recruited by Rossetti in Rome and transferred to Athens when Rossetti joined the Ast. An intelligent youth, probably an Arab, HAMLET had studied in Rome. He served the Germans out of idealism and would accept no compensation. HAMLET left Rome in 1942 aboard a submarine which was to take him to the Syrian coast. The submarine was sunk but, according to Rossetti, HAMLET survived and was interned by the British in Cairo. Nothing further was heard of him.

  The voyage of the U-372, which sailed from the Greek island of Salamis on 27 July, commanded by Captain Heinz-Joachim Neumann, was compromised at the outset by TRIANGLE and an ambush was prepared in which the submarine was attacked with depth-charges launched by the destroyers HMS Sikh and HMS Zulu and the escort destroyers HMS Croome and HMS Tetcott, and by depth-charges dropped from an RAF Wellington from 221 Squadron. The U-boat’s entire complement was placed aboard the Tetcott and disembarked in Haifa where they were questioned by CSDIC personnel. On 17 August Hamadi, when cross-examined by SIME, admitted his role, disclosed his call-signs and radio schedules, and volunteered to work as a double agent. According to him, he had been instructed to make his way to Ranlkin, a village about 45 kilometres from Beirut where he was to set up his transmitter and contact Athens on behalf of a man he knew only as ‘Paul’. SIME subsequently reported that

  the fates were unkind when, for a second time, a carefully-prepared plan had been made by CHEESE for his ‘petite amie’ to receive the money at a suitably selected café on the Pyramids Road in October, the enemy were distracted and apparently compelled to abandon their side of the plan because of the shock of the British victory at El Alamein.

  The next major Allied deception in which CHEESE was to participate was TREATMENT, a plan which suggested that Crete was about to be invaded. On 13 and 15 October 1942 CHEESE reported that troops would be landed on the island in early November, and on 17 October he described a commanders’ conference called for 26 October, and on 18 October he said that Montgomery also planned a simultaneous offensive in the south of the desert, and declared that the attack on Crete was scheduled for 8 November. In response, on 21 October Hitler ordered the 22nd Infantry Division’s occupation garrisons in Crete to be reinforced with valuable troops that might otherwise have gone to the Afrika Korps.

  While working on TREATMENT, CHEESE was also setting the foundation for the cover story for TORCH, the huge Allied landings in Morocco and Algeria in which 1,500 ships would deliver first 90,000, then a further 200,000 troops to North Africa. The planners presumed that the preparations for what was to be the largest amphibious operation ever conducted would be impossible to conceal, so they concentrated on offering alternative objectives. Codenamed KENNECOT, the campaign fell into two parts, with the first suggesting that Sicily and Italy were the targets, and TOWNSMAN, centred further east, implying that Crete was the likely objective of the massive invasion fleet which assembled in the Atlantic before venturing through the Straits of Gibraltar, which were under continuous scrutiny by the enemy. On 31 October CHEESE reported that the Allies intended to attack Italy, with Crete as a secondary target, and predicted large-scale American air raids. In fact, of course, the TORCH beachheads were secured with ease on 8 November, the Axis having been taken completely by surprise.

  Following the success of TORCH, and the liberation of Algiers, the local Deuxième Bureau agreed to cooperate with the British, and the result was the transfer from Cairo of Cuthbert Bowlby’s ISLD headquarters, and the opening of an SIS station under Arthur G. Trevor Wilson and his energetic subordinate, John Bruce Lockhart. The French turned out to be running a stable of about a dozen double agents, among them RAM, a French NCO who was notionally employed at the French GHQ; a Spanish officer codenamed WHISKERS who was in contact with the Abwehr in Morocco and had created an entirely bogus network of sub-agents; and GILBERT, a French army officer and St Cyr graduate who had been infiltrated into Tunis in April 1943 to act as a stay-behind agent and was in radio contact with his Abwehr controller.

  The Deuxième Bureau also controlled a SIM agent, LLAMA, operating from Tripoli and responsible for conveying a false order-of-battle to the Italians. SIM’s efforts in the region had not achieved much success, their large stay-behind network in Tripolitania had been rounded up. They were also discouraged by an attempt to drop three agents by parachute near Damascus in December 1942 who were to establish radio contact with a station on Rhodes.

  Levi was released from Tremiti on 15 May 1943 and, due to ill health, was put in hospital in Foggia but, due to heavy bombing, was transferred on 19 August to the military hospital at San Severo. In Cairo these events were monitored through TRIANGLE, and SIME reported that ‘at the beginning of April 1943 we received information from a most secret source that Levi was to be released from imprisonment, and Rossetti was asked by Scirombo of the Italian Intelligence Service what should be done with him. There was an obscure reference to his being transferred to the Germans. Rossetti requested that Levi should be sent to Sofia to be at Rossetti’s disposal. In mid-May 1943 Rossetti in Istanbul asked Athens if Levi could be brought to Sofia. Ten days later he again asked what had happened to Levi.

  Two days later, on 17 May, owing to a shortage of beds, Levi was sent to San Severo prison, where he was liberated by British forces on 17 October and given a post as an interpreter by a Civil Affairs officer, Captain Cooley.

  In October 1941 CHEESE was used as the principal channel for deception in conveying false information to the enemy on 20 and 27 October, in anticipation of the Allied offensive in the Western Desert planned for November. His role was considered both essential and completely successful, as was reported to MI5 in January 1942, although ISOS indicated some recent loss of confidence in him by the enemy.

  At the end of September 1942 ISOS indicated that Rossetti had been rebuked by his superiors in Berlin for his excessive travel, but his reputation had been restored after Walter Sensburg had seen Hans Piekenbrock in Berlin. According to his British file, Sensburg had been based in Brussels in 1940 and had distinguished himself by landing a group of wholly unprepared spies, Sjörd Pons, Carl Meier, Charles van de Kieboom and José Waldberg on the south coast of England where they had been quickly arrested.

  Between January and July 1942 the Abwehr, transmitting from a radio station in Bari, expressed renewed confidence in CHEESE and asked him to transmit daily, instead of merely twice a week, a request that was interpreted as proof of his perceived value.

  In April 1943 Dick White, while on a visit to Cairo, was briefed on CHEESE and brought back to London a summary of the case prepared on 1 September 1942 by Evan Simpson so that MI5’s B Division experts, T. A. Robertson among them, could offer so
me comments. Simpson’s Report on CHEESE covered twenty-four pages and started with a short account of Levi’s prewar history but concentrated on his activities during the two months he spent in Cairo between February and April 1941. Although, at the time that it was written, Levi himself had been sentenced to five years’ imprisonment and was still confined on Tremiti, Simpson had no knowledge of his fate beyond a single TRIANGLE intercept that suggested Levi had been detained for two to four months and then had been released. In his narrative Simpson concentrated on some of the technical issues that had arisen in developing the radio link, and catalogued the enemy’s perceived failures in their poor handling of what should have been considered an exemplary agent. Simpson opined that most likely Levi had been managed by the Italians, which explained some the shortcomings, but this was one of several issues that would be dealt with in greater detail by Tommy Robertson, a man with vastly more experience of the Abwehr. Simpson observed

  all outside evidence goes to show that, since the Italian declaration of war, and the prompt internment of Italians suspected of espionage, the enemy has been starved of reliable up-to-date intelligence from Egypt. He is consequently doubly greedy for it, inclined to swallow bad (in the absence of good) information and careless in his correlation and checking up. Though Levi was recruited by a predominantly German organisation it seems that the arrangements for his journey, wireless set, etc. were in Italian hands and he himself complained to Zähringer of the inefficiency of the arrangements. The handling of the CHEESE messages was done at Bari and linguistic evidence (the very Italianate idioms embedded in the French of the messages) seemed to show that the translation, and possibly the writing of the message was in Italian hands. Perhaps the whole case was being handled by Italians, and handled with a laziness and carelessness that the Germans would not have tolerated.

  Simpson was very dismissive of how the enemy had mismanaged CHEESE and speculated about his imprisonment and subsequent release, unaware that Levi had been incarcerated continuously since August 1941.

  It is possible that Levi’s release from prison with a new stock of plausible ingenuities on his tongue, may have had a hand in reestablishing what he had originally foisted on his employers, and found them even more ready to be deceived. All this must naturally remain pure speculation. If true it leads to the pleasing conclusion that ‘wishful thinking’ is not a monopoly of the democratic peoples. While on the subject of enemy methods, it may be remarked that had the enemy been handling genuine, loyal agents (instead of a pertinacious British organisation) he would almost certainly have soured their loyalty and lost their cooperation at a much earlier stage. The constant carelessness in encodement and (still more) in wireless procedure, endangered the whole communication and, strangely enough, it rose to a climax of inefficiency early in July just when the directorate of enemy intelligence seemed most eager for the information which was being transmitted. Meanwhile the failure to send money, the hollowness and monotony of the promises to do so, and the apparent indifference to the dangers which inefficiency would have brought on real spies concealed in Cairo, indicate a really lamentable lack of imagination and common sense.

  Simpson concluded his History on a final note of triumph; ‘CHEESE is still in action, and it is hoped that he will remain so, to the better confusion of His Majesty’s enemies.’

  The History was read in London by Tommy Robertson who drafted a six-page document of nineteen paragraphs dated 30 March 1943, quite critical of SIME’s handling of the entire case. He saw the matter as ‘an entirely Abwehr enterprise’ and one, more specifically, managed by Einz Heer II through the Athens Abstelle, as disclosed by TRIANGLE. Einz Heer II was the Abwehr’s military sabotage branch, and MI5 considered it significant that this was the section handling CHEESE, instead of Einz Heer I, the intelligence collection specialists. Robertson also took up the issue of the Abwehr’s ‘degree of incompetence in the CHEESE case’.

  In our experience the Abwehr is often recklessly careless both in the way in which it trains, or fails to train, its agents, and in the way in which it superintends, or fails to superintend, their later activities. These haphazard methods even extend into the sphere of communications … We also have suffered from an Abwehr officer who did not listen, could not operate, or called on the wrong frequency, and from cipher clerks who could neither encipher nor decipher correctly. The standard, of course, varies from station to station and, with the Abwehr wireless network as large as it now is, it would perhaps be unreasonable to expect the general average to be very high. Nevertheless is it surprising to find it as low as it is. We can confidently say that CHEESE’S experience with his control-station is a fair example of what any Abwehr agent may have to put up with.

  By way of further explanation, Robertson explained that Colonel Sensburg, the head of the Athens Abstelle, was a familiar character who had been known to MI5 since 1940 when he had gained an unenviable reputation at the Brussels Abstelle for sending poorly trained agents on badly planned missions. In particular, while at the Brussels Abstelle in September 1940 Sensburg had sent four ill-prepared and poorly equipped agents in a fishing-boat to land on a beach in Kent, and all had been arrested within hours or their arrival. Of the four, Carl Meier and José Waldberg had been hanged. Robertson had relied on TRIANGLE and information from an Abwehr defector to conclude that Sensburg was incompetent.

  Robertson then turned to what he termed ‘the greatest strength and the greatest weakness’ of the CHEESE case.

  That is the return of Levi himself to Italy and his continued presence there. It is unnecessary to emphasise how great a weakness this is or could have been. At any point after his return Levi, a man reputed to be not entirely trustworthy, could have revealed to his employers what the true position was. In theory at least he could still do so, though probably he now feels (particularly since his imprisonment) that he has committed himself so far as to make any withdrawal impossible. Nevertheless, any operational deception put over by CHEESE contains this element of risk. If Levi were, from whatever cause, to break down, any deception then being practised would not only fall but by its nature reveal to the enemy precisely what we were attempting to conceal from them. On the other side the advantages of Levi’s position (so long as he does not break down) are almost equally great. In the first place, the Abwehr is entitled to argue that no counter-espionage organisation and still more no double agent would have been willing to take the risks inherent in Levi’s return. Indeed, from Levi’s own point of view, the risks were fantastic. By returning to Italy at such an early stage he committed himself in advance to answering perhaps with his life for the good faith and reliability of an organisation of the nature and scope of whose activities he was necessarily ignorant. In these circumstances no one can blame the Abwehr if they regarded Levi’s willingness to return and remain in Italy as an absolute guarantee that CHEESE was not working under control. In my view it is this fact which explains the rather cautious tone displayed in secret services of November 1941. The Abwehr never said that V-mann ROBERTO himself was under control, but only that ‘the intrusion of the enemy Intelligence Service into the ROBERTO network is becoming clearer and clearer’. The picture in their mind was of SIME gradually penetrating the fringes of an organisation which was itself still sound at heart. Had they taken a different view, it would not have been possible for CHEESE at a later date to have restored his credit by discarding his network.

  Robertson, who had been the MI5 case officer running Arthur Owens, the Welsh double agent codenamed SNOW, was adept at the complexities of managing difficult individuals who demonstrated uncertain or shifting loyalties, made some relevant points about CHEESE;

  The return of Levi to Italy had, I think, another valuable result. With his going (since Fulvio Melcher had already fallen by the way) the Germans lost the only member of the Cairo organisation with whom they had any personal acquaintance. This, as we know from our own experience in England, is an incomparable advantage. A double agent whom the
Abwehr has recruited directly, whom they have trained, and with whose personality they are acquainted, is always a little circumscribed as to what he can and cannot do. He cannot step too far out of character without arousing suspicion; it is difficult for him suddenly to change the type or value of his reports; he must react to new situations in a manner in keeping with what the Germans know of his character. An agent whom the Abwehr have never seen, or better still a network of imaginary agents, offers a wide scope. The Abwehr know only as much or as little about them as it is expedient that they should be told. They have no means to assess the probability, physical or psychological, of their saying or doing one thing more than another. Moreover, when necessary, imaginary agents can be dispensed with, changed or even captured without imperiling the remainder of the network or giving rise to inconvenient administrative problems.

  Our experience in England has shown that the Abwehr, provided that its confidence in the original agent is unshaken, is very ready to accept new sub-agents recruited in the course of operations, and surprisingly unexacting in the particulars which it demands of their past careers and personalities. In this way we have been able to provide one agent with seven non-existent sub-agents, some of whom we have since put into direct communication with the Abwehr and who now constitute a network which has spread even outside the UK. Our experiences in this and similar cases also bears out the statement on page nine of the SIME report that ‘the enemy is curiously unwary and eager to accept stories of the disloyalty of disgruntled colonials, Irishmen, etc. and even of supposed ex-member of Fascist organisations in England’. It may be unnecessary to draw the attention of SIME to the equally wide field of opportunity offered by disgruntled allies, nationalist (and even Republican) Spaniards; Indian supporters of Congress; and German-born Americans. The Abwehr is also prepared to accept on the one hand agents or sub-agents who work without any other motive purely for money, and on the other such ideological prodigies as Poles inspired by a disinterested love for the Greater Germany. The explanation is probably to be found first in the degree (always noticeable) to which Abwehr officers are the victims of Dr Goebbels’s propaganda; secondly, in the slackness and increasing corruption of the organisation as a whole; thirdly, in the natural anxiety of an intelligence officer to see his estimate of the original agent justified by the skill and initiative with which he develops his work.