SIS’s enhanced performance in the Mediterranean and North Africa, which had become apparent by the end of 1942, was reflected elsewhere, as increasing numbers of networks were established, techniques were refined, liaison with Allies was improved and, despite the continuing (and in some cases intensifying) oppression of enemy occupation forces, growing numbers of individuals were bravely prepared to offer assistance in all sorts of ways to intelligence, escape and resistance organisations across Europe.
For most of the Second World War the main SIS representative in Switzerland was Frederick ‘Fanny’ Vanden Heuvel, based in Geneva. He and his staff established contacts with Swiss intelligence, mainly in order to get their co-operation in escape and evasion. The Swiss assisted by putting escapees quietly over the frontiers into France at appropriate spots and times. In the spring of 1943 Dansey produced a report on SIS activities in Switzerland which showed that the Service had agents operating in or into France, Germany, Italy, Yugoslavia, Belgium and the Netherlands, and contacts providing periodic reports from Turkey and Norway. The main effort was on Germany and France. In Germany there appear to have been agents in eight main industrial and communication centres, including four in Berlin. In Switzerland itself, the train-reporting service continued, while economic reporting, primarily for the local Ministry of Economic Warfare representative, had become a major commitment.
The files indicate good military intelligence on Italy too. A report of May 1943 from Geneva, marked up by Menzies as ‘v. important’, recorded that ‘Z.101’ had identified twenty-four special trains loaded with torpedoes and sea mines and that 120 German naval personnel had passed Brenner destined for Venice. Between 9 and 12 May, Z.101 noted approximately 25,000 German troops in trains to north Italy, en route to Verona, Piacenza and Bologna. Following the collapse of Italy, thousands of British prisoners-of-war were liberated, and many crossed into Switzerland where they were interned. Many of them, however, especially air crew, were badly needed in the United Kingdom. SIS, in conjunction with SOE, organised the clandestine escape through France of a number of these, it is said at the rate of some six every three weeks. On 26 October 1943 Menzies sent a personal message to Vanden Heuvel offering his ‘thanks and congratulations’ for the work of SIS staff in Switzerland. ‘After four years isolation and heavy work,’ he wrote, ‘I realise that all must feel the strain but the results should be a consolation and satisfaction to you.’
By this time SIS in Switzerland was receiving a detailed picture of internal German conditions and the effects of heavy Allied bombing. A source in Berlin reported that ‘90 per cent population still spend every night in cellars.’ The people were generally ‘apathetic’ and there was ‘no evidence [of] riots or demonstrations but could not visit working class quarters. Little hate against the British despite intensive propaganda, but people hope reprisals will be sufficiently effective to stop air raids.’ There was great faith in the possibilities of an unspecified secret weapon, which was ‘relied upon to work miracles but if it fails or proves non existent reaction will be serious and probably revolutionary’. The Air Ministry in London lauded these reports, affirming that they were ‘of considerable value and largely accurate’, since they added details to the more general evidence from signals intelligence ‘and provided an assessment that the RAF was unable to obtain from aerial reconnaissance when the weather was bad’. Another source visited Berlin from August to November 1943 and reported that there were no signs of internal demoralisation and that public services were functioning satisfactorily. Among workers and soldiers, he noted, Hitler was still considered a ‘demigod’. Heavy bombing, while materially devastating, seemed to have had the effect of stiffening resistance and determination for revenge.
Reflecting the continuing value of the close Anglo-Polish intelligence relationship, a precious supply of information from high German military circles came to SIS from an impressive Polish intelligence network code-named ‘Darek’, run by Major Szczęsny Choynacki, with cover as the Polish deputy consul in Berne. After the Germans (who knew the network as ‘Jerzy’) began in the early summer of 1943 to read Choynacki’s radio traffic and break up his organisation, they concluded that at least one of his agents had well-informed access to Hitler’s headquarters. This was probably ‘JX/Knopf ’, one of the ninety-three Darek sub-sources recorded by SIS. Evidence survives that Knopf reported between February 1942 and April 1943. He mainly supplied intelligence on the Russian campaign, but also considered a few other matters, such as the possibility of an Italo-German offensive aimed at the seizure of Suez, and troop dispositions in Tripolitania. A surviving Knopf report, dated 12 February 1943, and described as ‘from “contacts in the O.K.W.” [German high command]’, discussed the ‘state of mind’ at the headquarters following the defeat of the German Sixth Army at Stalingrad. ‘Despite their prodigious efforts, their heavy casualties, illness and, in some sectors, the loss of the greater part of their equipment’, German army morale was ‘far from broken; indeed it is very good. For this reason all rumours of an imminent break-down of the German military front in the East are quite untrue.’ MI14 (the German order-of-battle branch of British Military Intelligence) regarded Knopf as having ‘very good contacts’, and when ‘reporting from his usual sources, he is more often sound than not’. The information, however, could not ‘be accepted without some confirmation, but when it is clear and factual and is in line with our own views [a necessarily double-edged judgment this] or with information from other sources, a high degree of confidence can be placed in it’.1
One of the most interesting relationships was that between Vanden Heuvel and Allen Dulles, the OSS representative in Switzerland, who arrived in early November 1942 (and who later served as Director of the CIA from 1953 to 1961). Although they would ultimately share intelligence, from the beginning controversy surrounded Dulles. In January 1943 Broadway warned Vanden Heuvel about him, suggesting that he was likely to ‘lend himself easily to any striking proposal which looks like notoriety’. Although the SIS head of station and Dulles had mutual friends in the United States and were personally on friendly terms, Vanden Heuvel himself concluded that Dulles was ‘out for himself and clashes of interest are bound to come’. Competition for agents seemed especially acute. In typically acerbic fashion Dansey observed that the dollar was an ‘unfailing magnet’ and that Americans ‘everywhere swallow easily and are not critical’. He therefore thought it essential for SIS to work as closely as possible with Dulles (and the French, Poles and Swiss), otherwise the ‘only people who will profit by this madness will be the Germans and agents who get paid by three and even more masters’.
The closeness of the wartime Anglo-American intelligence relationship is confirmed by the friendly tone of this letter from ‘Stewart’ (Menzies) to ‘My dear Bill’ (Donovan).
The potential for disaster was demonstrated by Dulles’s first contact with the Abwehr representative, Hans Bernd Gisevius, in January 1943, which he reported back to Washington on a cypher which the British ascertained (through their Polish agent Halina Szymańska) had been broken by the Germans. In the middle of April Dulles told Vanden Heuvel that he had seen Gisevius, who had just returned from Berlin and had told him that forty large flying boats had recently been built in Rotterdam to be used for the heavy bombing of London manned by suicide squads. Despite having been alerted to the problem with the cypher, Dulles had reported this to Washington in two telegrams. Dansey thought that Dulles had been ‘stuffed’ by a deliberate piece of German disinformation. Clearly agitated, he told Vanden Heuvel: ‘could you report to the fool [Dulles] who knows his code was compromised if he has used that code to report meetings with anyone, Germans probably identified persons concerned and use them for stuffing. He swallows easily.’
Vanden Heuvel, nevertheless, was keen to exploit Dulles’s contact with Gisevius, who in July 1943 (responding to questions supplied by SIS) told Dulles about rocket projectile tests at ‘Peenemuende or Swinemuende’, for which his source was an Abwehr colleague. Gisevius expressed great concern lest Dulles’s communications were unsafe, since discovery would certainly lead to identification of the original source, and he suggested to Dulles that he should pass the information ‘through British channels for extra safety’ (which he did). The information stated that damage caused by air raids at Friedrichshafen included works producing the steering mechanism for rockets, which could lead to up to three months’ delay in them being put into operation. The main production of rocket bombs was taking place at Frankenthal. The source believed that no other important factories were there other than some sugar refineries and for this reason the source urged that the information must be used with the greatest discretion. The course also stated that Hitler was taking a close personal interest in the production of the projectile.
In September 1943 Szymańska reported that Gisevius, who was ‘highly nervous’, had (for the first time) told her about the ‘V’ (for ‘Vengeance’) weapons being developed. He said he was giving all the details he could ‘as [a] Good German’, because he firmly believed the new weapon was serious and would certainly lengthen the war even if it could not affect the eventual result. Although London remained desperate to glean more information on V-weapons, particularly over rocket sites, scales of production and the date for which firing was planned, Gisevius proved unable to help further. According to Szymańska, by February 1944 Gisevius had become persona non grata among his German colleagues and friends. Dansey told Vanden Heuvel that he feared Gisevius’s usefulness was ‘now quite impaired. 189 [Dulles] has compromised him beyond redemption.’ Dansey was further alarmed since Dulles was ‘flooding’ Washington with Gisevius information, representing it as being from ‘an important source’. When Szymańska saw Gisevius again in February 1944 his position (and that of Canaris) had worsened considerably, particularly after the defection of the Vermehrens in Turkey the previous month. A year later Gisevius told Szymańska that he had been deeply involved in the July 1944 plot to assassinate Hitler and that after its failure he had managed to evade arrest by lying low in Berlin for three months.2
Cyril Cheshire, head of the Stockholm station from December 1942 until the middle of 1945, expanded and developed the work started by his predecessor, John Martin. Reviewing the station’s output in October 1944, Bill Cordeaux at Head Office noted that Cheshire had worked up the number of reports from fewer than four hundred a month ‘to an average of about 700 per month, the majority of the increase being Armed Forces information’. Much of this resulted from the exploitation of Dutch, Danish, Norwegian and Baltic contacts, among businessmen and others who were able to travel between Sweden and German-occupied Europe. One especially valuable Danish agent, a businessman recruited in December 1942 and code-named ‘Elgar’, was a frequent visitor to Stockholm where his company had an office. Other business connections took him to Germany, Finland and Romania, and he acquired information from fellow countrymen working for German heavy industrial companies manufacturing war matériel. During 1943 he recruited some twenty sub-sources, including a diplomat in Hamburg; technical workers in Peenemünde on Usedom island in the Baltic, where the Germans had a weapons-testing site; an employee in the Berlin office of Donau Flugzeugbau, which manufactured Messerschmitt aircraft in Hungary; and a man in the Danish lighthouse service, able to debrief lighthouse-keepers for shipping information.
Conveying material to Cheshire in Sweden required some ingenuity. In the autumn of 1943, for example, Elgar brought three hundred reports on film concealed in glass bottles hidden in one of three barrels of acid imported for business purposes. In November Cordeaux noted that he was ‘among the first’ to report on the V-1 (described at the time as a ‘rocket gun’) and that he had supplied a photograph of a rocket which had landed on the Danish island of Bornholm in the south-western Baltic Sea. He had ‘provided valuable night fighter information, Finnish chemical warfare equipment and valuable reports from his sub-agents and his own journeys in Germany, Finland and Roumania’. One of his sources had provided ‘ground photographs of air raid damage in Hamburg’. But it was too good to last. Elgar was arrested by the Germans in January 1944 and his network disintegrated. Although roughly treated in captivity, he was transferred to a sealed camp in Sweden in April 1945, and was suspected by some of having been an enemy agent. The evidence from surviving accounts of German interrogations of him does not support this in any convincing way, and SIS afterwards decided that he had provided information to his captors simply ‘in order to save his skin’. While he gave the Germans accurate descriptions of SIS staff in Stockholm and of Danish contacts in Sweden and the United Kingdom, he also gave invented information with sufficient plausibility at least to worry the Germans, if not to convince them, claiming, for example, that SIS had ‘succeeded in establishing important groups in Berlin, Hamburg, Bonn, Königsberg and Vienna’ and that English-trained sabotage teams had been deployed in the Danish islands of Zealand, Funen and Falster.
Like all the other neutral capitals in Continental Europe, Stockholm was full of dubious characters offering information to whomever would pay. One such was agent ‘36439’, a Russian émigré who had been a Z Organisation source before the war. In late 1943 this agent claimed to have a penetration source in the Japanese legation, agent ‘Eve’, recruited in June 1943, who provided copies of Japanese despatches and re-enciphering codes; and an economist in Berlin, who, apart from reporting in his own right, retailed gossipy information from an old contact who was a housekeeper in the household of Reichsmarschall Hermann Göring, the head of the German Air Force.
From April 1943 a trained Section V representative, Peter Falk, was posted to Stockholm under Passport Control cover to run a separate Sweden station devoted to attacking the German intelligence service, not only in Sweden but across Scandinavia and elsewhere. Falk’s chief target was Dr Karl-Heinz Krämer, the German assistant air attaché in Stockholm since November 1942. Krämer was a flamboyant character who, as one of SIS’s contacts in the Swedish police complained, ‘always travels in a fast sports car and was therefore very difficult to follow’. He was, in fact, an officer of the Abwehr’s Air Intelligence Section (Luft I) and had been posted to Sweden to develop intelligence operations against Britain and the USA. SIS’s interest comprised the need not just to identify and (if possible) neutralise him as an Abwehr officer, but also to assist MI5 in closing down any potential intelligence operation in Britain itself. Late in 1943 signals originating with Krämer and containing intelligence from and about the United Kingdom were acquired by OSS in Switzerland. Further messages purporting to come from Abwehr agents in Britain (including one code-named ‘Josephine’), which contained a mixture of plausible speculation and some apparently factual statements, subsequently appeared among British signals intelligence. While MI5 eventually concluded that the alleged Krämer network was fictitious, suspicions remained that he was getting information from Swedish diplomatic sources in Britain. SIS in Stockholm meanwhile mounted a successful operation against him.3
In early December 1943 luck took a turn with a walk-in. A ‘strongly anti-Nazi’ Austrian woman, separated from her Swedish husband, came to the British legation and asked to speak to ‘someone who spoke German’. Fielded by an SIS officer, she ‘did not disclose the nature of her business until she had been safely conducted . . . out of the reach of S.O.E., YN [the naval attaché], etc.’. The woman then declared that her best friend, also Austrian, worked as a maid for Krämer and had chanced upon some important-looking documents which she had copied. These appeared to be drafts of telegrams about air intelligence from Britain and included a ‘request to Josefine [sic] for information’. This first contact duly led to a meeting with the friend and to the recruitment of both women. Throughout 1944 the maid supplied more message texts and other papers from Krämer’s desk, waste-paper basket and coat pockets. Moreover, after coolly borrowing and copying the key to his desk drawer which he always kept locked (by taking its impression in a dish of butter), she was able to abstract and copy by hand no fewer than eight current and old passports, which showed his travel movements since 1938. These included a short visit to England in the month before the outbreak of war, which naturally sounded alarm bells in London.
After the end of the war, when SIS were able to interrogate Krämer himself, and assemble reports from both German and Swedish sources, it appeared that the energetic and creative Krämer had been misleading both the British and his own Abwehr superiors. His reports from ‘very reliable sources’ were mostly reworked second-hand material acquired from intelligence-peddlers in Sweden and elsewhere, though amid this was information derived from genuine Swedish service attaché reports originating in London. Although MI5 had effectively rumbled Krämer before the end of 1943, there were residual worries about him up to the early summer of 1944, especially in the tense few weeks before Operation ‘Overlord’ (the invasion of France by Allied forces), when it was feared that even the slightest security lapse could jeopardise the invasion. Learning in May 1944 that Krämer had been ‘laying aside considerable savings every month’, and was suspected of pocketing ‘quite large sums from the expenses allowed for paying agents’, Peter Falk thought that the German might be susceptible to being ‘bought’ by SIS, especially since he had briefly been involved in running a concentration camp and believed that he was on the Allies’ list of war criminals. But Broadway vetoed the proposal. ‘We cannot do business with war criminals to save their necks,’ wrote a Section V officer. ‘There is surely nothing very important that this peculiarly unpleasant rat could give us if he was allowed to leave the sinking ship,’ added Cuthbert Bowlby on 11 June, ‘and the “leakage” is now so much less dangerous than before “Overlord” started.’
During the second half of the war, the Helsinki station-in-exile under Harry Carr continued to operate from Stockholm. In December 1942 Menzies congratulated Carr for ‘considerable progress in work of your station during last six months’. This was ‘particularly commendable bearing in mind handicap of working from outside your own country’. Carr’s product had been boosted by the re-emergence in May 1942 of his pre-war agent Outcast, who was now working for the Abwehr. Outcast was briefed to report specifically on the results of RAF raids on Berlin and other German cities. His early reports, sent by secret writing to a cover address in Sweden, were not very useful but his stock in London rose dramatically after some detailed and accurate reporting on the results of the Battle of Berlin, the RAF’s heavy bombing campaign launched against the German capital from November 1943, and on which he was debriefed by Carr himself in Stockholm. Carr provided a large-scale plan of Berlin while Outcast ‘produced little scraps of paper concealed all over his person with what seemed to be unintelligible scribbles on them in pencil’. The resulting report reached the Air Ministry before they got any clear photographic intelligence and ‘some weeks later’, recalled Carr, ‘we were delighted to see our typewritten report, almost word for word, in the Air Ministry’s periodic Intelligence Summary, which had been sent to the Air Attache in Stockholm’.
Following his May 1942 visit to Stockholm, Outcast whetted the appetite of his Abwehr employers with the prospect of securing information on the United Kingdom, and possibly even establishing a source there through his good Swedish connections. This scheme eventually developed into a fully fledged double-agent operation. With the help of T. A. ‘Tar’ Robertson of MI5, Outcast passed a mixture of genuine and fictitious information (‘foodstuff’) to the Abwehr, which enhanced his value to them and justified his regular visits to Stockholm. This, in turn, meant that he could be debriefed in person, and it thus improved his reporting. By the end of 1942 he had begun to include information on the power struggle between the Abwehr and the Sicherheitsdienst, a subject of great interest to him as well as to London, since he belonged to a circle of Abwehr officers regarded as close to Admiral Canaris. In early 1943 he was evaluated in Head Office as ‘the best source for information on the German interior produced so far in the war’.
Outcast was known to Swedish Military Intelligence as an Abwehr agent and he supplied them with information on the German-Russian aspects of the war. In turn, he reported to Carr the Swedes’ views on the progress of the war, which were of considerable interest to London, especially when the tide began to turn against the Germans. Late in 1942 Outcast established contact with the Japanese military attaché in Stockholm, General Makoto Onodera, who recruited him as an agent. By 1944 Outcast was providing Onodera with American deception material supplied through SIS. Outcast’s reports continued to be very well received in London throughout 1943. One in particular, on conditions in Poland and Polish attitudes towards the Russians and the Germans, aroused great interest in the Foreign Office and was shown by Menzies to Churchill. When in late 1943 Outcast, ill with tuberculosis, wanted to move with his family from Berlin to Stockholm, SIS was supportive, and he managed it early in 1944. But the Swedes refused to let him stay and he was evacuated to Britain, where he died the following year. In a postwar SIS evaluation he was described as ‘one of the most successful spies against Germany that the 1939-45 war produced’.
Other agents reported from inside Germany, as is illustrated by a telegram from Carr dated 1 October 1943 with a report from a source travelling ‘from Belgium to Stockholm who was in train standing in Hanover Hauptbahnhof during raid September 22nd/September 23rd’. The station remained intact ‘but railway lines about 150 metres outside station (in direction of Berlin) were smashed and five full trains standing on lines were destroyed or badly damaged by bombs and fire’. The district north of the station ‘suffered greatly from bombs and flooding resulting from burst water main pipes. Inhabitants seen escaping from this district wading knee deep in water.’ In early 1944 a report from a Finnish source raised a difficult dilemma for Harry Carr where professional duty cut across personal feelings. The agent told him that the Germans had taken a party of Finnish officers on a tour which had included launching sites for V-1 flying bombs in north-west France and had ‘boasted to them how London would be obliterated by the pilotless machines’. At the time Carr’s mother and sister were living together in a top-floor flat in Norwood, south London, which he realised was right in the line of flight of V-1 bombs heading for central London. Apart from looking after his sister (who was working as an ambulance driver), Carr reckoned his mother ‘had no special reason for remaining in the London area’, but he could not ‘disclose the secret agent’s report to them’. In the end he wrote her a letter ‘saying that I considered she had stuck to London long enough, and suggested that she should go away for a bit’, and stay with relatives in Bournemouth on the south coast. His mother ignored the advice and remained where she was. As Carr had feared, in June 1944 one of the very first ‘doodlebugs’ aimed at London landed near her flat, blew in the dining-room window and caused quite a lot of damage. Fortunately, however, his mother was in another room at the time and his sister was out.
While in April 1942 Bill Cordeaux had complained that there were only four SIS ‘stations’ (here meaning networks) in Norway, in June the following year Broadway sent Stockholm a list of twenty-three Norwegian stations whose members and agents might appear in Sweden seeking assistance. They extended from Kirkenes in the far north, through Tromsö (two stations), Trondheim (three), Bergen (two), to Stavanger, Kristiansund in the north-west and Oslo (four). There was no shortage of patriotic volunteers for these stations and the Norwegian section at Head Office (P.9) established a training school in London at 14 Brompton Square. While the ‘primary objective’ of the Norwegian shore stations was ‘the obtaining of sighting reports of major units of the main German battle fleet’ (especially the battleship Tirpitz), their activities varied significantly according to the personalities of their members and the range of their contacts. Intelligence on German naval movements was collected from direct visual observation as well as from agents recruited locally. This was all backed up with extensive logistical support. Motorised fishing vessels, motor torpedo boats, submarines and aircraft were all used to infiltrate agents. Once ashore each station needed support, for example, in charging accumulators to power their radio sets, as well as couriers to deliver intelligence and warning of German counter-measures, such as arrests, searches and roadblocks. Visual reports on the Tirpitz’s movements, along with Ultra decrypts and photographic air intelligence, locating the ship and its defences in Altafjord, northern Norway, contributed to the planning of a daring midget-submarine attack which in September 1943 disabled the warship for three months.4 For a year from late 1943, the ‘Aquarius’ station in Stavanger had a useful agent in the police headquarters whose command of German led to his being employed in liaison duties with the German military. Thus he was able not only to supply detailed information about the local German forces but also to influence the course of security investigations.
One of SIS’s most successful Norwegian agents was Oluf Reed Olsen, whose report of his second operation, a six-month spell from May 1944 based near the seaport of Kristiansand on the south coast of Norway, gives a flavour of the challenges presented by working in enemy-occupied territory. Olsen and his wireless operator were dropped in by parachute, and the operation (code-named ‘Makir’) began very well. ‘I had an excellent landing’, he wrote, ‘in the middle of a blueberry bush approx. 10 metres from one of the corners of the light triangle’ (set up to mark the landing zone). ‘The reception were on the spot and ready with a cup of warm tea.’ A few days later, Olsen, who was carrying his wireless set, found himself alone in a railway compartment with a German SS officer, who ‘became extremely interested in the bag containing the radio set and kept on staring at it’. Olsen managed to move to the next carriage, but mildly observed in his report that ‘it would be a great help if the radio sets could be packed into an ordinary leather bag with a good lock on it’. A camp was set up on a forested hillside near the fjord entrance to Kristiansand from where shipping movements could be monitored. A contact in the harbour police recruited two German sailors who had been planning to desert and who had sought his assistance in escaping to England. They ‘were keen to join and fight for the side which they thought right’, but the Norwegians persuaded them that they could best serve by staying on and collecting military information. ‘This they agreed to do, and all seemed so easy that we were almost wary of a Gestapo trap.’ Olsen told them ‘that if anything went wrong, they would be given a safe passage to Sweden’ but, ‘on the other hand, if they double-crossed us, their death warrant could be considered signed’.
Apart from reporting shipping movements and providing daily weather reports, the Makir station had also to report on U-boat activity along the adjacent coast, as well as German order of battle in the area. Olsen, who was careful about security, described the routine as follows:
The main contacts in the town collected reports from their respective sources and placed them in the post-box: twice a week the post-box in town was emptied by the man who was responsible for sending food to the camp, and the reports which had thus been collected were sent with the food courier twice a week. There were always two on this job as it was necessary for one man to go ahead empty-handed in case of bumping into a control. In this way four men were constantly employed on this part of the work. Every fourteenth day the timing was put forward one day to avoid regular traffic. The couriers met each other at pre-arranged places in the forest and at each meeting the place was changed again for the next time.
The men’s tents were ‘well camouflaged from the air and ground, which proved itself when the owner of the property, together with 2 berry-pickers, passed not more than four yards from the main tent without discovering the camp’. Silence was the rule; boots were forbidden; ‘and gym-shoes were worn whether it was wet or dry’. They also ‘had considerable trouble’ from one colleague ‘being prone to snoring, but the person concerned soon became used to being woken up 10 times a night’. At its peak, too, the station transmitted up to ten messages a day, and had to relocate more than once in order to evade German direction-finding units.5
The Low Countries
The Netherlands and Belgium together posed particular problems for British intelligence operations. Both were densely populated small countries, which made it extremely hard to find suitably remote locations for air dropping zones. Furthermore, as they both lay under the Allied bomber route to the German industrial heartland of the Ruhr, anti-aircraft defences were especially concentrated and made it exceptionally difficult for aircraft to grope around at low level trying to distinguish the faint lights of a reception committee. While both were geographically well placed for sea operations from Britain, the enemy in consequence specially built up maritime defences and security measures so that this method soon became prohibitive. The military importance of the two countries to the Axis also meant that a high priority was given to them by the German security services, and although the Sicherheitsdienst and the Abwehr were often at each other’s throats, their efforts constituted a very serious threat to SIS, which was compounded by the disasters suffered by SOE in both Belgium and the Netherlands.
By the autumn of 1942 German successes in Belgium had resulted in a situation when, as an official SOE history put it, the agency’s ‘organisation in the field was no more than a mirage created by the Gestapo’.6 In March 1942, following a wave of arrests, Claude Dansey began to worry about the SIS organisation in Belgium. He told Frederick Jempson (head of the Belgian Section in Broadway) that ‘unless some drastic steps are taken with or without the consent of [Fernand] Lepage [the London-based head of the Belgian Sûreté de l’État] you will soon be without any agents left in Belgium’. Dansey (wisely in the light of SOE’s experience) wanted to double-check whether any of SIS’s wireless operators had been compromised. Only then might they ‘devise some steps to save what is left’. But the news was not all bad. In April 1942 R. V. Jones in the Scientific Section noted that Jempson had produced two valuable reports, one of 17 February giving the position of the first known German Radio Direction Finding (RDF) station for controlling night fighters; and one of 28 March locating the first identified ‘decimetre’ directional wireless communications station in Belgium. Jones added that, ‘as all our attacks on the Ruhr’ had to pass through Jempson’s territory, ‘these constitute an important section of the main German air defence system and the more we know about them, the less our casualties’. Early the following year another agent, assisted by a compatriot employed in forced labour by the Germans building a new radar station at Lantin, near Liège, bribed his way into the building where over several nights he made detailed sketches and notes of all he saw. Supporting SIS’s case for a gallantry award after the war, the Air Ministry assessed his achievement as ‘making among the most outstanding air intelligence received during the war from clandestine sources’. This had enabled counter-measures to be developed which the ministry said saved ‘the lives of hundreds of Allied airmen and many of our night bombers’.7
The most successful Belgian network was Service Clarence, which was effectively administered by Hector Demarque after the First World War espionage veteran Walthère Dewé (‘Cleveland’) had been forced to keep a low profile. Dewé was later shot by the Germans while fleeing from arrest in January 1944. Clarence provided regular technical intelligence, including reports about the removal of uranium salts to Germany, and one agent working in a German headquarters near Breskens just across the Dutch frontier reported on German minefields and the disposition of enemy forces along the Belgian coast. A contact in a Belgian oil company provided not only information on military and other road traffic but also petrol and cars for the network to use. Information on enemy communications and movements, especially on the Belgian railway system, continued to be a priority target and, as in the First World War, Belgian agents produced most impressive results. Service Clarence, for example, continued to provide reports on the RAF bombing of German cities, including Hamburg, Düsseldorf and Cologne; intelligence on batteries at Zeebrugge and Ostend and on the defences at Knocke; and photographs of German anti-aircraft gunners, aerodromes and gun emplacements. The network also produced reports on V-weapons, including Peenemünde (4 January 1944), while from an entrepreneur working for the Germans and travelling in northern France an agent gleaned sightings of seventy-metre-long platforms (9 January 1944). Thereafter there were regular reports from Service Clarence on V-rockets.8 The ‘Luc’ network acquired information from the head office of Belgian State Railways and was able to give advance notice about the transport of German troops and equipment through Belgium and northern France. Luc agents also tapped into the State Railways’ private telephone system, as well as the teleprinter service between the State Railways and ‘the German Railway Transport Office which controlled the running of all military trains’.
The Belgian networks faced considerable problems in getting information out of the country. Towards the end of 1942 SIS began to send new ground-to-air ‘Ascension’ radio sets to the Clarence network. Developed by Gambier-Parry’s Section VIII boffins, this new device allowed agents to speak directly to an aircraft flying near by. Since it dispensed with the need for Morse code, it was easier for inexperienced operators to use, and because they did not require lengthy call-sign procedures, transmissions were more difficult to locate by direction-finding.9 The Ascension sets were used with some success in Belgium and elsewhere, but the system was not very useful for long messages which still had to be smuggled out by courier across long and precarious land routes. One officer in Broadway complained in January 1943 that, although they were receiving a large number of reports from Belgian sources, the information contained in them ‘has been so old that they have become valueless’. During the last week of December 1942, the Military Section had sent eighty reports to the War Office, the contents of which revealed an average time lag of three and three-quarter months. An officer asked if anything could be done to speed up delivery, but Hubert Hatton-Hall, head of the section, commented: ‘I imagine that we are very lucky to get this stuff at all & that there is little chance now of speeding up.’ In May 1943 Jempson told Dansey that all his courier lines with Belgium had broken down following arrests at collecting and forwarding centres in Paris, Lyons and Toulouse. Jempson therefore asked Demarque if he could open a northern route through Sweden. Eventually he found someone and sent an Ascension message to London that one of his agents would call on the British military attaché in Stockholm: ‘He will give the name Buelemans and will say: “Aucun chemin de fleurs ne conduit a la gloire.”’10 The contact appears to have been made but in the end little intelligence came through Sweden.
SIS’s problematic relations with the Dutch began to improve in 1942 after Bill Cordeaux took charge of operations to the Low Countries and Captain D. O. ‘Charles’ Seymour took over as head of the Dutch section. One of the section’s problems concerned the employment of Dutch nationals, whom the government-in-exile wanted to keep for themselves in specifically Dutch formations. To get round this difficulty, Seymour proposed in September 1942 to look for men with British passports, but who were, ‘to all intents and purposes, purely Dutch’, and he asked the service liaison sections in Broadway to ‘do all in your power to trace every likely individual’. The first trained agent of SIS’s new Dutch set-up left for the Netherlands in March 1943. Because it was (rightly) thought that the existing British clandestine networks in the country were compromised, ‘Hendrick’ parachuted in blind, ‘with no organisation to receive him and even without strong hope that his identity cards and other documents were correctly forged’. Cordeaux described him as a strong Calvinist, who was ‘utterly fearless and regarded his task as a mission from God’. He encountered great difficulties from the start, when his Ascension radio-telephony set failed to work. He managed, nevertheless, to build up a network involving some three hundred members, which by the end of 1943 ‘was providing a steady and useful, if not very prolific series of reports’. Hendrick’s attempts to return to England vividly illustrate that it was as difficult to get agents out of occupied Europe as it was to get them in. One scheme involved seizing an air-sea rescue motor boat in Scheveningen harbour, but the engine refused to start and Hendrick escaped only after a lengthy shooting-match. On a second occasion he got as far as the Pyrenees, when a snowstorm forced his party back from the Spanish frontier. Hendrick was arrested shortly afterwards and spent the winter of 1944-5 in prison subjected to intermittent bouts of violent interrogation before he was liberated by advancing Allied forces in the spring of 1945.
In August 1943 Menzies learned that the French Committee of National Liberation in Algiers, which had formalised the joint leadership of Giraud and de Gaulle over all anti-Axis French forces, was deliberating on the future of French clandestine services. He alerted the British Minister of State in Algiers, Harold Macmillan, to the importance of SIS’s relationship with the Fighting French. ‘Since 1940,’ signalled Menzies, ‘have maintained close collaboration with de Gaulle’s special Bureau here known as B.C.R.A. [Bureau Central de Renseignements et d’Action].’ ‘Organization’, he continued, ‘built up in France with much difficulty in collaboration my service and B.C.R.A. has now reached stage where most important results being obtained.’ He quoted a letter he had just received from the Director of Military Intelligence ‘saying that the intelligence supplied by this joint French-British service has been of greatest value to General Staff particularly during last month’, adding that the War Office ‘would view with great misgiving any action which might interfere with our relations with B.C.R.A. and perhaps decrease the flow of military information’.
Menzies, evidently worried that SIS’s liaison with the Gaullists in London (and its monopoly of the information produced by the link) might be jeopardised by some new arrangement based in North Africa, naturally emphasised the productivity of joint espionage operations in France, while not unduly troubling Macmillan with any of the accompanying problems. There was a universal expansion of clandestine activity in France during 1943, but that of SOE and resistance organisations, who concentrated on special operations more than intelligence, often cut across SIS’s information-gathering networks. There were also continuing and overlapping tensions between different groups, including Gaullists, anti-Gaullists, conservatives, Communists, former Vichyists and those with ostensibly no politics at all. Internal French political competition, moreover, was sharpened as victory over the Axis became more likely and also as the Gaullists began to get the upper hand in the battle for acceptance as the legitimate French government-in-waiting.
One of the most important Free French networks in the early part of the war, the Confrérie de Notre-Dame, remained important in north-west and northern France, but by the summer of 1943 London reckoned that the network was in some trouble. Fearing that he knew too much and was a security risk in France, Dewavrin had brought the inspirational leader Gilbert Renault out, but in his absence rivalries among the hierarchy undermined the organisation. In March 1943, moreover, ‘Espadon’, a former port employee in Bordeaux, who had been providing shipping intelligence and helping to courier information about coast defences in Brittany, was arrested after helping two British fighter pilots to escape.11
There was better news of the ‘Davis’ group, based around Nice, which had originally been associated with the Confrérie, but had become an independent outfit by mid-1943 when a journalist with the cover name ‘Chavagnac’ took charge. A handful of his original telegrams still remain on file, giving a flavour of the material coming from the network, and also illustrating the difficulty of disentangling intelligence from special operations. In late July 1943 Chavagnac offered to blow up railway locomotives which the Italians were trying to remove from Nice. On 28 July he noted the arrival in Nice of Field Marshal von Rundstedt, commander of the German Army Group West, and two days later gave precise details of his journey back to Paris by special train, information which could have been used to inform an attack on the general. But the timings were very tight. Chavagnac’s 30 July telegram noting that Rundstedt’s train ‘will leave Sete on the first of August at 1525 hrs. and will arrive in Paris (Gare de Lyon) on the second of August at 0925 hrs.’ took over a day to be processed and was passed to the Air Ministry only at 7.45 p.m. on 1 August. On 2 August Chavagnac reported that three German divisions were moving out of the Montpellier-Port Vendres region, and were thought to be going towards Italy. On 14 August he signalled that twenty-seven German trains with about eleven thousand troops and nine Italian trains with five thousand troops had passed through Cannes towards Italy between 6 and 9 August.
The Davis network was highly regarded in Broadway. A March 1944 minute noted that over the previous six months it had been the most productive ‘of any of our organisations’. In June it was added that Davis had ‘frequently received bouquets from the Air Ministry and War Office, who also often express astonishment at the volume of Davis’ output’. The Admiralty considered Davis to be in the ‘top drawer’ and the Ministry of Economic Warfare also commented favourably on the value and number of economic reports produced by the organisation. London believed that this success was due to Chavagnac’s ‘administrative ability and common sense’, combined with ‘a complete disregard for his own safety’. Early in 1944, however, ‘two minor members of the organisation divulged their knowledge of it to the Germans’. After failed attempts ‘firstly to evacuate them by a pickup operation and secondly to execute them’, and fearing that the network had been seriously compromised, Chavagnac was brought out and the organisation split into thirteen independent groups, which all continued to supply ‘valuable intelligence’.
Marie-Madeleine Fourcade was described by SIS’s Kenneth Cohen as the ‘copybook “beautiful spy”’. Top left shows how British forgers could change her appearance for false identity papers.
An annotated breakdown of Marie-Madeleine Fourcade’s Alliance organisation drawn up in August 1942 demonstrates the reach of one of the largest and most important SIS French networks. One hundred and forty-five individuals are listed, organised into ten different groups. The largest of these, led by ‘Panthère’, had thirty agents, among whom were an electrical engineer and wireless operator who ‘has some serious sources in the Beauvais region’; an ‘industrialist, Paris and Nord’; a Paris-based female friend of Panthère, ‘a very cultured woman and enterprising’ with ‘contacts in medical circles’; a businessman able to travel between the two zones of France and to Germany; a police commissaire ‘in charge of the surveillance of the coast in Brittany’; an engineer who divided his time between Lille and Grenoble; and a port employee in Brest who ‘possesses information on movements of all German boats which he can pass on to us’. At its peak in the spring of 1943 Alliance was reckoned to have over two thousand agents, couriers and contacts right across France. Organised into three main geographical regions, each of which was divided into a number of sectors, it was run along military lines and concentrated broadly on military intelligence, including information on U-boat bases in western France and, from 1943, V-weapons.
A snapshot of life for an Alliance agent on the ground is provided by ‘Pierre Verrier’ (‘Seagull’), a civil airline pilot before the war, who, among other things, helped make arrangements for landing operations. Debriefed on his return to England in April 1943, he reported that he had had ‘very little trouble’ with the French police, and that ‘the most patriotic and the easiest ones to get on with were those belonging to the pre-war regime’. He claimed that the Gestapo, who ‘combined excessive politeness and the utmost brutality in the accomplishment of their duties’, were ‘overwhelmed with work’. Sometimes they were ‘remarkably stupid’, as when they had come to arrest two brothers. At the house they ‘asked one of the men if his brother was at home’. He replied that he would go and see, and both men managed to escape while the Germans waited patiently for him to return. ‘Whilst they were in the house,’ moreover, ‘a young local boy came in, went upstairs to the attic, dismantled and packed the W/T set and walked out without being molested.’ More ominously, however, Verrier added that ‘in every small town and village’ the Gestapo had ‘at least 2 or 3 denouncers’. Verrier, too, had good advice for other agents. ‘When in lonely country districts where he was a stranger, looking for suitable landing-grounds,’ he said that ‘he made a habit of visiting 2 or 3 farms to enquire for food.’ This could provide him with an alibi. Should the police ask him why he was in the district, ‘he could answer that he was looking for food’, and ‘indicate farms where he had called and where, should the police check up on his story, they would find it to be true’. Another observation was that when a stranger went to live in any locality for a length of time, it was ‘not safe for him to be there without receiving correspondence’. If, ‘in the normal way’, he did not receive letters, ‘he should write letters to himself. A person who does not receive any correspondence’, asserted Verrier, ‘is apt to be looked upon as a person in hiding.’
Despite Verrier’s somewhat dismissive remarks about the Gestapo, the Alliance organisation was very badly hit by enemy penetration and arrests during 1943. Its sheer size and centralisation, along with some lax security, proved fatal for many of its members. The German occupation of south-east France in late 1942, moreover, deprived Allied sympathisers of what had been a comparatively safe haven. Early in 1943 it was clear that the Germans had become aware of a clandestine organisation whose members took the names of animals as pseudonyms. In January and February a series of arrests disrupted the network in Toulouse, Marseilles, Nice and Lyons, where the Gestapo chief, Klaus Barbie, directed the torture of two young female Alliance members, ‘Hummingbird’ and ‘Mouse’. German pressure forced the network to move its headquarters from Lyons to Paris, but in the summer there was a further series of blows. In June, ‘Elephant’, one of the principal figures who had remained at Lyons, and the network’s chief supplier of forged documents, was arrested. The following month the head of the south-west region in Limoges narrowly evaded capture by the Gestapo, but some of his staff were captured. Further arrests followed in Toulon, the Alpes Maritimes and even Paris. With the whole organisation under extreme pressure, its leader, Marie-Madeleine Fourcade, was brought out to London.
Only after the war did it become apparent that the disasters of 1943 were in part due to at least one traitor in the Alliance ranks. Elephant’s capture seems to have been down to an Alsatian, Jean-Paul Lien (‘Lanky’), who had penetrated the organisation for the Germans. Lien also engineered the arrest of Fourcade’s effective second-in-command, Léon Faye (‘Eagle’) on his return from a brief visit to London in September. Faye, whose party included ‘Magpie’, a British radio operator who had been sent to work with Alliance in October 1942, was picked up with incriminating documents, three million francs, arms and other equipment which they had brought over from England. Others awaiting Faye’s arrival in Paris managed to get away when the Gestapo began to search the block of flats where they were gathered. More arrests followed, and by the late autumn of 1943 most of the Alliance groups in north-west France and the Rhône valley had ceased to function. Many of the network’s members were shot by the Germans. Faye was executed at a camp in eastern Germany early in 1945. Magpie was more fortunate. In a rare wartime example of a spy-swap, the British, negotiating through the Swiss protecting power, managed to exchange him for a German officer and he arrived back in the United Kingdom in February 1945.12
The Davis and Alliance organisations came under Kenneth Cohen’s P.1 section at Broadway, which handled most of SIS’s French work. Numerous other intelligence networks developed under Cohen’s aegis, with and without Free French co-operation. One was ‘Sosie’, a large and extensive organisation working mainly in the north-east of the country, and the smaller networks included ‘Triboullet’ and ‘Jove’. Their primary means of communication was by radio, but some used courier lines through Spain, while others were provided with air and sea pick-ups. The growth of intelligence activities in France during 1943 produced a growing backlog of intelligence reports because of inadequate communications out of the country. The problem was eased by the development of an independent organisation to handle the material. This was set up by a French diplomat in Madrid (using the code-name ‘Alibi’), who had already run a successful line across the Franco-Spanish border, and who managed to establish wireless communications with networks in France.
All through the war Biffy Dunderdale, P.5 at Head Office, continued to work with separate French networks. In June 1943 he reported that, apart from Polish intelligence, he still got information from his old Deuxième Bureau contacts (with a special line to Bertrand), and also from the Jade organisation. The Polish service had some thirty networks in occupied and neutral Europe, and three hundred agents were employed primarily on SIS work. Dunderdale’s section organised their travelling and transport facilities, while a centre at Stanmore, just north of London, produced agent equipment including radio sets. Some of the Polish networks were very productive. One based in the south of France run by ‘Lubicz’ (Zdzisław Piątkiewicz) had 159 agents, helpers and couriers, who in August and September 1943 provided 481 reports, of which P.5 circulated 346. Dunderdale’s other organisations were rather smaller. He recorded the Deuxième Bureau as having nine stations and Jade four in mid-1943.13
In the autumn of 1942 Claude Lamirault had begun to rebuild the Jade organisation. One group on the Swiss frontier was organised to smuggle urgently needed precision instruments out of Switzerland and across France in fruit boxes, though this was not a great success. There were also sub-groups in Marseilles, Bordeaux, Lyons and Paris (it was in the capital that Lamirault established his headquarters). Each group had independent communications with London, and relied on the centre only for funds and instructions. The Jade Bordeaux group, which actually came to be centred on Paris, was led by Philip Keun (‘Admiral’ or ‘Deux’), a Dutchman raised in England and France who had worked for the French Marine Deuxième Bureau and whom Bill Cordeaux described as ‘an international adventurer with more than his fair share of charm and cunning’. Along with Roman Czerniawski, who had been part of the Polish F2 network, Keun set up his headquarters in a convent which, happily, the Gestapo regarded as above suspicion. Czerniawski had been arrested in 1941 and recruited as an agent by the Germans, but got back to London in October the following year where he pledged to continue to work for Polish and British intelligence and was thereafter run as double agent ‘Brutus’ by the XX Committee.14
One of Lamirault’s most impressive achievements, which helped enormously in his work, was the acquisition of an official identity card and badge used by the Vichy police. Both card and badge were sent to London by courier, faithfully copied by SIS’s false document section and returned to France. The false badges allowed Lamirault and a few chosen members of the Jade group to pass through police cordons, but in the end it was the fake identity card that caused his downfall. He was arrested in Paris on 15 December 1943, trying to force his way into a hotel as a police inspector, but suspicions were raised, and in resisting arrest he shot a real policeman. The following day he was handed over to the Gestapo, who transferred him to Fresnes prison. Attempts were made to rescue him but they failed. He was deported to Dachau, but survived the war only to be killed in a road accident back in France in May 1945. Philip Keun was also arrested, in June 1944, and murdered at Buchenwald the following September.
In November 1942, Menzies asked that Dunderdale should provide him with a fortnightly progress report, drawing his attention only ‘to items of outstanding interest’. ‘Where there is no such item’, he instructed that he ‘would prefer the paragraph to be as short as possible’. Although it is difficult to ascertain exactly who was providing what for Dunderdale’s P.5 section, the surviving reports give a flavour of the intelligence his agents were sending back. In October he reported telegrams from ‘Bertie’ (Gustave Bertrand) giving the names of the heads of French armed forces intelligence and the Vichy French Intelligence Service itself. The next month he stated that his ‘Vichy France Section’ was providing ‘outstanding’ information on French North Africa, including photographs of beaches on the Moroccan Atlantic coast from Agadir to Mazagan; the boom and net defences at the port of Oran; ‘description, equipment and cargo movements’ at Nemours in Algeria; from occupied France ‘a scheme and report of coastal defences’ around Bayonne, and a report about construction work on a submarine shelter in Bordeaux; and from the then still unoccupied zone a report detailing the difficulties of recruitment for the Vichy army. A telegram from his Polish source ‘Rygor’ (Major Mieczysław Słowikowski) on 8 November stated that from 6 November all shipping movements in the port of Algiers had been suspended. Clearly the authorities there had got a sniff of the impending Torch offensive. Ships at sea had received orders to return to the nearest port, while the French submarines Caiman and Marsouin were being held at readiness in Algiers.15
During November 1942 P.5 agents reported on coastal defences at Toulon and Marseilles; the names of commanders of all ships of the French naval forces larger than minesweepers taken from a secret document held by the French Ministry of Marine; details of a German staff headquarters in the Château de Kerlivio near Brandérion in Brittany, with a sketch showing strongpoints which should be destroyed in the event of operations; changes in the armament of the destroyers Gerfaut and Guépard and the cruisers Algérie and Dupleix at Toulon, with plans; reports on the arrival of the Italian Piave and Legnano divisions in France (with the Italian headquarters at Cannes and staff headquarters at Nice), along with the general Italian order of battle; German troop movements on the French Mediterranean coast and air order of battle across France. From Germany itself, Dunderdale’s agents reported on the output of the Schichau shipyards at Elbing in East Prussia, and details of ships under construction or repair in Kiel, Stettin and Lübeck. There were also reports of extensive preparations being made in Germany for gas warfare, including one that ‘Infantry Battalion 151’ had carried out exercises with poison gas at Białystok in Poland; positions of U-boats and their bases in Danzig, with yard serial numbers of submarines under construction; and an OKW appraisal of Soviet and German intentions and the Eastern Front situation dated 30 November 1942.
In 1943, there were reports on Italian naval units on the south coast of France (‘a prompt reply to a query’, according to Dunderdale); the position of ships and air-raid damage in Genoa; more reports on German preparations for gas warfare; and submarine construction at Gdynia, Königsberg and Katowice. In France there were reports on the coastal defences in the Marseilles region and the political situation in both France and North Africa following the Allied landings of Operation Torch. In February 1944 reports arrived on German troop movements to the Eastern Front and a survey of the situation in Italian-occupied Corsica, giving the strength of the occupation forces (20,000-25,000) and identifying units and their locations, with notes on coastal defences, accompanied by two maps and twelve photographs. In March there were reports on the state of the French fleet at Toulon after scuttling; Gestapo arrests in the French Foreign Ministry; a copy of a questionnaire submitted to German SS agents in France; and the drinking-water supply in Tunisia. From Germany came reports on the shipyards at Pillau and Königsberg in East Prussia; the position of the German naval headquarters at Copenhagen; and military convoys for the Eastern Front.
During the spring reports came in on coastal defences along the Mediterranean west of Nice; the defences of Fréjus-Saint-Raphaël aerodrome and beach; and radio-location stations south of Morlaix in Brittany. Names of the German naval staff at Toulon were provided, together with Axis units and the principal radio installations at the Toulon naval base as well as details of torpedoes used in German motor torpedo boats. A Jade agent provided details of damage at the U-boat base in Bordeaux during a daylight raid on 17 May. From the early summer onwards there were increasing numbers of reports on coastal defences, not only along the Mediterranean, but also along the English Channel, the Atlantic in Brittany and south to the Spanish frontier. In July and August reports were received on German troop movements in France; aircraft production; ‘danger areas’ in Brittany; German-occupied aerodromes; the manufacture of propellers; the movement of troops to Italy and the Italian-occupied zone in south-eastern France; and German air order of battle in Poland. At the end of June there was a report on the state of work at shipyards in Bordeaux, Nantes, Saint-Nazaire, Saint-Malo, Le Havre, Rouen and Dunkirk. There was also intelligence about telecommunications - buried cables at Lorient, details of lines between Rouen and Cherbourg and a repeater station at Le Mans.
From the autumn of 1943 reports about German rocket weapons came with increasing frequency from Dunderdale’s P.5 networks. The Germans had, in fact, been developing two V-weapons: the V-1, a pilotless jet-propelled plane launched from a ramp; and the V-2, which was a longer-range rocket. Both carried about a ton of explosive. SIS played a major role in providing intelligence about these programmes. In 1940 the Service had passed on vague reports of German experiments with long-range weapons, but it was not until late 1942 onwards that more specific information began to be received. Elgar, the Danish engineer who was able to travel in Germany, provided three reports about a new rocket weapon between December 1942 and March 1943. Another source reported in January and February 1943 that a factory had been built at Peenemünde to manufacture a rocket, and a third source in February reported that a rocket with a ten-ton warhead and a range of a hundred kilometres had been developed there. During March Broadway acquired a much more detailed report from ‘a most reliable and expert source which has provided most valuable information over a long period’. The report, which described a series of rocket trials, drew on information provided by forced labourers from Luxembourg who had been drafted to work at Peenemünde. Further information came through Switzerland. As recalled by one SIS officer, the SIS station in Berne received from a Luxembourger ‘a very dirty ragged sketch plan of some big contrivance being constructed by the Germans at an island call[ed] Peenemuende sometime in 1943. This filthy sketch plan seemed to make little sense and it referred to a launching of a “pilotless aeroplane” which had a range of 250 kilometres.’ The head of station, Fanny Vanden Heuvel, was away at the time and his assistants decided ‘to telegraph the most we could make out of this piece of paper. When Fanny returned he chided us for being such fools as to believe such nonsense, and said we would undoubtedly get a rocket from Dansey for wasting cypher groups!’ But Broadway’s reaction was quite the opposite: Berne was congratulated for ‘most valuable information’ and told ‘that anything further on this subject should always be telegraphed “Most Immediate”’. Corroborative information about the installations at Peenemünde from Polish sources and RAF photographic intelligence enabled an extremely detailed plan of the facility to be prepared and used for an Allied air raid on 17-18 August 1943 which so badly damaged the site that the weapons programme was put back by at least two months and the Germans had to disperse both the experimental facility and the manufacturing side to other locations across the Reich.16
SIS agents also helped to pinpoint V-1 and V-2 launching sites in France. The most outstanding contribution was made by Michel Hollard and his ‘Agir’ network. Hollard, sales representative for a company that made gas generators for motor cars (a job which allowed him to travel all over France), offered his services to the British assistant military attaché in Berne in January 1942. He was at first rejected, but the report he left, on the French motor and aero industries, was graded so highly that, when he turned up again in May, SIS recruited him at once. Hollard built up a network, including many railway workers, and began by reporting on economic matters and German order of battle in France. In August 1943, however, one of his agents drew his attention to a site near Rouen where the Germans were imposing unusually precise specifications on the French building contractors. Masquerading as a labourer, Hollard investigated for himself and found mysterious ‘miniature runways’ being constructed which were all carefully aligned on bearings pointing towards London. The Berne station was not particularly impressed by the information, but when it was sent on to London it was realised that they were launching-ramps for the V-1 flying bomb. Hollard was instructed forthwith to concentrate exclusively on locating other such sites, which he and his colleagues did with extraordinary success.
By October 1943 Agir had located over a hundred V-1 sites, and Hollard, who had ‘an arrangement’ with the Chef de Gare at Rouen, was able to measure and take a series of photographs of a railway truck-load of V-1s. In addition, he stole an architect’s blueprint of a V-1 site which allowed London to construct an exact scale model which, complemented with photo-reconnaissance, enabled the RAF to carry out effective bombing of the sites (which Agir in turn reported on). One of Hollard’s sub-agents, ‘Z.187’, was an eighteen-year-old Frenchman who came to the British mission in Berne hoping to go on to England to join the Free French forces. When told that this was not possible he offered to work as an agent. SIS arranged for him to be taken on by Swiss Military Intelligence (which facilitated crossing and recrossing the frontier) and simultaneously join the Agir organisation. He helped identify V-1 sites and pinpointed the first V-2 launching site at Watten, near Calais. He also regularly supplied conventional military intelligence, among which was a German General Staff map showing the deployment of an infantry division in the Pas de Calais region – alleged to have been specially prepared for an inspection tour by Hitler – in ‘complete detail down to platoon formations, with all the defence works drawn in, including underwater anti-tank and anti-personnel obstacles along the shore’. Z.187 was caught in June 1944 and never heard of again, presumed shot by the Gestapo. Hollard was more fortunate. Having crossed the Swiss frontier ninety-eight times before being betrayed and arrested in February 1944, he survived the war.17
SIS and D-Day
Operation Overlord, the successful invasion of France on 6 June 1944, in which some 150,000 troops were landed at a cost of little over 5,000 casualties, was a stunning achievement in terms of Allied and inter-service co-operation, careful planning, logistics, timing and, not least, the courage of the individual soldiers, sailors and airmen involved. It was also a tremendous intelligence success. Indeed, no previous military operation in history had been so well supported and sustained by the intelligence available for the assault and its range and quality were a testament to the combined efforts of many agencies: signals intelligence from Bletchley Park; aerial reconnaissance from the RAF; deception operations run by the XX Committee; and detailed surveys of the beaches themselves obtained by COPPs (Combined Operations Pilotage Parties), sometimes lying offshore for days in midget submarines measuring tidal currents and beach gradients and taking sand samples.18 But there were limitations to what could be done. Signals intelligence, for example, which was so valuable where the enemy used wireless, could provide only fitful coverage about the deployment of static German forces, as in northern France, who relied principally on land-lines for their communications. Much of the information about this came from the kind of material that was SIS’s stock-in-trade: old-fashioned human intelligence – humint – gathered by a host of individual agents and helpers over the preceding two years and more.
Information on German deployment, coastal defences and communications in northern France, such as that provided by Dunderdale’s P.5 networks, was carefully assembled by section MI14 in the Directorate of Military Intelligence (for German order-of-battle information), and the Combined Intelligence Section (CIS) of General Headquarters, Home Forces, which had been given responsibility for collating all intelligence along a thirty-mile-deep coastal strip from Den Helder in the Netherlands to the mouth of the River Loire in France. From June 1942 to May 1944 CIS produced a series of weekly reports, code-named ‘Martian’, which systematically laid out the intelligence under seven headings: strategic survey; enemy forces; topography and maps; transport and industry; police and civilians; air; and naval. The reports were illustrated with aerial photographs, reproductions of documents and plans of coastal defences supplied by agents.19 From March 1944, MI14 began to issue more detailed ‘weekly notes’ or ‘weekly summaries’, classified ‘TOP SECRET(U)’, which provided a greater emphasis on German troop dispositions, general order of battle and the nature of fortifications on the northern French coast.20
In order to sustain the supply of tactical intelligence for the invasion forces during and immediately after Overlord, and aiming also to provide an alternative network in the tactical areas around the Normandy bridgehead should existing clandestine organisations fail as a result of enemy counter-measures, in late 1943 the ‘Sussex scheme’ was developed. This was a joint scheme between SIS, the American OSS and the Free French BCRA. Fifty intelligence targets were identified, half in the planned operational area of the British 21st Army Group and half in that of the United States forces. Each was to be covered by a French two-person team. But there were delays in assembling the agents and wireless operators, as not only had the BCRA in London to refer all matters to Algiers for a decision, but it had been decreed that General de Gaulle and his staff were not to be given details of Overlord until the actual day the invasion was launched. There were also problems with the personnel themselves. Just before Christmas 1943 Kenneth Cohen (Chief Staff Officer, Training, and chair of the inter-Allied Sussex Committee of Three running the scheme), complained to Tony Morris in Algiers that the men the Free French were supplying were of ‘low medical category or indifferent morale’ and that ‘at present rate we shall fail to implement our operational undertakings to the Chiefs of Staff, and responsibility for this will lie with the French’. For his part, Morris thought that ‘the whole B.C.R.A. attitude of mind seems to be much more directed towards internal politics and the Mouvements de Résistance than towards the collecting of military information’.
So concerned were Cohen’s committee about the risk of contacts with potentially compromised existing agent networks in France (and in keeping with the intense security maintained for all aspects of Overlord) that they considered the risky possibility of dropping the teams in blind. In the end, however, they decided to send in a preliminary ‘Pathfinder’ mission to prepare the way, when the twenty-five-year-old ‘Jeanette Gauthier’ was parachuted in on 6 February 1944. Based in Paris, she travelled west to Alençon, and south to Lyons, Bordeaux and Châteauroux, locating suitable dropping zones and arranging reception committees. Between February and April 1944 she personally met the first nine Sussex teams that parachuted into France and led them to safe houses in the Paris region.
Considerable effort was put into agent training, which was modelled on SOE experience and run by British and American officers based at Prae Wood in St Albans, close to the Sussex administrative headquarters at Glenalmond, a commodious Victorian house where Section V had been located earlier in the war. The training programme included a week-long exercise where pairs of potential agents were scattered across England, given locations of safe houses and instructed to send information (for example about military convoys) back by radio. A report of the fourth of these exercises, based in the East Midlands in early May 1944, gives a flavour of what was involved. With an emphasis on ‘tradecraft and its practice’, the students performed well and no one was ‘arrested in “flagrante delicto”’. The local police were in on the exercise and the trainees ‘all had to be brought in on an obviously framed up charge in order to undergo the experience of interrogation’. The police reported afterwards that the teams had stood up well to questioning (one individual was ‘interrogated for some 8 hours’ by the Nottingham Police) and they were commended for the hiding of ‘maps, documents and compromising papers’, as ‘searching of their rooms in their “safe” houses in no case produced any result’. On the radio side, ‘105 messages were sent, of which only 2 were indecipherable and of the remainder 70% were without fault’. The students were carefully debriefed afterwards. One female radio operator had put her partner in danger by telling the police that she had met him in Nottingham, while unbeknown to her he had claimed to have no ‘friends or acquaintances’ in the city. This was a salutary lesson and the training officer observed that the woman now seemed ‘to have grasped the importance of detail and of co-ordinating cover stories’. Only one participant refused to play the game, and had made ‘little or no attempt to learn his cover story or to apply it under interrogation’. His attitude was ‘that the matter was quite childish and he was merely there on an exercise which in fact was well known to the police concerned’. This did not go down at all well and he was ‘severely talked to about his general manner during the exercise’. He admitted ‘that he played the fool’, and his partner thought he had offended ‘from a feeling of “je m’en foutism”’, but ‘that he would in fact take the matter very seriously if it were the real thing’.
At the time of D-Day, although only fifteen Sussex teams had been dropped into France, ninety-five students had completed their training, and by mid-June twenty teams were in the field, reporting both by wireless and using the new ground-to-air Ascension device. Between 5 and 8 June, one team, ‘Ossex/6’, sent in reports which, along with German order-of-battle details, first identified the Panzer Lehr Division, about which one Supreme Headquarters Allied Expeditionary Force (SHAEF) officer remarked: ‘If the Sussex scheme produced no other intelligence, it would have been worth while for these reports alone.’ Following the invasion, other teams supplied immediately useful information, such as ‘Brissex/20’, who on 7 July reported on German barge traffic along the Seine, as a result of which ‘the R.A.F. took immediate action’, sinking twelve barges ‘during the following 24 hrs. from the details given in his message’. By the end of August 1944 over thirty of the forty-five Sussex teams despatched had made radio contact with base, and up to 800 messages had been sent.21
During the run-up to Overlord the possibility of a co-ordinated assassination operation was discussed. Targeted killings had been contemplated at intervals since the beginning of the war and this was in part what SOE potentially existed to do. In May 1942 SOE-trained Czechoslovaks had assassinated Reinhard Heydrich, the German Protector of Bohemia and Moravia. Another SOE operation, ‘Ratweek’, in 1943-4, aimed to kill as many collaborators and members of the German security forces as possible across Europe, though in France it was successful only in Lyons where one agent ‘disposed of’ eleven individuals.22 In April 1944 SHAEF, suggesting that likely victims might include Rommel, Rundstedt and possibly also Vichy collaborators, asked the Foreign Office to raise the matter with SIS (and curiously not with SOE). Menzies wanted clearer guidance as to what precisely was required, but, assuming ‘that they must be thinking of the removal between now and D. Day of personalities whose liquidation might actually assist the Overlord operation’, he said that ‘the compiling of lists would seem to be a fairly simple matter’. Nevertheless he was worried that any such action ‘might automatically lead to reprisals against hostages or United Nations personnel in enemy hands’, and wished to know ‘exactly what effects it is hoped to attain by the action proposed’. SHAEF came back with the proposal that, rather than German military commanders, the sorts of people they had in mind were ‘para-military and civilian German personnel in key positions in France whose removal at the critical moment might really be a blow to the German war effort’. These might include ‘Abwehr and S.D. [Sicherheitsdienst] characters, important political figures, transport chiefs, heads of supply and other economic organisations etc’. They did not want any French names, as they were ‘strongly in favour of leaving it to French resisters to select their own victims’.23
On 11 May Menzies told Peter Loxley that while SIS had prepared a list of possible targets (which he was not as yet forwarding to him), the Service did not believe that their removal would ‘have much, or indeed any effect on the efficient functioning of so widespread and highly organised a machine’ as the Germans had in France. He also recommended that Loxley consult Bill Cavendish-Bentinck, the chairman of the Joint Intelligence Sub-Committee, on the matter. Cavendish-Bentinck agreed with Menzies ‘in disliking this scheme, not out of squeamishness, as there are several people in this world whom I could kill with my own hands with a feeling of pleasure and without that action in any way spoiling my appetite’, but because it was ‘the type of bright idea which in the end produces a good deal of trouble and does little good’. Above all there was the risk of bloody reprisals. Nevertheless, if the French liked ‘to assassinate Germans or collaborators’, he added, ‘we should not deter them’, but ‘we should steer clear of this business’ and ‘should not ourselves designate persons to be liquidated, either German or still less French’. He noted the likelihood ‘that for every successful assassination’ there would be ‘two or three failures’ and observed that ‘if assassination were easy many statesmen and high officers would have come to a violent end’. Looking to the future, Cavendish-Bentinck worried that if German civilian officials were murdered in France, it would ‘probably start a wave of murderings’ which might continue after the British had occupied enemy territory, ‘with the result that members of our Control and other Commissions will become poor risks for the insurance companies’. Cadogan shared these negative opinions and SHAEF were informed accordingly.24
SIS’s part in Overlord drew on its experience in North Africa and the Middle East and reflected the successful organisation which had been developed there in 1942-3. Thus, as well as the Sussex teams, echoing SIS’s No. 1 Intelligence Unit which had been attached to Alexander’s 15th Army Group in Italy, No. 2 Intelligence (Unit) Section (or, more succinctly, but no more revealingly, No. 2 I(U) Section) was attached to the headquarters of General Montgomery’s British 21st Army Group, and provided the link to SHAEF on intelligence matters, both with London and in the field. On 8 June Menzies appointed Colonel Guy Westmacott to take over the section, with an officer who had hitherto been running the Sussex programme as his deputy. In a move that signalled the start of the Service’s anticipated peacetime deployment in formerly occupied Europe, the deputy was given the symbol ‘27000’, which belonged to the head of the Paris station. He was charged with obtaining ‘information deep behind enemy lines’, both ‘by the recruiting and infiltration of special agents and by establishing direct contact with S.I.S. organisations previously controlled from the U.K.’. At Broadway, No. 2 Intelligence Unit came under Commander Cohen, who now as Controller Western Europe (CWE) was responsible for SIS operations ‘in France, Belgium, the direct penetration of Germany and Czechoslovakia; also the clandestine activities of M.I.9. in those areas’.
Three other types of SIS section served with Allied formations in the field. There were a number of Special Counter-Intelligence Units, attached to major headquarters, which distributed sensitive signals intelligence material from Bletchley Park and were responsible for the security of that material. There were Special Liaison Units which handled Bletchley Park air intelligence and passed it on to theatre air headquarters. Finally, there were the paramilitary Special Communications Units which had been set up by Gambier-Parry to handle all SIS’s field communications.25 No. 2 Intelligence Unit’s role in the provision of tactical battlefield intelligence fell away after the Allies had broken out of the Normandy bridgehead in mid-August and begun rapidly to move eastwards, soon overrunning most of the Sussex teams in the process. Thereafter, it acted as a link for the passing of information from London to 21st Army Group. Following the Allied landings in southern France on 15 August, and the liberation of Paris a week later, by the end of September all but a few parts of eastern France had effectively been liberated. On 23 October the Allies recognised General de Gaulle’s administration as the provisional government of France and for SIS the task now became a combination of closing down wartime French networks, settling remaining obligations to agents (including those who had worked for SIS in preference to de Gaulle), sorting out its future deployment in the country and establishing the basis of postwar liaison with the French intelligence authorities.