Codeword Overlord

by Nigel West

  … in spite of all the attendant risks, the enemy will probably attempt a second landing in the 15th Army’s sector, all the more so as public opinion will press for the elimination of the sites of the long-range weapons firing on London. The disposition of the forces still available in England suggests attacks primarily against the sector between the Somme and the Seine by divisions assembled north of the Thames, but also against Belgium and southern Holland. At the same time, however, surprise attacks designed to effect the capture of one of the large ports in Brittany cannot be ruled out.10

  Hitler discussed this assessment on 12 July with Admiral Dönitz, noting that the latest intelligence included reports that American troops had moved from south-east England to new positions north of the Thames, and speculated that this shift could have been prompted by the discomfort of the V-1 attacks, or a preparation for another amphibious assault. The debate ended in an agreement that the issue could only be resolved by aerial reconnaissance, but the matter was pursued again in conversation the next day when Dönitz raised the possibility of a landing in the Skagerrak, while Hitler dismissed the idea and repeated his belief in a second wave in the Pas-de-Calais or, less likely, an attack on the Belgian–Dutch coast, which suggested a route to the Ruhr.

  9

  MUSGRAVE

  ‘Our infiltration of the French resistance movement also

  provided us with some very valuable indications and clues.’

  Major Roger Michael

  Deputy Chief, FHW

  One of the relatively unknown dimensions to OVERLORD was the contribution to be made by the French resistance. Special Operations Executive (SOE) had been planning its own participation in D-Day since August 1943, and had established a planning section, designated MUSGRAVE, to give the invasion the benefit of its organisation in France, which would amount to 137 separate wireless transmitters located in occupied territory.

  Allegedly, SOE’s assets in France were very considerable and by May 1944 were estimated at 100,000 men available to take orders directly from London. Another claim, made by the head of the maquis, Michel Brault, was 35,000 to 40,000 ‘well-armed’ men, of whom 10,000 had sufficient ammunition to continue fighting beyond a day. He also added a further 350,000 unarmed supporters, plus 500,000 railwaymen and 300,000 trade unionists.

  In terms of weapons, SOE calculated the delivery to France of huge quantities of materiel by 10 May, including 74,131 Sten guns; 27,047 pistols; 16,495 rifles; 3,205 Bren guns; 572 Bazookas; and 404 light machine guns. Responsibility for co-ordinating the irregulars in the field was delegated to a new entity, Special Forces Headquarters (SFHQ), which was created on 10 January 1944.

  However, enthusiasm for SOE’s activities was not universal, given the reservations of both MI5 and SIS concerning the level of enemy penetration, and the organisation’s role in OVERLORD was discussed by the JIC on 22 April, as was noted by Guy Liddell:

  The JIC have tried to estimate the potentialities of French resistance. They say that up till the end of March some 80,000 weapons had been distributed to French resisters. This figure does not include known losses but a deduction of 20 percent must be allowed for unknown losses and for deterioration. There have been further deliveries of about 20,000 weapons from North Africa and there are also believed to be certain stocks of French arms already in the hands of the resisters. It is considered, therefore that in France about 100,000 men are provided with arms and ammunition. The total numerical strength of the resistance movement is much greater but cannot be considered of military value without arms. There is said to exist an organisation to co-ordinate activity of all-important groups. It has been built up from its beginning since about a year ago but the standard of efficiency varies widely from bad to excellent in individual groups. There is a tendency especially in the Armee Secrete to create a large central directing organisation in France with various committees and a staff on semi-military lines. This is an obvious danger from the point of view of penetration. A good many political differences exist, particularly between the supports of Henri Giraud and Charles de Gaulle. The Communists are hoping and working for power after liberation but are putting the elimination of the Germans first. They appear to be the most efficient organisation. Vichy are taking strenuous counter measures which are supported by Darnand. The effect of present-day activities against the German war effort is thought to be appreciable, although it cannot be stated that the German military machinery is thereby seriously impaired. The general conclusions are that day-to-day activities are contributing to the weakening of the German war effort but that the real value of resistance lies in the building up of the means to strike on D-Day. It is felt that at the worst, resistance cannot fail to be of value owing to the extent to which it diverts German energy, and at the best it may so derange the German lines of communication that it cannot fail to be a determining factor in the battle. The above information is I presume mainly based on SOE reports which have presumably to be taken at their face value.

  Certainly SIS was not prepared to take SOE’s exaggerations at face value, and on 8 February 1944 the ever-sceptical Vice Chief Claude Dansey circulated a report based on information from the Free French intelligence adviser André Dewawin that predicted only 2,000 men could be relied upon to take up arms against the occupation when called upon to do so.

  The misgivings about SOE’s integrity extended to the Free French authorities and the decision had been taken to exclude Charles de Gaulle, then in Algiers, and his staff from any involvement in D-Day until the operation had begun. The embargo also extended to their subordinates, but not initially to the participation of SOE’s RF (Republique Français) section, which worked closely with the Free French Bureau Central de Renseingement et d’Action (BCRA). However, a planning meeting held at his suite in Claridge’s by Colonel Bill Stirling, the commanding officer of the 2nd SAS Regiment, with a French BCRA agent, Lieutenant Jean Rosenthal (code-named CANTINIER), would change the situation dramatically. The issue under discussion in the hotel had been the SAS’s future deployment in France, and Rosenthal had shocked those present by pointing to a map of Normandy and appearing to know precisely where the invasion was headed, with his thumb actually placed on the beaches. This incident was reported immediately by SOE’s security director, John Senter, to MI5, and consequently both SIS and SHAEF imposed a ban on the infiltration of any French agents until after the landings, on the assumption that at least some of Rosenthal’s BCRA colleagues shared the same knowledge. A Parisian jeweller by trade, Rosenthal was later dropped very successfully from a Hudson bomber with SOE’s Richard Heslop (XAVIER), on 21 September, into the Haute-Savoie, but was never challenged about this potential lapse of security.1

  Given the level of German influence over SOE’s French networks, the temptation was to regard the circuits as potential channels of deception to the enemy, but this approach was rejected, based on the experience gained in September 1943 when just such an attempt had been made in an effort to convey the proposition that the Allies were planning an imminent cross-Channel invasion in the Pas-de-Calais, with TINDALL, a simultaneous threat against Norway. A sub-plot, code-named WADHAM, was for a follow-on landing in the Bay of Biscay by American troops led by General Jacob Devers to seize the deep-water port of Brest. The aim of the three inter-connected schemes was to pin down Axis troops in north-west Europe so as to relieve the pressure on the Russian and Italian fronts. Another objective was to lure the Luftwaffe into the air within the RAF’s fighter range, thereby offering an opportunity to inflict serious and lasting damage and establish air superiority.

  The overall plan, code-named COCKADE and approved by Freddie Morgan and the Chiefs of Staff, was executed on 8 September 1943 when an unconvincing number of landing craft sailed into the Channel to join an assembled convoy, in what was described locally as a large exercise. In the total absence of the Luftwaffe, the undertaking was considered somewhat a failure, although in retrospect it did firmly create two imaginary British armies and
had the merit of providing cover for the arrival of large quantities of American troops. The deception narrative was that Eastern Command, based at Luton, had been transformed into the Sixth Army, and Scottish Command in Edinburgh had become the Fourth Army, and these reforms were accepted by the FHW, and would prove helpful when the Sixth Army was allocated to FUSAG. To boost the credibility of the Fourth Army, General Sir Andrew Thorne was publicly appointed its commander. Thorne was well-known to the Germans, having served as military attaché in Berlin from 1932, so the narrative appeared plausible.

  It was hoped too that even as a large-scale exercise, STARKEY, a component of COCKADE, might be seen by the enemy as significant in the Allied planning process for the invasion summer of 1944, but Ops (B) was too sanguine about that interpretation. An unexpected consequence of STARKEY, originally peddled as a real invasion and not just an exercise, was the loss of GARBO’s first agent, ONE, who was supposedly a KLM steward acting as a courier, posting ARABEL’s mail in Lisbon. He had been the principal double agent deployed in support of STARKEY, and when the operation failed to materialise, he resigned on 19 November 1943. However, the most controversial aspect of STARKEY was SOE’s involvement, which was marginal.

  SOE agreed to ‘increase its encouragement to Resistance’ in the target region and, through the transmission of bogus wireless messages, noticeably escalated the volume of signal traffic so the enemy intercept operators would observe the increase. This controversial policy, of heightening expectations and then delivering disappointment, was considered likely to undermine confidence in the Allies, so it was agreed that, nine days before the operation, SOE would circulate a ‘stand-down’ message and the Political Warfare Executive (PWE) would have the RAF drop leaflets to the civilian population explaining that no invasion was contemplated. Additionally, the BBC was invited to report this development in its regular news bulletins. This measure, intended to retain the confidence of the French population, was also calculated to confuse the enemy, and it probably did. It also caused dissention within the PWE and BBC, resulting in the replacement of the head of the organisation’s director of the French regional department, Colonel Gielgud. Written off as a propaganda blunder, the embarrassing episode demonstrated the hazards of employing SOE as an instrument of deception, consciously or unconsciously.

  As a rehearsal for D-Day, STARKEY provided some valuable lessons, and produced a directive setting out three conditions that would prevail if the same issues emerged during OVERLORD: To avoid costly reprisals, the French resistance would not be called to action exclusively to promote a deception scheme; any participation in a deception scheme would not be at the expense of operations in support of OVERLORD; and no action in support of deception would be taken prior to D-Day by any resistance group. In practice this meant that F Section could indicate ‘a very slight bias to the Pas-de-Calais area’ and avoid prejudicing FORTITUDE, but little more.

  The debate prior to D-Day concentrated on tactics. Was SOE to focus on building up its existing reseaux in the urban areas and conduct small-scale sabotage when called upon, or should it develop links with the maquisards in the countryside in the hope that, having been alerted by radio, they would emerge to engage the enemy, thereby tying them down far behind the battle front? The former strategy had a high nuisance value but the destruction of power installations and the mining of roads, bridges and other minor enemy assets would be scarcely likely to influence the outcome of the greater conflict. It was criticism of this kind that had prompted the Minister of Economic Warfare, Lord Selborne, to circulate a review of SOE’s sabotage coups in January 1944 in which he listed thirty-seven instances where F Section had inflicted significant damage on the German war machine during the previous twelve months. Yet, beyond acts of sabotage, there were tremendous risks involved in developing and encouraging large bodies of ill-disciplined men, who could hardly be expected to take on regular troops on equal terms. Even with the advantage of local knowledge and support, the guerrillas would be decimated by enemy armour and air power if drawn into any extended confrontation. It was a choice between pinpricks and potential massacres that, with the uniquely complex elements of French politics and the active participation of American and British personnel, made such decisions concerning strategy all the more crucial.

  The attraction of harassing the enemy behind his own lines, and thereby preventing mobile reserves from reacting swiftly, was obvious, but had to be balanced against some potential disadvantages, such as the security implications of giving advance notice to SOE circuits that probably had been penetrated to some extent. German counter-intelligence had already taken a heavy toll on F Section, even if the management in Baker Street was itself unaware of the actual scale, which would not become evident until the end of hostilities. At the end of the war Robert Bourne-Patterson, an F Section planning officer, drafted an after-action report in which he found that of the 393 officers sent to France by F Section, 102 had been killed, and a further seventeen had been arrested and survived. These dismal statistics suggested that German counter-intelligence had been rather more efficient than appreciated hitherto.2

  Not fully aware of the challenge, SFHQ adopted a series of projects, including Plan VERT, which involved attacks on railways, and Plan VIOLET, designed to disrupt enemy communications. A component of Plan VERT, put into action before D-Day, was Plan GRENOUILLE, the sabotage of railway turntables, as recommended by the rail company itself, which produced a secret scheme detailing all the SNCF system’s vulnerable points, probably motivated by the hope of ending the hopelessly inaccurate carpet-bombing of marshalling yards, as advocated by Air Marshal Tedder and the RAF’s scientific adviser, Solly Zuckermann. Similarly, just as the SNCF privately assisted SOE’s search for targets susceptible to sabotage, the PTT offered the services of 700 technicians who maintained France’s long-distance underground cable network. The objective was to erode German confidence in the 6,800 miles of landlines laid across France since 1940, and discourage the employment of vulnerable couriers, forcing them to rely on wireless communications, which could be intercepted. On D-Day itself, pre-planned saboteurs went into action and cut twenty-eight trunklines, including the crucial telephone wires linking Avranches to Saint-Lô, and Cherbourg to Saint-Lô and Caen. These tactics resulted in a greater use of Enigma wireless circuits, and of the single Geheimschreiber radio-teletype link between OB West and Berlin, which had been installed in October 1943, and read as JELLYFISH at Bletchley Park from April 1944. Access to this new source would profoundly improve MI-14’s somewhat patchy assessments of the enemy’s order of battle in Normandy.

  A product of the newly introduced Colossus apparatus, JELLYFISH would provide a vast quantity of decrypts, including numerous contemporaneous military situation reports addressed to the OKW, and current intelligence assessments from Berlin to von Rundstedt’s staff, a category of high-grade material not previously entrusted to Enigma. It proved its worth on D+4 when the decrypt of a signal sent the previous evening identified the location of PanzerGruppe West’s new headquarters in an orchard at the Château La Caine, just north of Thury-Harcourt, which resulted in an immediate raid by RAF Typhoons that killed eighteen staff officers and seriously wounded the commander, General Geyr von Schweppenburg, causing major disruption. PanzerGruppe West then withdrew to Paris, making a planned counter-attack impossible, being unable to resume operations until 28 June.

  JELLYFISH would not last long, and new security procedures suspended the source in July, but a similar FISH non-Morse teleprinter link, from Berlin to Field Marshal Albert Kesselring’s headquarters in Italy, designated BREAM, would include assessments relating to the battle in Normandy and thereby assist the MI-14 analysts.

  In order to avoid compromising the PTT personnel, and to ensure they were available to restore full service when the Allies arrived, SFHQ imposed strict security conditions on agents inserted into France, and on the volume of routine radio traffic that was enhanced, when necessary, by nonsense transmissio
ns.

  These measures were intended to protect OVERLORD but, almost inevitably, there were some breaches. However, the scale of the enemy’s penetration and control over F Section networks was far beyond anything imagined in London, with the SD’s Dr Josef Goetz estimating that he had fifteen circuits under SD control that might give him advance notice of the imminent attack. Based on the second floor of the SD’s Paris headquarters at 84 Avenue Foch, and assisted by his interpreter Ernst Vogt, the former Hamburg schoolteacher successfully circumvented SOE’s radio security procedures, and his confidence was founded on his experience of the principal German radio counter-intelligence organisation, the Funkabwehr. In September 1946 he would give SOE’s Vera Atkins a detailed account of his successes, some of which he credited to an SOE traitor, John Starr (EMILE, later BOB) of the ACROBAT circuit, who had willingly collaborated with his German captor, and only narrowly had escaped post-war prosecution.3 Another co-operative prisoner was Gilbert Norman (ARCHAMBAUD), who did not survive Mauthausen. Goetz’s third SOE source was Henri Déricourt (GILBERT), the air movements officer responsible for supervising F Section’s transport aircraft, who was recalled to London in February 1943, by which time he had handled fifty-four individual agents. All had completed their journeys unhindered, courtesy of the sponsorship of Goetz and the Paris SD chief, Karl Bömelberg. As well as compromising everyone who took Déricourt’s flights, he also dealt with SOE’s confidential mail, all of which was sorted and copied by SD analysts before being returned for onward delivery.

  According to Colonel Ernst de Bary, who commanded the Funkabwehr’s counter-intelligence branch from 1942, his organisation made ‘30 direct arrests in 1941, 90 in 1942, 160 in 1943, and approximately 130 in 1944’.

  In all this amounted to some 410 cases, in about 20 per cent of which the civilian police forces lent their assistance. Moreover, indirect arrests could be made on the basis of the information compiled by the final evaluation section. This source contributed approximately 140 additional cases during the same period. Thus a total of 550 arrests stemming from radio counter-intelligence were effected in four years. The figures for arrests made by other counter-intelligence agencies and the [RSSA] are not known.