The Mystery of the Undetected Radios (full)

Chapter 1

A G.P.O. detector-van, circa 1925. Note the well-camouflaged postilion on the roof of the vehicle.

The successful invasion of France by the Allied Forces in May 1944 was achieved largely because of the successful project to mislead the Germans about the planned landing site – the Pas de Calais rather than the actual beachhead in Normandy. Operation OVERLORD was a winner because of Plan BODYGUARD and the latter’s Directive for FORTITUDE. Yet the ability of the Abwehr – and hence the Wehrmacht – to be deceived so comprehensively by the group of double agents recruited by MI5, and run by Britain’s Double Cross (XX) Committee, opens up the management of these agents to some searching inspection. The relentless question, posed by many historians of this period, can be expressed essentially as follows: Why did the XX Committee allow such intensive wireless transmissions (especially from agent GARBO) to take place, knowing that any respectable interception agency would have located them and arrested the operators?

The Strategic Dilemma

In last month’s blog, I started analyzing the statement by Sir Michael Howard, the author of Volume 5 of British Intelligence in the Second World War (‘Strategic Deception’) concerning the challenges the Abwehr faced in exploiting its agents in Britain. Howard wrote: “The most satisfactory channel was radio transmission, but for this three problems had to be solved. First, the agents had to be provided with transmitting and receiving sets, and after June 1940 this was easier said than done. Secondly their missions had to evade detection by the security authorities; and finally they had to communicate in a secure cypher.” What were Howard’s sources in divining this strategy?

Howard was strangely subdued as to how the Abwehr went about its mission, or how the German intelligence service overcame these supposed challenges. He records a few sporadic observations, and refers to Professor Sir Harry Hinsley’s and C. A. G. Simkin’s Volume 4 (‘Security and Counter-Intelligence’), which also writes about the XX Committee, but says little about Abwehr strategies. As I have pointed out before, it seems that the authors of this volume did not want to engage the technical challenges too energetically.  The discussion of such is relegated to a very amateurish and inadequate Appendix 3 (‘Technical Problems Affecting Radio Communications by the Double-Cross Agents’), written by an anonymous ‘former MI 5 officer from his personal experience’, which only skims the surface of the intricate subterfuges and negotiations that were undertaken to allow the agents to communicate with their supposed controllers in occupied Europe.

In fact the wireless strategy of the Abwehr was haphazard, tentative, and frequently incompetent. At the end of 1940, when the main wave of agents arrived on UK shores, mainly by parachute, most of the spies were equipped only with transmitters, not receivers, as they were planned to be used only as temporary informants before the imminent invasion. Howard’s description of the Abwehr’s objectives thus misrepresent its intentions. The chances of a heavy wireless set, strapped to the parachutist, surviving the fall from a low height, were not considered healthy. When one of the surviving sets (belonging to Wulf Schmidt, agent TATE) was at last made operable, the Abwehr gave advice for charging it up from motor-cycle batteries that, if followed, would have burned the valves. When Lily Sergueiev (agent TREASURE) was, after an inordinate amount of bureaucratic muddle, finally given her visa by the Abwehr to leave France for Madrid, as late as October 1943, her advised method of communication was still invisible ink, even though she had been trained in wireless operation. The acquisition of agent GARBO’s (Juan Pujol’s) wireless set had to be arranged by MI5: TREASURE likewise had to convince her Abwehr boss that she could go to Lisbon to pick up a set, under cover of being employed by the Ministry of Information. Agent BRUTUS (Roman Czerniawski) appealed to his controller to send him a more powerful transmitter, but the Germans prevaricated. It was as if the Abwehr only very late in the war considered seriously the use of wireless as a means of communicating intelligence.

The British, on the other hand, as they started to realise that the double-agents could be used more for strategic deception than simply gathering information about the enemy’s intentions, concluded that critical information needed to be forwarded in a timely fashion. Letters with invisible ink or microdots, sent to intermediaries, simply took too long, and the XX Committee thus inveigled the Abwehr into more intensive use of radio. Yet this was a very delicate path to follow, for the double agents would have to deploy their wireless sets as if the whole exercise had been initiated by the Abwehr itself, complemented by the ingenuity of the spies themselves. Thereafter, once radio communication had been set up, the XX Committee, and the officers of B.1.a of MI5 who controlled the agents and liaised with the Radio Security Service (RSS), had the outwardly conflicting goals of a) ensuring crisp and reliable communication between the agents and their handlers, and b) pretending that their powers of radio-detection and location-finding (generally known as ‘D/F’, for ‘direction-finding’, even though the term does not explicitly include ‘locating’) were so poor that their own agencies were incapable of intercepting the transmissions and hunting down the culprits.

One can thus present the dilemmas faced by the architects of strategic deception as follows:

  • The authorities had to ensure that no unauthorised enemy transmissions were made from UK soil. Hence good detection and direction-finding were paramount.
  • They had to be confident that the Abwehr trusted the communications of their agents and that they were kept in place providing ‘useful’ information. Hence D/F had to be shown to be inadequate, or extraordinarily imaginative transmission techniques, masking location, or using wavelengths close to dominant broadcasters, had to be used.
  • At the same time, MI5 had to discourage illicit broadcasts by embassies and governments-in-exile, since information might be passed on that would undermine the ‘bodyguard of lies’ being woven by the official deception agency. In order to do this, an effective interception and D/F operation had to be managed.
  • Thus all illicit broadcasts, by such agencies, or by rogue private operators, had to be shut down. If news of this got out, the Abwehr would no doubt hear of it, which would lead them to conclude that the operation was an effective instrument of surveillance. How, therefore, could the Abwehr be convinced that the British D/F operation made sense?
  • The British experts needed to keep informed about the capabilities of the Nazi D/F operation. This process would mature soon, when SOE, the Special Operations Executive, started up its insertion of wireless agents into France, Belgium and the Netherlands, and later in the war when the shared Soviet-British espionage network in the neutral territory of Switzerland was pinned down by Gestapo technology and silenced in 1943. Would the Germans not assume that its enemy’s capabilities were as advanced as its own? Thus the XX Committee, abetted by the RSS, focused on such practices as reducing radiation emissions as far as possible without weakening the signal so much that it could not be picked up across the Channel.
  • Meanwhile, doubts lingered over the efficacy of the domestic interception operation. RSS was known to be very capable at locating, fairly broadly, transmitting stations in occupied Europe. It also gave great assistance to the MI5 in testing the strength of agents’ signals when the location of the transmitter was known. But how good – or committed – was it at detecting all other sources (such as the Communist transmitters that MI5 was nervously following)? If known operators could bypass RSS detection, what unknown agents were doing the same? This knowledge of undetected transmissions (some acknowledgeable, others not) increased suspicion of the efficacy of RSS processes.

This chapter starts to explore the evolution of this tangled operation. The official histories provide little guidance: there is no comprehensive account of the RSS organisation outside some mainly affectionate memoirs. Frank Birch’s multi-volume history of British Sigint is opaque, and often self-contradictory: it is overloaded with obscure organizational subtleties, and fails to make crisp conclusions. Some facts can be gleaned by a close inspection of the agents’ folders at the National Archives: occasionally a fascinating handwritten note of ‘Copy to Wireless Folder’ can be seen on documents, but no such Folder has been released. Several documents listed in the Indices of the agents’ folders have been plainly destroyed (some entries having a line through them with the descriptor ‘DEST’!). Yet enough tidbits of information can be gathered from the National Archives, including the unedited (but redacted) original of Guy Liddell’s very revealing Diaries, to indicate that the challenge of masking the D/F operation was taken very seriously by some intelligence officers. Strangely, however, many reckless decisions were made, too, that could have jeopardized the whole campaign.

Four Phases

I have divided the period in question into four segments. The divisions are in some respects arbitrary, but they do delineate some clear shifts in trends, and in the conduct of the war (up to the Normandy landings).

Phase 1: Learning the Ropes (September 1939 to the end of 1940):

This phase is characterised by a fear of invasion, and of supposed ‘Fifth Columnists’ assisting it. It is a period of organisational dysfunction, with no clear command over the personnel and technology required to intercept illicit transmissions, or the detection of strategic wireless communications from overseas. British Intelligence quickly learns, from its experiences with the suspected triple agent SNOW, and the rather undisciplined attempts by the Germans to land spies in the UK, the lessons deriving from analysis of radio traffic, and the role of detection-finding. The W Board and the XX Committee are set up as a structure to explore ways of handling double agents, but by then MI5 is losing control of an important asset with insights into the problem of interception.

Phase 2: Conflicts and Tensions (January 1941 to June 1942):

Organisational change brings improved management and leadership to MI5, but the placement of RSS under SIS (with which MI5 was initially happy) leaves the Security Service with a sense of lost control. SIS’s tightness over security means MI5 does not receive the decrypts it regards as essential to the task of running its double agents, and RSS’s mission shifts more to overseas work. Both MI5 and SIS try to deal with their prima donnas. Even the Controlling Officer of Deception faces political attacks. RSS is outwardly cooperative in direction-finding, but MI5 questions its ability and commitment to the Security Service’s aims. The situation regarding the Soviet Union is clarified by its entry to the war as an ally, but is then complicated by the wireless activity of Soviet spies.

Phase 3: From Defence to Offence (June 1942 to May 1943):

Progress is made: Masterman persuades the head of SIS to release decrypts to the XX Committee, and he confidently declares that the operation controls all German spies. The new Controlling Officer of Deception (Bevan) brings energy and imagination to the overall deception plans for OVERLORD: the strategy for double agents evolves to using them to mislead the enemy about the proposed landings in Europe. Knowledge of Abwehr communications has increased. With the arrival of GARBO, the XX Committee develops plans to expand the usage of wireless telegraphy among the agents it is controlling, as the method of communication will be faster, and more reliable.

Phase 4: High Stakes (May 1943 to June 1944):

As the pressures on the security of the Double-Cross operation increase, doubts surface. MI5 expresses anxiety about RSS’s abdicating monitoring of Army transmissions, a loophole the Abwehr seems to be aware of. Gambier-Parry, head of Section VII in SIS, is not fully trusted. Concerns intensify about the volume of traffic being sent, but that remain undetected by British surveillance. Concerns are expressed about rumours of suspicions within the Abwehr about the reliability of their spies in England. Interception of Abwehr messages, however, appears to confirm that the messages of the double agents are overall being trusted. Transmissions by third parties (embassies, Soviet spies and visitors) alarm MI5, which reflects on its technical lack of expertise. While periods of radio silence have to be imposed, the double agents (especially GARBO) continue to send long-winded messages, remaining on the air for hours. Yet the deception is successful.

This chapter covers Phase 1. The other Phases will be examined in future postings.

RSS & The Fifth Column Threat

For the first nine months of the war, MI5 was focussed almost entirely on the risk that a Fifth Column, taking its instructions from German radio broadcasts, might aid the Nazi invasion when it came. The Germans had set up several radio stations broadcasting in English, of which the New British Broadcasting Station was the most prominent. Apart from the obvious propaganda in its messages, MI5 believed that the signals included coded instructions that subversive refugees – and ardent British Union of Fascists – would decipher. The Security Service even believed that some privately-owned transmitters were sending information back to their masters, even though the use of unregistered transmitters was illegal. Prime Minister Chamberlain chartered his Minister Without Portfolio, Maurice Hankey, with investigating such leaks, and in November 1939 a new unit, MI8(c), was set up to take over responsibility for wireless interception from MI1(g). Guy Liddell, head of Counter-Espionage in MI5, did not think anyone was taking the matter of disguised radio codes seriously enough: no one – neither MI8, nor GC&CS, nor SIS, nor MI5 itself – was adequately equipped or motivated to assume the work.

The problem endured into Churchill’s administration, which took over the reins in May 1940. Ironically, Churchill was the most vehement about the Fifth Column threat, installing a new layer of management over the intelligence services, and firing the veteran head of MI5, Vernon Kell. An apparently valuable expert in telecommunications and cyphers, Lt.-Colonel Simpson, had recently been lost from the service. As a possible replacement, an officer from the BBC, Malcolm Frost, who started to work with Liddell, reinforced the claim that the NBBS was sending coded messages to subversives, and Frost probably saw an opportunity to enter the limelight by taking on the challenge of deciphering them. Yet, by the spring of 1940, the RSS (the Radio Security Service) had concluded that the lack of domestic wireless transmissions suggested that the threat from a Fifth Column was minimal. It came to these conclusions in an imaginative fashion, but the logic behind that judgment was to come to harm relationships between SIS (to whom RSS eventually reported), and MI5, responsible for domestic security.

The exact origin and identity of RSS are murky, some accounts suggesting that it was subsumed into MI8(c), and in fact became the bulk of that section, which officially reported to the War Office – a clash between militarism and amateurism that would lead to later tensions. (Please see SoniasRadioPart2 for a fuller account of the origins of this unit.) Some official archives suggest that RSS had been set up in 1938, and was a team composed largely of Voluntary Interceptors – amateur radio hams – who watched the ether for unusual signals, and a team of mobile direction-finding units nominally reporting to the General Post Office. That latter part is clearly true, but Frank Birch’s Official History of British Signals Intelligence indicates that, when the financial approval for an organisation called IWI (Interception of Illicit Wireless Communications) was granted in March 1939, that unit was soon named MI1(g), later reassigned MI8(c), and that the definitive organisation permanently known as RSS developed under MI8(c)’s control.  (Some archival documents tantalisingly refer to an entity titled the ‘Radio Section’, as if it had been a department of M11g, and then lent its name to the larger organisation.) Internal MI5 documents, such as Dick White’s Notes for Counter-Espionage Training in 1943 (KV 4-170), strongly assert that RSS was set up only at the outbreak of war. In any case, the section known as RSS had moved to Wormwood Scrubs, in the same building as MI5, in September 1939, and thus enjoyed close collaboration with MI5’s officers while keeping its separate identity. The official history informs us that the completion of the transfer from MI1(g) did not take place until November.

The critical conclusion that RSS made was based on the interception of transmissions from the German intermediary ship, the Theseus. The relevant sentences from Hinsley’s history are worth quoting in full: “The organisation responsible for the interception of illicit wireless transmission, the future RSS, continued to be controlled by the War Office – by MI 1(g) till November 1939 and by MI8 (c) after that date – with the GPO acting as its agent for the provision of men and material and the maintenance and operation of the intercept stations. By the outbreak of war its headquarters staff had been located close to GC and CS, which was to be responsible for cryptanalytic work on the intercepts, and it had finally established the beginnings of a network with the decision in March 1939 to establish three fixed and four mobile stations and the recruitment, from June 1939, of an auxiliary observer corps of amateur radio enthusiasts. But it had listened in vain for transmissions from the United Kingdom – in vain because it was still the case that no transmissions were being made apart from those on Snow’s set which was operating under MI5’s control. Since Snow’s signals had not been heard before MI5 took control of him, the failure to intercept others was understandably attributed to the inefficiency of the watch or to technical problems, notably the difficulty of picking up low-powered high frequency signals except at very close or very long range. By December 1939, however, it had been recognised that the difficulty did not apply to transmissions made from Germany to agents: they had to be able to receive their control stations’ signals, and if they could hear them, so could the RSS.”

What Hinsley does not explicitly state is that the task of deciphering these messages was undertaken by a Major E. W. B. (Walter) Gill of RSS, in conjunction with his aide, Hugh Trevor-Roper. Gill, who had served as a wireless intelligence officer in World War I, had been recruited by Colonel Worlledge only that same month, December 1939. Exactly why Gill had joined at this critical moment is unclear from the familiar accounts: John Bryden, in Fighting to Lose, describes Gill as primarily ‘a Canadian army signals officer’, and explains that a unit in Ottawa had picked up ‘the clandestine wireless traffic from Canada’, whereupon Gill had arrived at the War Office ‘looking for advice on how to handle it’. Bryden clearly states that the clandestine traffic was the Germany enemy-agent transmissions, and that Gill’s mission was to ensure that RSS abandoned its beacon-searches for the newer phenomenon. If Gill needed advice, why did the War Office not turn to Simpson, first, rather than planting Gill at MI8c with a directive role? Why could it not give those instructions to Worlledge directly? Who was making these critical decisions?

E. W. B. Gill

Bryden’s explanation does not really make sense. Indeed, it does not appear that Gill was shipped over from Canada. According to a biographical article by Dr. Brian Austin, Gill was aged 56, and employed as Bursar of Merton College when he volunteered for duty on the outbreak of war. Indeed, Gill had plenty of relevant experience for his role as head of the ‘discrimination section’ at RSS. In WW I, he had been instrumental in interpreting the wireless messages of the Zeppelins, and had also set up wireless intercept stations in Egypt. After demobilization, in July 1919, Gill was put in charge of the wireless intercept station at Devizes, where, as Austin notes, the attention of the listening devices including listening to allies as well. With an OBE awarded, Gill then returned to civilian life in the Electrical Laboratory at Oxford, and in 1934 published a short memoir of his life in the military titled War, Wireless and Wangles. Dr. Austin, who has performed intensive research into Gill’s life, reports that he was identified in a scheme of Lord Hankey’s as a potentially useful asset in signals intelligence (sigint), and assigned to RSS to work on discrimination, an aspect of traffic analysis that isolates signals of interest.  Given Hankey’s charter at that time, as described earlier, that makes excellent sense.

Trevor-Roper, who described Gill as ‘a genial philistine with very little respect for red tape, hierarchy, convention or tradition’, confirms for us that Gill was Bursar of Merton College when he invited Trevor-Roper to join him at the RSS, an observation that does not sit tidily with that of a sudden visit from Canada, and an order from the War Office. MI5 records that GC & CS turned down RSS’s requests for assuming the task of inspecting the messages, as it was too busy, and thus Gill and Trevor-Roper set about decrypting them themselves. By late January, 1940, Gill and Trevor-Roper had solved the cipher, and thus informed GC & CS of their achievement. That provoked Denniston’s ire. (Gill had performed a similar act in World War I, but the War Office had reacted positively to his breaking of the rules.) Perhaps as a punishment, Gill was then ordered – on loan –  to Oxford to set up a radar-training school, but, on returning to duty, was demoted and sent to the Siberia of Catterick. He must surely have upset someone with influence, and Hankey could not save him.

Yet, if RSS dabbled dangerously into GC & Cs’s domain of cryptography, it perhaps departed too rapidly from its own mission of interception and counter-espionage. It overlooked a very pertinent fact. Gill’s report, written in November 1940, states, on the basis that ‘it takes two to make a wireless communication’, that ‘if the agent can hear his replies, so can we, and the watch on the German agent stations is thus of first importance to see if they are working to any station we may not have heard’. The serious flaw in RSS’s logic, which I do not believe anyone has picked up, is that SNOW had been supplied with a transmitter only. Since any undetected agents would likewise probably have no receiver capability, there would not have been any messages sent out to them by their ‘control stations’, and thus absence of evidence of acknowledgment or guidance from Abwehr controllers was no solid indication that there were no other agents in possession of transmitters. Gill’s conclusions about the likelihood of undetected wireless agents in Britain may have been sound, but it was based on the assumption that these agents had receivers. If this was an acknowledged flaw in Gill’s reasoning, he could have been reprimanded, and the decision overturned. But it was not: his recommendations were adopted, and echoed by the official histories.

Thus Gill, with his disdain for the proper procedure, was ultimately responsible for a major strategic decision while gaining enemies on all fronts. At exactly the same time that Lt-Colonel Simpson was bolstering RSS and pressing for tight domestic surveillance, Gill turned its attention elsewhere. He incurred the annoyance of Denniston in GC & CS for stepping on its turf, and, with his boss, Worlledge, later touting his achievements in a case involving espionage in Morocco, he also trod on the sensitive toes of Major Cowgill in SIS. While the known technical difficulty of picking up medium-range signals could still have inhibited the detection of active agents infiltrated by the Germans, Gill persuaded his superiors that interception efforts should be focussed overseas. This new policy was articulated in the following account of the decision (at MI5’s 1943 training session of intelligence officers): “As far as stray agents in the U.K. were concerned it was held that rather than try to get on to their ground waves, they would watch the controls in Europe and would get the reflection of the existence of an enemy agent in the U.K.” Yet Dick White’s report includes a very misleading and surprising statement, relating to Gill’s discoveries of early 1940: “He [Gill] therefore obtained from M.I.5 (Captain T. A. Robertson) full information concerning double-cross W/T agents run by M.I.5, and directed the machinery of R.S.S. to a systematic study of first the control, then the other out-stations, of the enemy W/T system thus penetrated.” White is unambiguously referring to the time when the detected traffic was sent to GC& CS, and rejected, early in 1940. There was, however, no network of double-cross agents being run at that time. SNOW was the only candidate. What was White’s intention here in misrepresenting the facts, so soon after the event? Might have he wanted to inflate the breadth and depth of RSS’s capabilities, and to underline the correctness of its new mission?

Nevertheless, out of convenience, and because of the difficulties in picking up short-wave radio signals from close proximity, a policy was adopted of abandonment of any attempt to detect illicit wireless at source, replaced by a reliance purely on reflected signals.  Liddell hints at a tortured fear several times in his Diaries without every describing the explicit reasons for his sense of horror – namely, that he knew agents might have transmitters only, and that not all dangerous illicit transmissions were actually issuing from enemy (i.e. German) agents. Moreover, this concern echoed further, and was even represented by one of the historians (Curry) as a disaster of almost existential proportions.

Gill’s demise is astonishing. Here was an officer with an outstanding WWI record in wireless interception, awarded an OBE, bearing an impressive résumé of original scientific analysis in the inter-war years, sponsored by an influential minister, Lord Hankey, and recognised for some important analysis of German radio traffic. He was then dumped unceremoniously, not even being informed of his sacking, demoted from Major to Captain, and despatched to the Royal Signals Training Centre at Catterick. The obituaries written about him all point out his puckish humour, and his impatience with any cant or humbuggery. He must surely have spoken up in inappropriate terms about Denniston, or made other unpublished criticisms, to incur such treatment, but Denniston himself was under the gun, disliked by the Armed Forces staff, and shortly to be demoted himself. It is a mystery that suggests there was more going on than has been recorded. Was Gill really such an unpopular performer in the eyes of the Top Brass?

Such tensions between cryptography and interception had been highlighted by ongoing disagreements between GC & CS and the intelligence units of the Armed Forces, who were all investing more money and personnel into sigint, but who were resenting the amount of control that GC & CS wielded over the committees that made decisions about interception. The Y Committee, which was responsible for wireless interception policy, had held a meeting on December 28, 1939 (chaired by Denniston), that did not succeed in reconciling the disparate views expressed, representable mainly as the conflict between Service independence and inter-Service centralised control. In familiar tradition, the Minister Without Portfolio, Lord Hankey was asked to arbitrate. Hankey was a committee man, and his recommendation of strengthening the Y Committee, under a new chairman from the Admiralty, and joint secretaries nominated by the War Office and the Air Ministry, was adopted on March 1, 1940. In May, this new committee officially recognised RSS’s vital role in exploring these overseas groups before handing them over for attention by the Service analysis stations.

Double Agents

Meanwhile, MI5 had been exposed to its first experiences with double agents. (The primary reference for the double-cross operation is John Masterman’s The Double-Cross System, but, while giving a first-class breakdown of the mechanisms and principles of the operation, it is as much a work of public relations as it is formal history. Ben Macintyre’s Double Cross is engagingly written, and an excellent guide, but contains many mistakes.) This period was dominated by the case of agent SNOW, a Welshman named Arthur Owens. Owens, who was a businessman specializing in batteries, had been an occasional agent for SIS, but was discovered by MI5 to have been in contact with the Abwehr on business visits to Germany. He had been given a wireless transmitter by his Abwehr controllers, and started signalling in early September 1939. He was by then, however, under MI5 supervision, and his messages were initially sent from Wandsworth Prison. (A lively account of Owens’s career as a double, and possible triple-agent, can be found in James Hayward’s Double Agent Snow.) What is important for the story of detection and deception is what MI5 learned early in the cycle, before the mass of would-be spies arrived in the autumn of 1940, with the result that the Security Service was prepared when the tide arrived.  It was at this stage that many of the formative ideas about deception, and what was required to make it successful, were forged.

Agent SNOW (Arthur Owens)

SNOW’s exchanges with the Abwehr also provoked some highly important breakthroughs. This particular aspect of how the SNOW experience assisted cryptology generally has been told concisely and comprehensively several times (for example in Nigel West’s MI5), so I shall simply summarise it here, and add some commentary. The knowledge of the codes that SNOW used in his communications facilitated for Gill and Trevor-Roper, and then Oliver Strachey and Dillwyn Knox in GC&CS, the task of deciphering Abwehr messages. Some of these were based on use of the Enigma machine, but communications with agents in the field, and outlying bases that would not have been secure enough to be entrusted with Enigma machines, used hand cyphers (such as pinwheels with codes).

Early in 1940, RSS’s team of Voluntary Interceptors had been able to ‘pinpoint’ [a term that Nigel West provocatively uses] a vessel, the Theseus, lying in the North Sea as the originator of the transmissions received by SNOW, and the source of messages to other agents in Western Europe. It is, however, extremely unlikely that location finding was accurate enough at that time to give precise co-ordinates of any transmitter without local sniffers being required. It is not clear from the accounts whether a broad area was identified, and the precise location of the German vessel established by aircraft inspection, or whether a purely electronic identification of the location of the Theseus had been made. ‘Pinpointing’ is a regrettable term. Indeed, Frank Birch offers the following laconic observation about the state-of-the-art at this stage of the war: “The optimism of enthusiasts as to the pinpoint [sic] accuracy of D/F fixes was shattered early in 1940 by the Norwegian campaign.” Nevertheless, through this successful detection exercise, RSS was able to supply GC&CS with a constant stream of traffic to the cryptanalysts in Bletchley Park.

Yet the questioning of SNOW in early 1939, when he had informed his contacts in MI5 of the immediate plans of the Abwehr to deliver to him a wireless-set, are also very revealing, in that they show both the mixed ambitions of the Abwehr as well as the ignorance of MI5 about wireless matters. The set itself was delivered to a left-luggage locker at Victoria Station, and MI5 arranged for the equipment to be removed and inspected by SIS before allowing SNOW to explain its workings, and hand over its codes and callsigns that he was supposed to use. The device was small, and portable, and was claimed to have a range of 12,000 miles, using an ordinary 350-volt battery, and also to be activatable by plugging into a normal lamp-socket. Yet it was a transmitter only: SNOW was to inform his masters when transmissions would start by means of the regular mail service, and, in time, acquire a short-wave set that would allow reception. This is quite an extraordinary revelation, showing how unambitious the Abwehr was in its wireless plans at this time. A transmission without any mechanism for immediate confirmation was a highly quixotic venture, and the Abwehr’s relying on its agent to construct a receiver (a more complicated piece of apparatus than a transmitter) and manipulate it properly betrays an overall lack of seriousness that again belies Howard’s confident assertions about Abwehr strategy.

An earlier interrogation of SNOW had been carried out, in September 1938, by Edward Hinchley-Cooke, an enigmatic figure in the whole saga. Hinchley-Cooke is a puzzle, primarily because the authorized historian of MI5, Christopher Andrew, gives him no coverage at all after the early 1920s. He features regularly, up until 1943, as an interrogator of Germans in Liddell’s Diaries, but Nigel West (who also edited the published version of the Diaries) never places him in any of his organisation charts in his own history. Hinchley-Cooke had a German mother, and spoke German very fluently, which is probably the reason that he was brought into so many of the interrogations and prosecutions of Nazi agents. John Curry, in his history of MI5, suggests that Hinchley-Cooke was ‘attached to’ B Division in 1939, while working for the War Office, because of his interrogatory skills, but then clearly states that he was on the Director-General’s Staff after Petrie’s reorganization of summer 1941. John Bryden indicates that Hinchley-Cooke was the sole MI5 officer working on German counter-espionage up to the outbreak of the war. Moreover, Hinchley-Cooke’s questioning of SNOW was not very subtle. He failed to follow up on SNOW’s evasive answers, and it is clear that Hinchley-Cooke had no understanding of the principles of radio communication and codes. He was accompanied by an Inspector and Superintendent from Special Branch, but their names are redacted, and they contributed little to the proceedings. This lack of technical expertise would come to dog MI5 in a big way.

The Strange Career of Lt.-Colonel Simpson

Yet MI5 did possess competency – for a while. Even more astonishing than the oversight with Hinchley-Cooke is the failure of the authorised historian to include any reference to a key figure behind the events of 1939, one Lt.-Colonel Adrian Simpson. Perhaps Andrew’s omission (quite probably a matter of strong guidance to the authorised historian by MI5’s mandarins) is due to the fact that Simpson appears to have been appallingly mishandled. We owe it to Curry’s ‘official’ history, published for internal use in 1946, to describe for us how Simpson was appointed to advise MI5 on all matters relating to wireless after the Security Service had declined to take on the responsibility for establishing the Radio Security Service in late 1938. Simpson was well qualified, having been head of MI1(b), the code- and cipher-breaking agency in WWI, and an executive with the Marconi company between the wars. Nigel West’s Dictionary of  Signals Intelligence informs us that in 1915 ‘Simpson established a General Headquarters cipher bureau at Le Touquet to analyse material collected from intercepted enemy landline communications’, and that ‘within a year, MI1(b) had built direction-finding stations at Leiston in Suffolk and Devizes in Wiltshire, with a control facility on the roof of the War Office in London’. MI1(b) was a core group that was amalgamated into the Government Code and Cypher School (GCCS) in 1919. So Simpson was eminently qualified to define the next generation of interception facilities. And it should be noted that Walter Gill had been the head of the Devizes station, possibly appointed by Simpson: one might expect him and Simpson to have been collaborators, even friends.

Simpson’s efforts appear, however, to have been wasted. Curry would go on to write: “One of the conspicuous illustrations of these tendencies has been the refusal in December 1938 to grapple with the problem of wireless and the consequent establishment of R.S.S. under M.I.8 with results recalling the principles of Greek tragedy.” This extraordinary uncensored commentary on ‘Greek tragedy’ must hint at disasters undocumented. If the war was won, and the Double Cross operation judged to be an utter success, where did the calamities lie? Which character would suffer in Act V? Would it be Liddell’s failure to be appointed Director-General of MI5 in 1953? Where were the bodies buried? Why did MI5 allow this judgment from Curry to appear?

Curry states that B.3 (which Simpson headed, a section under chief of counter-espionage Guy Liddell) was not set up until the beginning of the war, but Simpson was clearly active in some influential capacity throughout 1939. He wrote (at least) three important papers, none of which appear to have survived. In an October 1938 report that surely provoked the December decision, he had crisply laid out the investments, equipment, and organisation that an effective Security Service would require to defend the realm against illicit wireless, pointing out that technology had advanced considerably in the past few years. This scenario would include three fixed Direction-Finding stations, and a corps of several dozen Voluntary Interceptors to track the airwaves. Hinsley and Simkins reinforce the importance of Simpson’s recommendations, writing that his report ‘reached the disturbing conclusion that interception arrangements were so inadequate that had recent developments led to the outbreak of hostilities a skilled agent could have established a reliable wireless service and maintained it for a considerable period with almost complete immunity; he added that such a service might well be already in existence.’

Simpson took over responsibility for B.3, a section that was set up to liaise with the RSS, and to deal with suspected illicit transmissions, in person being involved with any search and prosecution decisions. He was clearly closely involved with the SNOW case during 1939, but was moved to write another report, dated February 2, 1940, which harshly criticized ‘the state of affairs concerning the detection of illicit wireless’, although he laid most of the blame at the General Post Office for its failure to provide the appropriate skilled staff in operating the sniffer-vans that would hunt down transmitters to individual residences. His career with MI5 effectively ended at that point, as he was reportedly moved over to General Wavell’s army in the Middle East: whether he was pushed out, or resigned in frustration, is not clear. The source of this story may be Stephen Dorrill, who writes in his 2000 history of MI6 (SIS) that Simpson was appointed by Wavell to prepare to counter possible Soviet intervention in Transcaucasia. Since Dorrill then states, however, that Simpson, ‘a former managing director of Marconi’ [correct], ‘had been ADC to the Grand Duke Nicholas in the Russian Army’s Caucasus “Savage Division”’, Dorrill may have got the wrong Simpson. That experience does not sound as if it comes from the ‘Memoirs of a Wireless Interception Man’. In any case, Curry’s observation that MI5 ‘lost his services’ at that time suggests that he resigned. An intriguing correspondence that Mark Rowe, author of Don’t Panic: Britain Prepares for Invasion, 1940, discovered in Bristol record offices, indicates that, in April 1940, Simpson was still recruiting Voluntary Interceptors to the RSS organisation. Maybe he did not move to the Middle East, but worked for a while championing what he saw as RSS’s true role, and applying pressure to his successor, Malcom Frost (see below).

Curry’s suggestion that Simpson stated that the fault lay with the staff operating the sniffer-vans may have been a political comment that veiled the truth. If sniffer-vans were going to be effective in following up triangulations of illicit transmissions, they would have to work in real-time in close communication with the Y Service that tracked signals. Sending them out the next day to try to detect noise would be a fruitless task unless the service expected the offenders to transmit at the same time that day. Moreover, the sight of such vans would immediately have deterred further transmissions, as we learned from the activities of the communist Green network (see SoniasRadioPart9). The Gestapo would soon perfect such an operation, with radio contact between vans and central control (which I shall describe in a later chapter), but one can easily imagine a more casual approach in island Britain at this time. Simpson’s criticisms, and imminent departure, hint at such more serious problems. Perhaps he had identified the inevitable conflict between efficient location-finding and controlled double agents using wireless, and his name has thus to be excised from the record, like one of Stalin’s commissars disappearing from a photograph?

What is even more astonishing is Guy Liddell’s almost complete exclusion of any reference to Simpson in his Diaries. The complete (but redacted) version of the diaries at the National Archives contains just one reference to Simpson by name (when he is called to investigate Verey lights at Harwich Harbour), and one veiled reference to his positional identity (B.3) when, on March 20, 1940, shortly before he resigned, Simpson attended a meeting with Liddell, Worlledge of M.I.I.8, and G.C.& C.S. officers and ‘cypher experts’, to discuss decrypted messages from Germany. Yet the organisation of B3 is very puzzling. If Simpson headed it (as Curry clearly states), T. A. (‘Tar’) Robertson must have been his subordinate, yet Robertson signs off his reports as ‘B3’ in the autumn of 1939, while a couple of anonymous memoranda, signed off as ‘B3.a’ while Robertson was away, may have been written by Simpson. Robertson worked closely with Simpson on the SNOW case: Robertson refers to Simpson’s attending a meeting at Robertson’s house without clarifying the management relationship.

Yet there may have been problems with authority and rank. Simpson was a Lt.-Colonel with a proper military background, while Robertson was only a Captain at this time (soon promoted to Major after Simpson left). In the rank-obsessed climate of wartime Britain, that would have been a problem if Simpson had truly been subordinate to Robertson. Curry muddies the waters even more, since elsewhere he writes that a subsection B.3.B was responsible for liaison with RSS. That is how the structure appears in his diagram of the organisation after the Petrie decisions in July 1941: I have found no specific reference to B.3.B in the time that Simpson was around. Maybe with some purposeful vagueness, without giving a precise date, Curry writes: “It [B.3.B] derived from the section under Captain later Lt.-Colonel Robertson and Lt. Colonel Simpson which, before and soon after the beginning of the war, was concerned with the arrangements for developing the R.S.S. organisation and for maintaining liaison with it . . .” If anything, it points to an awkward compromise joint leadership, akin to the role that Liddell was sharing with Lord Swinton’s pal Crocker at the time. William Crocker, a solicitor, was another disastrous imposition forced upon MI5, this time by Sir Joseph Ball, who was responsible for handling the Fifth Column ‘menace’ on Swinton’s Security Executive.

Liddell frequently talked to Robertson about the SNOW affair, but ignored – or bypassed  –  the expert brought in to design the RSS architecture, and makes no mention of his career, or the reasons for his leaving, even though what occupied Simpson’s time (the laxity over tracking down illicit wireless) was a subject that worried Liddell just as much. Robertson himself is recorded as speaking to Liddell in a fashion that passed on Simpson’s opinions (such as the criticism of the sniffer vans), and it appears that Robertson was content working under/with Simpson (unlike his relationship with Simpson’s eventual successor, Malcolm Frost). Thus Liddell’s studied rejection of Simpson’s significance is even more surprising. Did he perhaps resent an officer being foisted upon him? Did Simpson argue and activate too strongly for taking on RSS within the B3 section? It all points to a mysterious clash of personalities, or a disagreement over policy, not just a later embarrassment that might have required his name to be redacted. One must also wonder whether Gill and Simpson had crossed swords at some time. Gill, as I pointed out earlier, had been head of the interception station at Devizes, which was one of the monitoring posts established by Simpson. The highly oppositional strategies of a) RSS being consumed by foreign broadcasts, and being passed to SIS (Gill), and b) MI5 securing its control over illicit transmissions in Britain by taking over RSS (Simpson), would have clashed mightily. Is it possible that Gill was inserted into RSS to ensure that the unit did not fall into the hands of MI5? Moreover, the neglect by the authorised historian, Christopher Andrew, to write anything about B3 section must count as either a colossal oversight or an act of censorship – especially since Andrew recognised Simpson’s intellectual contribution in his earlier (1999) Introduction to the publication of Curry’s History.

SNOW’s Radio Activity

To return to SNOW. The coverage of SNOW’s radio activity after MI5 took control is infuriatingly elusive in the books that write about him, from Nigel West’s rather choppy MI5 (1981), through Volume 4 of The Official History of British Intelligence in the Second World War, by Hinsley and Simkins (1990) and Christopher Andrew’s authorized history of MI5 Defend the Realm (2009), to James Hayward’s breezy Double Agent Snow (2013) and John Bryden’s Fighting to Lose (2014). The archives on SNOW are typically disorganised, with much repetition, as well as many undated and anonymous reports, and it is consequently very difficult to identify exactly what wireless equipment is being referred to in the various documents.

The narrative on his wireless activity appears to run as follows: As outlined earlier, the Abwehr originally, in January 1939, provided SNOW with a transmitter only, suggesting that he himself construct a receiver. SNOW had been apprenticed as an electrical engineer, and was an expert on batteries, but constructing a reliable transmitter was no simple task. In the interim, it would mean that confirmation of receipt, transmission times, etc. would have to be conducted by letter, through SNOW’s purported business contacts in Germany – an extraordinary convoluted process, but one which was acceptable during peace-time. The Abwehr apparently had plans to send SNOW to the Americas at one stage: hence the extraordinary wide radius the transmitter enjoyed. The set was flexible and portable. It could be tuned to different wavelengths, unlike later models used, which required individual crystals. But it was unreliable, burning up under SIS/MI1(g) tests, and the boffins had to restore a resistance unit so that it would do the same again when SNOW tried to use it. In fact, SNOW was still having problems with it in July 1939, when he wrote to his contact Auerbach saying that he had at last rectified the faulty resistance. And transmitting successfully over 12,000 miles, had SNOW been able to smuggle his set overseas, would have required a very large antenna.

SNOW’s career was then disrupted by family matters: a jealous wife reported him to the authorities, telling them that he had disposed of his wireless set. MI5 tracked SNOW down to Surbiton, whither he had moved with his mistress on August 29, and, with his guidance, Robertson and his colleagues discovered a receiver in the bathroom cupboard, and his original transmitter buried in the garden. (Hayward notes that the receiver was a ‘crude’ device, and that SNOW had ‘apparently’ constructed it himself: maybe the experts from the Royal Signals had actually delivered it for him.) When war broke out, SNOW was arrested, and MI5 started broadcasting on his behalf, officially using him as a double agent. After the initial broadcast from Wandsworth Prison, the officers feared that the Germans might be able to triangulate the origin of the signals, and then ask themselves how a clandestine transmitter could be allowed to operate from such an institution. In fact they were being unduly cautious: locations could be identified only to the level of a large conurbation, and (certainly at this stage of the war) it would have taken a platoon of sniffer-vans, supported perhaps by portable equipment, to narrow the search to a particular building.  Moreover, German goniometric techniques were inhibited by geography: it took at least three receiving stations to plot an accurate fix, and their dominant Eastern orientation meant it was more difficult for them to triangulate transmissions from the UK. The British authorities would nevertheless have been mindful of the successful, but highly complex, process that allowed them to home in on the illicit Soviet MASK transmitter in Wimbledon a few years before.

Hereafter the story becomes contradictory. SNOW did make contact with his Abwehr controllers on September 19, but, given the problems he was experiencing with his apparatus, Hamburg promised to send him a new transmitter. MI5 reported how unreliable the current transmitter was. On February 29, Liddell noted that SNOW’s set had blown up, and a telegram had had to be concocted to send to the Abwehr to indicate that he had not been raided. His apparatus required a 98 foot antenna, which did not work reliably if misaligned. (The device had a knob – a ‘tuner’ – to control frequencies, but required corresponding changes to the antenna length if a frequency was switched. Using a knob would have been less reliable as a way of selecting a frequency than the insertion of a fixed frequency crystal.) Signals were not strong: Hamburg said they were weaker than those coming from Ireland. MI1(c) had been monitoring SNOW’s transmission: they said that jamming by a powerful station was causing interference. A note of February 29, 1940 indicates that the intrusion of dampness caused the equipment to burn up, with advice to use an outside antenna to avoid the use of the relay circuit.

A typical British suitcase wireless transmitter/receiver of early WWII

Yet, in another Case History, undated, but probably written in April 1940, as it refers to events that month as in the recent past, and describes how ‘every two or three months’ SNOW travels to Antwerp to meet Dr. Rantzau (whose real name was Ritter) – a record which must have preceded the Nazi invasion of the Low Countries. Here the writer tells us that SNOW ‘broadcasts every evening’. At some stage, SNOW’s set must have been improved after the stumblings earlier in the year: the archive notes that seamen couriers (quaintly described as ‘lascars’) did bring over new parts in April 1940, but the arrangement of having a separate receiver and transmitter was clumsy, and maybe the range of the machine made it more liable to direction-finding. Back in March 1939, MI5’s B.3 (i.e. Lt.-Colonel Simpson) had sought the opinion of Colonel Yule of MI1(g) as to how long he thought it should take for ‘our internal intercept and D/F organisation’ to locate SNOW’s transmitter, clearly concerned about what the Germans were thinking. Yule had organized some rather casual efforts to track SNOW’s frequency, and even mentioned detector vans, but the initiative appeared to fizzle.

Despite his studied ignoring of Simpson, Guy Liddell himself showed remarkable foresightedness in understanding the sensitivity of this issue, and the value of downplaying the radio-detection capabilities of the British security organs. In a diary entry for October 28, 1939, he wrote: “Brigadier Martin of MI.1 has suggested that a representative of the News Chronicle who thinks he had detected an illicit wireless station, should be shown the apparatus we use and taken round in a van in order to get a cross-bearing. He would then write up the story in the Press. D.S.S. telephoned Martin to say that we had strong objections to any publicity being given to this matter. It was in our interests that the Germans should regard us as grossly inefficient in these matters, particularly as ‘Snow’ is sending them weather reports. If they thought our organisation was that good they might well ask how it was that he managed to get his messages through.” This episode shows how quickly Liddell summed up the value of subterfuge against the obvious appeal of propaganda, at a time when the British press was very keen on providing the public with ammunition against the Fifth Column threat.

Nikolaus Ritter, Chief of Abwehr Air Intelligence

Direction-Finding

The British were not the only group to be thinking about wireless detection. When SNOW visited Rantzau in Antwerp in early April, 1940, prepared by MI5 to probe the enemy’s thoughts on detection-finding, Rantzau told him that he should not be concerned about being detected, as ‘as it was a very difficult thing to track down short wave wireless sets’. This information – that shortwave sets were immune to detection and direction-finding – was one he had originally given to SNOW as early as January 1939, a revelation that SNOW had passed on to a sceptical Robertson. Now, in April, 1940, Rantzau even mentioned the Abwehr’s strenuous efforts to track down such sets closer to hand. The details are redacted, but these were probably sets managed by the Soviet Red Orchestra. Rantzau told SNOW that a transmitter had been detected in the Wilhelmshaven area, but it had been impossible to run it down. In the light of later experience with this communist network, and with SOE wireless operators inserted into Nazi-controlled territory, primarily in France, this rather sanguine opinion would need to be changed.

“This is nonsense”, declares Bryden, perhaps too brusquely, implying that Rantzau was being devious, and in his book he gives an oversimplified account of how triangulation worked. In the early part of 1940, techniques were surely not that advanced. I quote Bryden’s summary in full: “Obviously, in order to survive in enemy territory, it is helpful for a spy to change frequencies and call signs as often as practical., but the most important necessity is to send from different locations. DR. RANTZAU was not asked the most critical question: Was it safe for JOHNNY – the name Ritter preferred to use for Owens – to always be sending from the same place? The Germans were soon to provide the answer when Britain’s sabotage agency, Special Operations Executive, began landing its agents into occupied Europe. Their wireless transmissions were DF’d and they were caught by the score. The only MI5 officer with the technical clout to challenge DR. RANTZAU’s advice – Colonel Simpson – had left. In his absence, Robertson chose to believe his German opponent.”

What is extraordinary is that, the very same month (April 1940), SNOW’s transmissions had been picked up by the French ‘illicit wireless service’, as Liddell reported. The French were, of course, conveniently at a distance where clear signals could be picked up. Fortunately, the French had sent a report to GC&CS, whence Commander Denniston forwarded it to Gill in RSS, who contacted MI5. “We are telling them to lay off”, wrote Liddell. Robertson sent a letter to Major Cowgill of SIS, telling them that MI5 knew all about the station. But, if a French service had been able to pick up SNOW’s signals, and to determine that it was probably an illicit set operating from the United Kingdom, why did MI5 not imagine that the Germans would conclude that the British should have been able to do the same, making allowances for the dispersion of their interception stations? (This is a vital point that Bryden makes, although he does not discuss the subject of ‘dead’ zones.) And was it not a careless mistake to brush off the interest of the French so casually? It could have been a leaky organisation, and the rumour that the British were manipulating a German agent could have spread.

Despite the provocative but fortuitous French experience, the problems in performing accurate direction-finding of short-wave radio signals were officially well recognized at the time. Frank Birch, in his Official History of British Sigint, 1914-1945, wrote: “Intertwined with the problem of interception was that of D/F, greatly complicated since 1918 by the development of shortwave transmissions and the general awareness among signals personnel of the need to defeat, as far as possible, D/F operation”, rather cryptically hinting at defensive methods that British signals would need to employ against German capabilities. Surprisingly, Birch did not explain why short-wave transmissions were less easy to detect: it was because their signals were bounced off the ionosphere, which gave them a greater range, but made their isolation more difficult. M.R.D. Foot, in his book on SOE, informs us that this phenomenon is called ‘skip’:  the signals bounce between the ionosphere and earth, and create ‘zones of silence and zones of good reception that may alternate all around the globe – and may vary according to time of day, season of the year, or prevalence of sunspots.’ (As will be shown below, Simpson offered a similar explanation of this phenomenon.) These signals were also subject to interference from local electronic activity, but, if that were too intense, it would have affected reception on behalf of the intended audience as well. Birch also pointed out that, while the Germans were able to specify their frequencies to the level of one kilocycle, the British were precise only to five, and ‘and in practice measurements were often up to a hundred kilocycles out’. Birch went on to write: “Now, its frequency was an attribute of a signal on which both traffic analysts and cryptanalysts partly based their work, and by February 1940 the nuisance of inaccuracy had become generally recognised as acute. But the remedy was still far to seek.”

Earlier in this piece, I quoted the MI5 report about giving up on trying to ‘get on to their ground waves’. The way that short-wave transmissions worked, ground waves would be emitted from the antenna, as a source of radiation, and could be picked up until the curvature of the earth (or unusual geological formations) attenuated them completely. In order to reach a remote target, the antenna would also emit skywaves, which would use the ionosphere to ‘bounce’ those signals beyond the horizon. But there would be a dead zone between the area where the ground wave penetrated and the larger expanse where the signals could be picked up – both by the desired station, as well as by interception stations with roughly the same distance and sphere of receptivity. (see diagram) These areas are technically called ‘lobes’, and their dimensions are dependent upon whether the antenna is placed horizontally or vertically. And that is why the detection and location of illicit radio were problematical. Interception stations within the skip zone would receive nothing if they were also beyond the ground wave range. And it would require at least two stations in the right place to attempt a fix, while the distortions of the skip zone could confuse the analysts.

Ground waves, sky waves and the ‘Skip Zone’

Birch was even more outspoken elsewhere, and it is worthwhile quoting in full an important paragraph: “On its own, D/F has been described [here he cites a Naval source] as ‘by far the most important source of communications intelligence, if the cryptographers are out of form.’ This is no doubt true, but truer still is another authoritative statement that ‘independently of Special Intelligence, D/F was useful, but in a much narrower field’. Optimistic illusions as to its accuracy were shattered early in the war. ‘It was the exception rather that the rule for a fix to be obtained that could be classified as being within ‘forty-mile radius’, and there were many occasions on which even such fixes turned out to be very wrong indeed’. In spite of the multiplication of stations carefully sited so that as many as 30 or more bearings could be taken of a simple transmission, in spite of improved equipment by technical staffs and the working out of a mathematical method of calculating the position of a unit, based on the ‘weight’ or class of bearings, investigations by experts revealed that ‘a good operator was nine-tenths of the battle.’ In plotting, skill and experience mattered more than gadgets and quantity of bearings. In short, D/F ‘as practised from 1939 to 1945 was an art – not a science.’” That judgment would appear to contradict directly the rather overawed conclusion that the American intelligence officer Norman Holmes Pearson offered in his Foreword to John Masterman’s book: “The techniques of intercepting messages sent by wireless were highly developed. So was the science of direction-finding by which the location of the transmitting instrument could be determined.” That is how mythologies begin.

Simpson had himself contributed to the debate. Again, we have to rely on Curry. Before he left B.3, Simpson wrote the third report that we know of: ‘Notes on the Detection of Illicit Wireless 1940’, with a view to investigating reports of suspected illicit wireless transmission. In Curry’s somewhat clumsy words: “He explained the problems connected with Ionosphere or Reflected Ray communication and ground rays, and suggested that secret agents would be able to avoid bulky or intricate apparatus and that only low-power would be employed. He said that, assuming an efficient receiving station in Germany, it would be possible to select a suitable wave-length, having regard to range and seasonal conditions, which would give a regular, reliable service. If such a station were to be established in a carefully chosen locality in this country it would very likely not be heard at all by our permanent interception and D/F stations. Such a station could be situated in the centre of a densely populated area or alternatively installed in a small car. He set out detailed instructions for procedure in dealing with investigations in these circumstances.” What those instructions were, we shall apparently never know. But his advice appears to have been ignored – or to have been politically unsuitable.

Simpson’s message was picked up by Hinsley and Simkins in Volume 4 of their History. “Since Snow’s signals had not been heard before MI5 took control of him, the failure to intercept others was understandably attributed to the inefficiency of the watch or to technical problems, notably the difficulty of picking up low-powered high frequency signals except at very close or very long range”, they wrote, to which Bryden has a riposte. In a note concerning the official claim, he comments: “This is correct, but it does not mean that the transmitters could not be located. The British Post Office was already using mobile direction-finding units to pick up local transmissions, and the Germans in Holland and France were to develop the technique to a fine art.” He goes on to say ‘the spy could change frequencies by changing the crystals in his set.’ For 1940, Bryden probably anticipates a little too much, and credits the Post Office vans with a little too much finesse. (It is one thing to roam around potentially busy locations, like Embassy districts, and another to chance upon illicit transmissions from private residencies around the countryside.) We now know that Simpson criticized the skills of the mobile direction-finding units (as did Liddell), but they may have been limited by technology. Furthermore, changing crystals had other implications, and was not available to every operator at this time. The commentaries on SNOW’s apparatus inform us that, in order to change a frequency, if a crystal were inserted (or the change made by the advanced facility of using a knob), the length of the antenna had to be changed as well. Matters were not as cut and dried as Bryden represents them: the state-of-the-art was probably more in the vein of what Rantzau and Simpson independently stated at about the same time.

The confusion is reinforced by other conversations. Liddell had discussed the question of detection with Menzies, the head of SIS, in April 1940. By then, Menzies claimed that it had wireless sets ‘operating from German territory and all over the continent’, a boast that would incidentally appear to be belied by Keith Jeffery’s authorised history. Thus Menzies’s statement was probably more goal than reality. Menzies echoed the general confidence when he told Liddell that he thought that SIS’s newest sets were ‘extremely difficult to pick up’, and he doubted very much ‘whether any monitoring system however widespread will be effective against them’. Where Menzies derived this science is not explained (he had no doubt been briefed by his head of Section VII, the telecommunications expert Richard Gambier-Parry, who was known for treating non-technicians with some arrogance) but it led Liddell to compare SNOW’s set, even with its new valves, very unfavourably against the SIS’s latest technology. Liddell concluded that diary’s entry with a not unreservedly confident belief that the British were ahead of the Germans in this matter. Both Rantzau and Menzies would later have to revise their opinions.

[This whole puzzle of direction-finding is tantalisingly highlighted by the titles of a set of lectures given by Herbert Hart, Major Morton Evans and Major Frost at the MI5 training session for regional officers on January 5th, 1943. They were, respectively, ‘Intercept Intelligence and its uses’, ‘The work of R.S.S. Interception and discrimination of Axis secret communications and its bearing in detection of illicit W/T’, and ‘Investigation into illicit W/T’. Unlike Dick White’s comprehensive notes, the details appear to have been destroyed.]

In any case, SNOW was provided with a new transmitter/receiver in August 1940, when the courier BISCUIT (Sam McCarthy) went to Lisbon, and was handed over a suit-case containing a new apparatus, known as an Afu. This is the equipment described in a later 1941 report: “His [SNOW’s] second set was a mains operated transmitter and receiver of excellent construction, the reception frequency being 5,800 kcs and the transmission frequency either on 6636 kc or on 6536 kc. The even frequency was used on odd days, the odd frequency on even days. The antenna used consisted of about 90 ft. in one leg and counter-poise 15 ft. long in the other. There was no difficulty in erecting this or its subsequent many locations and no difficulty experienced anywhere.” Hardly the discreet apparatus that could be easily concealed from the landlady, but it presumably gave much more reliable service until SNOW was closed down in 1941, and his wireless taken over for other agents. But there was a cost, too, since only two frequencies were offered, appearing to be delivered by separate crystals rather than a controlling knob: more reliable, but with fewer options.

Thus the Abwehr’s initial experiments with wireless were very tentative. It was as if they did not take the need for close two-way communication very seriously. They did not supply SNOW with reliable equipment, and accepted long periods of silence with equanimity. This was, of course, when western borders of Europe were still open for travel. The Afu set was ideal for smuggling overseas, concealed in a suitcase, but was probably not robust enough to survive a parachute jump. The pattern of the next wave of agents to arrive would reinforce this phenomenon, however. Spies were not planned to be long-term subversives: their terms of activity were expected to be short as they facilitated conquest. Hitler was expecting a successful invasion of the British Isles after he moved through Western Europe, and won the air-battle. Investing in a miniaturised, robust and flexible combined transmitter/receiver was not a priority. This lack of imagination about the potential of agents equipped with wireless would require the British to take the lead, and help put ideas in the minds of their adversaries.

The XX Committee

Meanwhile, MI5 was increasing its attention on the strategic challenges of handling double agents. The original idea had in fact come from the French, In June 1938, the intelligence office Paillole had visited MI5 to instruct them on such a policy and practice, and in October 1939 Dick White, Liddell’s right-hand man, had gone to France to get a refresher. On January 10, 1940, Liddell entered the following observation in his Diary: “With a view to supplying double-agents with information, commands have been asked to send in reports on local information or rumour which they, as ordinary members of the public, can pick up from observation or from gossip. Xxxxxx xxxx xxxx [redacted], this information will be supplied for transmission to the enemy. It is hoped that once confidence is established in this way it will be possible to mislead the enemy at a critical time. In addition it is felt that the reports will provide M.I. with some picture of the extent of leakage that is going on.” Yet MI5’s resources in this field were now scant. SNOW, the only agent with a wireless snow under double control, was acting suspiciously. Moreover, his wireless blew up on February 29, and a false telegram had to be concocted to show that he his operation had not been raided. A series of adventures would ensue, but he was finally cut off on April 13, 1941. MI5 officers by then considered that his Abwehr control, Rantzau (= Ritter), had wised up to what was happening.

The Fifth Column hysteria confused practically everybody about a native threat to abet the coming invasion. Throughout the year, Liddell would report regularly on illicit broadcasts detected, but they would inevitably end up as being harmless, normally foreign embassies trying to break the rules. In July 1940, Malcolm Frost, the BBC man, was appointed to head a new branch (W) to coordinate SIS and MI5 activities concerned with Radio Security, supported on a committee by representatives from the RAF, the Army, and the Royal Navy. It looked like an auspicious move, although at about the same time, Churchill rather unhelpfully told the British public that the Fifth Column menace ‘had always been exaggerated’, perhaps forgetting that he himself had been the prime cause of that hyperbolic reaction. On October 1, Liddell rather enigmatically reported that a W Committee had been set up, at the instigation of the Director of Military Intelligence, ‘to control false rumours and disseminate false information’.

Tensions arose almost immediately. It did not make sense for a man recently recruited from the BBC for his radio expertise, with no experience in counter-espionage fieldcraft, to be put in charge of the new group responsible for locating enemy agents. In the summer of 1940, Liddell had reported that he was impressed with Frost (‘strikes me as an extremely able and knowledgeable person’). In July, Liddell and Frost discussed the new group that Frost would lead ‘under the guidance of RSS [MI.8]’, with Robertson as his deputy. As soon as they heard about this, MI.8 applied fresh pressure on MI5 that the Security Service should absorb RSS completely (including all the civilian Voluntary Interceptors). MI5 was generally still very wary over taking responsibility for activities deriving from foreign soil. Soon, moreover, Frost was bridling over the demands from MI.8 for him to hire dozens of military people to carry out the RSS’s mission – a demand that had been ironically crafted by Lt.-Colonel Simpson. Frost declared that he wanted to hire his own people. Frost was appointed head of W Branch in August, but it was soon subsumed into Liddell’s organisation as a section (B3) – Simpson’s old organisation, where Robertson worked. The W Committee soon morphed into the W Board, and delegated its detailed work to the XX (Double-Cross) Committee, which started to meet in January 1941. This was a dysfunctional mess.

By then, B3 was in tatters. Frost had always wanted his own Branch, reporting at the same level as Liddell, and his ambitiousness, arrogance, and intriguing started to grate. Lord Swinton, the head of the Security Executive (who had encouraged Liddell to charge Frost with setting up the new group on radio security only a few months before), said that the bumptious Frost had to go: Tar Robertson found Frost impossible to work for. Frost had of necessity (as head of B3) been an integral part of the debates over the German agents in the preceding months, but he had no expertise in this area. In December 1940, Liddell reported that Robertson and his group of agent handlers had been moved over to B Branch, as the new B1a section, ‘back where they belonged’.  Yet the initial handling of captured agents was in fact carried out by a group named B8L: you will find no mention of that entity in Curry’s, West’s or Andrew’s book. [One should not trust official authorities on the structure of Liddell’s ‘B’ branch: memoranda in the TATE file show Robertson, after being identified as ‘W’ in November 1940, reporting as ‘B2a’ long into 1941, which would have put him under Maxwell Knight’s ‘Agents’ section. Andrew’s implication that B1a had been run by Robertson since early 1940 (p 249) is patently false. McIntyre makes the same blatant mistake (p 38).] It took the early 1941 arrival of David Petrie, as the new MI5 chief, to bring some permanent structure to the service. Yet Curry’s organisation chart for July 1941 still shows the unpopular survivor Frost in charge of B.3 (Communications), including B.3.B (Illicit Wireless Investigations; R.S.S. Liaison). Robertson is only then under Dick White in charge of B.1.a (Special Agents).

In the middle of November, the W Committee had set up objectives for the planned Double-Cross System, and, at exactly the time that Malcolm Frost was falling into greater disfavour, the Oxford don John Masterman (who had been tutor to Dick White) was interviewed to help the project. He became the highly successful chairman of the XX Committee, which had its first meeting on January 2, 1941, and was to convene weekly throughout the war, the last meeting being on May 10, 1945. The W Board (the new name for the W Committee, to which the XX Committee reported) was responsible for setting overall policy, but it left the details of managing double agents to Masterman and the team of B.1a, later led by a happier Tar Robertson.

Sir John Masterman, Chairman of the XX Committee

By then a new tactical thrust from the Germans had taken place. The Battle of Britain had started on July 10, 1940, and, in anticipation of a swift victory, Hitler ordered that spies be inserted into Britain to inform the invasion force of weather conditions, troop movements, the condition of aerodromes, etc. Between the beginning of September and the end of November, about a couple of dozen agents arrived on British shores.  MI5 had used SNOW, however, to pass on examples of identification numbers for ration-cards to the Abwehr, so that the British were able to detect forgeries supplied to captured agents during this later swarm. This gave MI5 an inkling of the possibilities of deception, with some important feedback, and constituted one of the most important coups of the campaign, as it verified the trust that the Abwehr held in SNOW. Owing to SNOW’s information, predominantly the supply of false identity card numbers, and the decryption of the Abwehr hand cyphers, the British were ready for the infiltrators. A few arrived by sea, the majority by air. Most were arrested within hours. Several were executed: a couple committed suicide. And some were turned into double agents.

One remarkable aspect of the project was that, while most of the spies were equipped with wireless apparatus, it normally consisted of a transmitter only. That decision had been made primarily out of optimistic pragmatism: the Nazis expected the invasion force to arrive shortly afterwards, and there was no need for the spies to receive additional instructions after they had supplied the information they had been instructed to gather. But it was not an exclusive policy, nor one governed by the logistics of carrying a set by boat, or landing with a heavy unit by parachute. (Agent TATE was told that the type of set they were to use in England depended on whether they crossed by sea or by air, as there was not yet any shock-proof apparatus available. But that was clearly not the whole truth.) The first wave that arrived on the Kent coast on September 3rd all had transmitters only. Gosta Caroli (SUMMER), who was parachuted in three days later, brought a combined transmitter/receiver with him, strapped to his chest – but it knocked him unconscious when he landed. Amazingly, he had never experienced a practice jump. TATE (Wulf Schmidt), underwent a bad landing from low altitude on September 20, owing to poor weather, and he hurt his hand and ankle because of the weight of the equipment strapped on to him. Dr. Ritter (aka Rantzau) had informed TATE that arrangements were being made for him to take with him to England a separate transmitter and receiver and also a large transmitter (called a ‘Z.B.V.’) which would be dropped separately and which he could destroy if the smaller sets were unbroken after landing. How exactly a probably injured airman was supposed to grope around the countryside looking for another parachute with transmitter/receiver, and then conceal them, before looking for a safe haven, without provoking interest from the local population, is not addressed. The spies were expendable.

Yet MI5 had to move fast. The spies were expected to report quickly (in about a week) to their masters, and the Security Service did not want any uncontrolled messages going back to Hamburg. SUMMER had forewarned them about TATE’s arrival, but a tight procedure of isolating the agents from each other was required. If they could not be quickly turned, and with confidence, they would have to be detained, isolated, and probably executed. Having been interrogated and threatened at Camp 020 on Ham Common over a period of thirteen days, TATE was turned, and eventually made contact with the Germans on October 16, sending a message under the control of B1.a of MI5 that he was safe but had been hurt on landing. A similar excuse had to be concocted for SUMMER (who was by no means a Nazi like TATE, and more easily turned). SUMMER was viewed by the Abwehr as an agent in SNOW’s network, and the authorities in the UK had dithered. SUMMER was installed at a safe house in Hinxton Grange, where his operator sent a message on his behalf on September 27, claiming that Caroli had recovered from his injury and was now lodging near Cambridge.

SUMMER had provided his interrogators with vital information. After gaining consciousness, he realised he had missed the first agreed time for transmitting (2-4 am), but he was arrested before the second (6-8 pm) arrived. The notes in his Kew archive declare that he agreed to give all information and details of his code on condition that his colleague coming over soon (TATE) should not be shot. He had known the colleague only since July; he would also have receiver and transmitter. Details of contacts, the type of cipher used, and frequencies, were all to be ‘found on file’. Here can be seen callsigns, the cipher wheel used for encryption, the choices of frequencies (4000 or 6000 kcs). TATE would use the same kind of cipher, he said, but with a different key.  He was given no instructions on what to do when his batteries ran down – again a probable sign that the Abwehr thought that the invasion was imminent. Poor SUMMER had been told that if he was caught with his set he would be shot. If he was caught without his set, he was told to tell a tale. Early the following year, SUMMER would try to escape and attempt suicide. But that is for the next episode.

John Bryden , in Fighting to Lose, argues that Ritter and the Abwehr were in control all the time. “Ritter found he could plant double agents on MI5 by allowing the British to intercept wireless messages in easy-to-break ciphers that referred to his spies before they set out for England,” he writes. His case is based on the fact that, since Hitler called off the invasion on September 8 (and that Canaris, the head of the Abwehr knew that), the landing of spies was a sop. Ritter was not taken in by the fake identity papers; Caroli and the others who arrived after him were intended to be caught. Ritter’s objective, by sending back lengthy questionnaires to SNOW, was at least to gain some useful intelligence. Bryden’s criticism of MI5 are justified, but his account does not explain how well TATE (especially) was able to mislead the Abwehr so successfully later in the war. And Germany’s attentions were now moving elsewhere: on December 18, 1940,  the directive for Operation Barbarossa – the invasion of the Soviet Union – was created. Britain was, temporarily, no longer subject to invasion.

Summary & Conclusions

Four major conclusions can be derived from the analysis of Phase 1:

  • The British authorities lost control of the supervision of the interception of illicit domestic wireless transmissions. MI5 fumbled by claiming it had neither the expertise nor the bandwidth to assume the responsibility. To help, it recruited an expert (Simpson) who was cast aside, with his recommendations being overthrown by the well-intentioned thoughts of a wireless interception expert but counter-espionage amateur (Gill), and an ineffective Signals officer (Worlledge). All three (as well as Denniston of GC & CS) were soon unceremoniously discarded. The official accounts of the decision to concentrate interception away from Britain points to a clumsy and grotesque attempt to conceal what must have been a major embarrassment. Curry referred to a ‘Greek Tragedy’, which suggests more than the casual sacrifice of Simpson: it suggests an undocumented drama affecting the essence of MI5. Did Simpson point out the irreconcilable conflict of having an efficient RDF capability, and running double agents through wireless transmission? Or do Curry’s words hint at a completely undocumented fiasco, such as an abandonment of surveillance of Soviet spy networks? The fact that the Double Cross Operation has been lauded as such a success would suggest the latter. The act of representing Britain’s D/F capabilities as ineffectual may have helped the deception campaign, but nothing I have found in the archives suggests that it was a deliberate decision, as opposed to an accident of circumstances. The answer may not be knowable, but my further research, as this saga evolves, may throw up something. Curry’s ‘Tragedy’ must mean something, and Andrew’s unwillingness to follow up on the issue is itself significant.
  • MI5’s counter-espionage section showed a combination of strategic imagination and operational clumsiness. For a group that was so unprepared as to what it should do when it found it had recruited Soviet spies to its corps, B Branch’s Liddell showed some farsighted insight when it came to the possibilities of using double agents for deception. Liddell quickly recognised the necessity of keeping British radio detection-finding capabilities under wraps, well before the massive deception campaign associated with FORTITUDE was conceived. Yet he was overcome by events rather than preparing for them: his mismanagement of Simpson and Frost, his reluctance to engage with Gill, and his slowness in reorganising his officers dealing with Nazi espionage, show that he was an ineffectual leader who did not show decisiveness in putting the structures and personnel in place for the smooth execution of its mission. True, he was operating under extreme pressures, but it is the duty of leaders to rise above them.
  • The Abwehr started off at half-cock with its wireless strategy, and displayed no firm intention that radio communications by spies would play a large part in the war. The agents they tried to insert at the end of 1940 were of low calibre and motivation, poorly prepared, and supplied with inferior equipment. The Abwehr’s treatment of SNOW remains an enigma. There were many in British Security who thought that SNOW was so unreliable that he should be dropped immediately. Yet MI5 and B.1a persevered. It is clear that Ritter and his cohorts had suspicions that SNOW was being run by their adversaries, yet they ignored the obvious signs. Why did they not suspect that the ID numbers passed on were phony? Or maybe they thought that it did not matter. John Bryden’s arguments should not be discounted completely, but the Abwehr’s behaviour was in many ways as naïve of that of the British.
  • While the evidence is sometimes contradictory (maybe deliberately so, some of it written long after), it seems that precise direction-finding and location of short-wave radio transmitters was at this stage of the war still something of a black art. The techniques of sending low-powered short-wave transmissions to bounce off the ionosphere, and skip large areas of potential interception, represented a considerable exposure to any nation’s domestic security. Both the German and the British intelligence organisation showed awareness that technology in this area was going to be a critical factor in the espionage and counter-espionage wars.

Thus Phase 1 of the saga came to a close. The focus in 1941 will be primarily on TATE and SUMMER, although it is in fact one of the quieter periods of wireless espionage. SNOW was determined to be too much of a risk, and he was soon taken out of commission. The second wave of double agents had yet to enter the stage. And the unsettling matter of reorganisation, of MI5 as well as the placement of RSS, would bring some clarity, some improvements, but also a lot of tension to the management of double agents.

(I am very grateful to Dr. Brian Austin, a retired academic from the Department of Electrical Engineering and Electronics at the University of Liverpool, for his recent guidance to me on matters of wireless telegraphy. Any mistakes or misrepresentations made in this piece are my responsibility alone.)

Chapter 2

Cyril Fairchild of RSS: Ditchling Beacon, 1939

When I set about my research into the puzzle of the apparent failure of the British Radio-Direction Finding mechanisms to detect the German agents incorporated in the Double-Cross System, I thought it would turn out to be a relatively straightforward case of guile – foolish, perhaps, and lucky – but still a feint. Yet my readings led me to conclude that here was a multi-dimensional enigma, involving the following conundrums: the bizarre and humbling treatment of Gill, after he made a breakthrough analysis; Gill’s mistake over the assumption that Hitler’s agents all had receivers as well as transmitters; the mystery of Lt.-Col. Simpson, who made a significant impact, but was almost completely removed from the records; the deceptions of Dick White about the timetable of the Double Cross System; the misrepresentations of Guy Liddell about his organisation; the official exaggeration of the Abwehr strategy, and finessing of some technical aspects of their agents’ method of operating; the contradictory representations, by various ‘experts’, of the state-of-the-art of wireless direction-finding; and the scanty coverage of the topic by the authorised historians.

Yet perhaps the most extraordinary finding was the almost apocalyptic observation that appeared in John Curry’s confidential history of MI5 compiled at the end of the war, asserting that the decisions made about the responsibility for the Radio Security Service (RSS) had caused a tragedy of Greek proportions to take place. This judgment was made when the war had recently been won, and the activities of the Double Cross Committee, in exploiting the agents under its control to promote the message that a dummy army (FUSAG) was assembled to invade the Pas de Calais, had been a primary contributor to the success. Was Curry hinting at the Cold War, and the betrayal of Eastern Europe by the Allies? Was he suggesting that British Intelligence had abdicated its responsibility for monitoring illicit Soviet transmissions? Did a careless decision not to deploy the RSS with the correct discipline allow the Soviets to transmit undetected, or did a careful decision to soft-pedal RSS in order to allow the spies to be surveilled open up a different exposure? Or was he simply lamenting the handing-over of control of RSS to SIS, with the struggle over the release of ISOS (Intelligence Service Oliver Strachey) material implying a colossal failure in joint intelligence? Given the political climate at the time, it is difficult to posit any other scenario beyond these. And, in fact, archival documents that have recently come to my attention firmly suggest that it was complacency about German agents that led to carelessness over other threats.

In my May blog, I had referred, in passing, to three documents written by the enigmatic Lt.-Col. Simpson that I believed were no longer extant. In June, through the agency of Dr. Brian Austin, I managed to contact a wartime RSS operator, one Bob King (who can be found in Pidgeon: see below) now in his nineties, who passed on to me a few files. One, though undated and unauthored, was surely an early draft of a contribution by John Curry of MI5 to his 1946 history of the institution (as the style was unmistakeable), but included comments that did not find their way into the eventual published version. The second was the 1938 report by Simpson on the threats constituted by the use of low-powered and miniaturized wireless transmitters in time of war, and what infrastructure, technology and organisation would be required to take on and eliminate such a menace. The discovery of this document is as if one of the lost plays of Aristophanes had suddenly been found. Likewise, I had not been able to locate this report from the Index of the National Archives at Kew, but, if any of Simpson’s contributions have been made publicly available, it astounds me that no historian appears to have grasped the significance of both these pieces. Another absorbing item is a report by an engineer who worked on a secret wireless interception project under the Metropolitan Police. I have no doubts whatever as to the authenticity of these documents, and shall use them (and others) to update the story in this entry. Moreover, in an email communication, Bob King assured me that Sonia’s illicit messages were picked up by the RSS, but the unit was told to ignore them. This nugget of information has enormous significance, and I shall address it in a future episode.

I had originally intended that this chapter would move the whole story – including progress in wireless transmission and detection techniques made by British, Soviet and German espionage and counter-espionage agencies – up to June 1942. The discovery of these new sources, however, means that this piece is dedicated to a deeper analysis of the evolution of RSS leading up to its transfer to SIS in the spring of 1941, and the immediate decisions made in the months afterwards.  I shall return to a full discussion of Phase 2 (January 1941 to June 1942) in a couple of months’ time.

RSS finds its Home

  • Background

For my research on the RSS as displayed in ‘Sonia’s Radio’, I had relied primarily on the Introduction to Hugh Trevor-Roper’s The Secret World, subtitled Behind the Curtain of British Intelligence in World War II and the Cold War, written by its Editor, Edward Harrison, for much of my information on the evolution of the Radio Security Service in the first two years of the war. That was complemented by a revealing chapter in Nigel West’s GCHQ, although West probably ascribes too much importance to the role of Lord Sandhurst, since West enjoyed exclusive access to the Sandhurst papers, and relied on them for much of his narrative. I found valuable, but mainly anecdotal, evidence in Geoffrey Pidgeon’s The Secret Wireless War, some rather fragmented accounts in Frank Birch’s often inscrutable Official History of British Sigint (which frequently reads as if it had been poorly translated from a foreign language, probably German), and some revealing but often imprecise material in Professor Hinsley’s official history of British Intelligence in the Second World War. Philip Davies’s MI6 and the Machinery of Spying is overall very thorough and contains good corrective analysis. But Keith Jeffery’s Secret History of MI6 is very disappointing in its coverage, considering that it is the authorised history, and that RSS was an integral part of SIS after the spring of 1941. I had inspected some of the source material at the National Archives on a visit in 2017, but, since little of it has been digitised, I have not been able to analyse any other since, apart from a few pieces shared by other researchers.

I recently discovered (thanks to Stan Ames, an RSS enthusiast) a longer paper published by Harrison, which appeared in the English Historical Review Vol. CXXIV, no. 506, dated January 13, 2009. It is titled ‘British Radio Security and Intelligence, 1939-1943’, and provides a very comprehensive account of this critical era in wireless and intelligence. Harrison, who suggests that his contribution ‘fills the gap’ in offering an academic article ‘dedicated to the organisation’ of RSS, generally provides an insightful guide to the literature, and skilfully exploits a broad number of sources. He crisply explains the evolution of RSS, taking the line that MI8 tried to find it a  home in MI5; that MI5 resisted, because of issues of overstretch and competence; how Walter Gill, introduced to the unit late in 1939, brought to it new skills in discrimination (isolating and organising signals of relevance from among a vast noise in the ether); how Gill’s findings shifted efforts towards Abwehr signals abroad rather than illicit transmissions from the UK; and how, because of this geographical re-focusing, with the approval of the imminently-to-be-appointed chief of MI5, David Petrie, RSS was handed over to SIS early in 1941, with official approval occurring in May. He then relates the continuing battles between MI5 and SIS – primarily through the personalities of Guy Liddell, Hugh Trevor-Roper and Maurice Cowgill – over the availability of ISOS decrypts that MI5 thought were critical for the smooth running of the Double-Cross system. It is a masterful and highly valuable contribution to the history.

Yet Harrison’s story does, I believe, not perform full justice to RSS, or describe accurately the manoeuvrings that went on behind the scenes to determine the control of RSS. It is a more a study of the relationships and tensions between MI5 and SIS than of the machinery and contributions of RSS itself, and Harrison is perhaps a touch too respectful of Trevor-Roper’s role, describing him as ‘the intellectual inspiration of RSS’. Moreover, Harrison largely ignores some of the figures who participated. He says nothing about Lord Sandhurst, who was appointed to RSS, and played some role in recruiting or training the Volunteer Interceptor force in the first months of the war. (As indicated above, this may have been a sagacious choice, as Sandhurst’s involvement remains somewhat controversial.) Harrison does not mention, however, the greater contribution of Lt.-Colonel Adrian Simpson, who wrote the seminal paper that defined the structures, technology and organisation that he felt were vital for protecting the nation’s defences. Harrison seems to be unaware of SIS’s own clandestine interception capabilities constructed in cooperation with the Metropolitan Police, documented by Kenworthy, and chronicled in the National Archives, which throw a bizarre light on the whole issue of MI5/SIS territorial control. He rather bizarrely devotes a section to Malcom Frost’s late efforts to increase the efficiency of the mobile detection units without offering an explanation of what illicit operators they were supposed to be pursuing. He mentions Richard Gambier-Parry, who headed SIS’s Section VIII, under which RSS resided, only in passing. He offers a restrained analysis of John Curry’s highly provocative assessment of the ‘Greek tragedy’ that resulted from SIS’s takeover of RSS, an opinion that Curry himself appeared to abjure elsewhere.

Moreover, Harrison brings to the surface a number of anomalies and paradoxes that are not satisfactorily addressed in his paper, and I have to backtrack a little to the topics I introduced in the first chapter of this saga to refresh the story. I should point out that I am not attempting to offer a comprehensive account of RSS’s history, but to focus on the questions highly relevant to radio interception and direction-finding policies in WWII. Who drove the takeover of RSS by SIS? Why were domestic interception and detection so casually executed? Why were Sonia’s radio transmissions overlooked? Why did British intelligence believe it could convince the Abwehr that the Double Cross agents had not been detectable?

  • The Strange Decline of Lt.-Col. Simpson

Simpson’s Report: First Page

Now that one of Simpson’s papers has come to light, one can understand his considerable strengths, as well as what probably caused him to fall into disfavour. (If not familiar with him, readers should inspect Chapter 1 of this saga first.) His October 1938 report to the Director of Security Service at the War Office, titled ‘Illicit W/T Communication’, is a masterful explanation of the way developments in wireless technology could allow a nest of foreign spies to remain undetected in Britain. He pointed out that low-power transmitters would be able to broadcast to receiving stations overseas (in Germany) while remaining difficult to detect locally via normal ground waves. He recommended the establishment of three fixed Direction-Finding (DF) stations, each complemented by a pair of portable (i.e. mobile) stations, that in turn would be supported by a set of hand apparatuses that could be used for house-to-house search. Landlines to connect the DF stations would be essential, and a line would also link the main DF station with the fixed Interception station. The project was to be enabled by the recruitment of ‘some 50 or 60 picked amateurs out of the 4,000 now existing in this country’; Simpson did add, however, that he believed that the creation of such an organisation was already under way.

Lt.-Colonel Simpson’s Plan for Interception, 1938

Simpson expressed concern about the suitability of the G.P.O., the institution currently chartered with executing MI5’s requirements in this area, since it had a more regulatory and bureaucratic approach to the issue of frequency usage. ‘Our objective’ (which should probably not be interpreted as ‘MI5’s objective’, but as a national interest), he said, is to prevent any unauthorised transmissions, not just investigate them after they had happened. That is why he focused on developing a more elite, professional staff from among all the amateurs who held experimental licenses. He did add, however, the intriguing comment that one of the objectives would be to ‘locate the source of transmission with the least possible delay, but not necessarily stop it’, hinting at the notion of possible control of alien broadcasts, but in fact suggesting a desire to distort the suspected propaganda signals to make them unintelligible. His final appeal was for centralised control over the whole process of interception, direction-finding, and message gathering, and that, when the collection ‘of a certain class of highly confidential intelligence’ had been made, it would be conveyed to the appropriate department ‘to take the necessary executive action’. Lastly, he nominated three very distinguished names to serve on a Technical Advisory Committee, Dr. James Robinson, Director of Wireless Research at the Air Ministry, Captain Round, an expert in DF and interception work, and Mr. K. Tremellen, ‘the greatest practical authority alive on the subject of short-wave communication’. Strangely, none of these names appears in the authorised histories.

Some of Simpson’s ideas would be echoed later (e.g. the need for unification of resources, the professionalisation of voluntary interceptors), but his recommendations were perhaps influenced by two notions that were gradually becoming obsolete: i) a too technical approach that emphasised that the problem was one simply of interception and location, not foreshadowing the technique of traffic analysis, and the way in which that process, alongside (even partial) decryption, fed back into the act of discrimination, and ii) the belief, perhaps encouraged by WWI memories of German spy threats, that the country was riddled with German agents, equipped with wireless, who were ready to spring into action. What is also significant that he articulated the mission as ‘closing . . . all illicit channels of communication with the enemy in time of war, and of locating sources of political propaganda in time of peace’. What he did not include was the need to protect the realm from hostile (not necessarily declared enemy) communications designed to help subvert the country – i.e. transmissions by Communist spies, whether in time of peace or war. This must have been a failure of knowledge or imagination, and it is astonishing that, since he was offering his report on behalf of MI5, he was allowed to make his submission to the Director of Security Service at the War Office without this oversight being pointed out.

John Bryden, in Fighting to Lose, suggests that MI5 rejected his ideas there and then, ‘being firmly of the view that German agents would only be using the mails or couriers to send in their reports’, and that the matter was turned back to the War Office. But that does not make sense. The source that Bryden provides for this explanation (Curry) does not give that as the reason: Curry blamed it on the administrative burden and financial commitments required. Moreover, despite the fact that the War Office approved Simpson’s recommendation that the RSS unit be set up, it did not endorse his ideas of ‘unified control’, and when MI5 declined to become involved, Simpson stayed on as the Security Service’s expert. He was surely happy to see his recommendations accepted, no matter where the unit reported. (His perspective on MI5 ownership is a little ambiguous: at one stage in his report he refers to ‘our’ DF or interception stations, but then goes on to write that they would be used ‘in conjunction with M.I.5.’ It appears he had an open mind on the command structure.) Bryden and Curry do agree, however, that the founding of MI1(g) was attributable to MI5’s lack of eagerness to take charge. Accordingly, RSS started collaborating with the Post Office in March 1939, with MI5 demoted to the sidelines, waiting for results.

Simpson may have been somewhat deflated, but thus hung around in MI5 (though without warm recognition from Liddell, his boss in MI5’s Counter-Espionage B Division). The fragment from Curry indicates that he was vigorously promoting his original vision of unified control, and stressing the importance of the Post Office in harnessing the appropriate resources to tackle the threat of illicit transmissions by supplying suitable personnel, and moving to build the new facilities required. Indeed, Curry reports that Simpson was the main muscle behind the establishment of the Voluntary Interceptor system: a recognition that other commentators have overlooked. As B3b, he was actively supplying the liaison between MI5 that was later mirrored in SIS’s Section V. As MI5’s representative on the Technical Committee on Leakage of Information (TCLI) that the War Office set up in October, 1939, he was quick (in February 1940) to try to persuade the Ministry of Home Security to bring pressure on to the GPO. He attended the critical meeting on March 20 at Bletchley Park after which GC&CS agreed to set up the ISOS decryption unit. Yet his stubbornness in believing that a domestic German menace was being overlooked (when none existed) must have clashed with the messages coming from RSS. His emphasis on the need for widely dispersed Voluntary Interceptors to pick up illicit ground signals turned out to be something of a luxury, although the wide dissemination of interceptors greatly aided the ability of the unit to avoid omissions provoked by the whimsicality of ‘skip zones’ and the presence of thunderstorms. His expressed frustrations with the GPO’s lack of urgency in constructing new DF and Interception stations was probably on target, but his insistence that the detection of illicit wireless was ‘extremely unsatisfactory’ was not.

Maybe the SNOW affair changed Liddell’s mind somewhat. Simpson’s ideas must have had a slight resurgence with the ‘Fifth Column’ scare in the summer of 1940, but Liddell’s entering discussions with ‘the BBC man’ Malcolm Frost in May 1940 suggests that Simpson was no longer around. (Frost had been the BBC representative on the TCLI, and thus presumably had caught Liddell’s eye as a possible replacement for Simpson.) Indeed, the system of Regional Security Liaison Officers that MI5 set up by Guy Liddell in June 1940, specifically to address the threat of illicit wireless (and which was headed by Jane Archer, mysteriously sidelined from her expert role in tracking Communist subversion) mapped very closely to Simpson’s areas of demarcation. But when that was shown to be a false alarm, his whole infrastructure was seen to be somewhat redundant, especially in the light of the lessons being learned by Gill and Trevor-Roper in the RSS organisation. Interceptors were needed in large numbers, but did not have to be located so evenly around the country in order to pick up ground waves. Simpson’s attendance at the meeting at Bletchley Park where the revelations about the discovery of Abwehr traffic were made is the last reference that Liddell makes to him in his Diaries.

Still, Simpson’s omission from the record books (outside Curry) is extremely puzzling, and some of his contribution remains uncredited. For example, his report clearly refers to the 4,000 amateurs known to the Post Office who had the potential of providing the elite force that Simpson needed. Yet most histories and memoirs attribute the imaginative idea to Lord Sandhurst, who was reportedly recruited by RSS at the outbreak of war to develop a professional force of interceptors to replace the largely part-time group assembled by Colonel Worlledge. Sandhurst, who had also been instructed to liaise with R. L. Hughes of MI5 (who, Curry informs us was B3b, responsible for liaison with the RSS and the BBC, and thus working directly for Simpson at that time), soon approached Arthur Watts, the President of the Radio Society of Great Britain. Watts had ‘several thousand’ members who were radio hams, so Sandhurst then began to select the most suitable for training. Thus Simpson’s contribution is overlooked: Davies, like Harrison, remarkably makes no mention of wireless expertise in MI5 before Frost.  Simpson will turn up again in this account, when I write about the negotiations to find RSS a suitable home, but the verdict on his contribution must be that he was technically correct, but strategically wrong. He brilliantly assessed the state of the art of short-wave wireless telegraphy, and its potential subversive use, but he was caught up in the tide of searching for a phantom menace – the German W/T agents installed in the English countryside – and failed to gain the confidence of his colleagues in MI5. The irony was that the flock of interceptors he identified to protect the nation did not need to be precisely dispersed to detect ground waves, as there were no illicit operators at large at that time, but the volume and placement of such individuals did turn out to be essential to pick up the mass of signals originating from overseas.

  • The Rise and Fall of Walter Gill

Walter Gill

Walter Gill, on the other hand, was (in a specialised sense) technically wrong, but strategically correct. It still comes as a surprise to some observers that nearly all the Abwehr agents infiltrated by air or sea in 1940 were equipped only with a transmitter, and not with a combined transmitter-receiver, or with a separate receiver. Operating ‘blind’, without any confirmation that one’s message was being received at all, or perhaps not clearly enough (and thus needed to be re-sent) would appear to reflect a less than serious objective by the perpetrators of the scheme. And that is one interpretation that can be cast on the German planning, as I have suggested. (Preparations for sending agents into Britain did not begin until July 1940.) Yet that phenomenon is confirmed by the archival reports, as well as by the memoirs of some of the members of RSS. While Gill showed great insight over the question of the source of beams guiding German aircraft, his thesis, that if the supposed German agents could hear their controllers’ signals, then so should the operators in RSS have been able to, and that there were therefore none operating, was based on a false assumption. The focus on enemy signals originating abroad, and the eventual deciphering of many of them (ULTRA), was, however, a major contributor to the success of the war.

Gill’s policy must come under continual scrutiny, however. I have recently read accounts of two Abwehr agents who parachuted on to English soil before the main wave (Operation LENA) that arrived in early September 1940. Each of this pair was reported to have brought a working transmitter/receiver unit and successfully exchanged messages with his controller. Such transmissions were presumably not detected by RSS, since Gill claimed the unit had not identified any unexplained outgoing Abwehr signals. Such agents might therefore have been able to transmit undetected for some time, contrary to the accounts that the authorised and semi-official historians would have us believe. I shall investigate such adventures in my next chapter, to judge whether this was all an elaborate hoax. It should perhaps also be noted that Gill came to his breakthrough conclusion about the absence of German agents in Britain in December 1939, when SNOW was, almost certainly, the only wireless operator recruited by the Abwehr. His report, however, was not written until November, 1940, when the experience of Operation LENA, under which a dozen or more spies landed on British soil, would have sharpened sensitivities in MI5. Indeed, as early as July 13, 1940, Liddell felt compelled to record in his diary the following: “While I feel it is likely that there are a few German agents here, possibly transmitting by wireless, I do not envisage anything in the nature of large bodies of individuals going out to stab us in the back as soon as the Germans invade this country.” That observation indicates that the Gill doctrine had not been accepted wholesale at that time, and Liddell did not have complete trust in the energies of RSS.

I have little here to add to my account of Gill’s demise that I described two months ago, but the account that Hugh Trevor-Roper gave of Gill’s departure is somewhat paradoxical. Trevor-Roper was known for his caustic dismissals of many of those he encountered in wartime, especially the blimpish characters he considered to be his intellectual inferiors, but he clearly had some admiration and affection for Gill. Gill had been a lecturer on electricity at Oxford University, and a successful Bursar at Merton College, although Trevor-Roper had diminished his overall academic qualifications by writing that he ‘could only by a charitable laxity of definition be included among the educated’, a harsh and inaccurate judgment (as revealed in Dr. Austin’s detailed profile of him), which sheds more light on Trevor-Roper’s arrogance than on Gill’s cultural accomplishments. Yet they worked well together as a team. Trevor-Roper, however, when commenting on Gill’s clumsy and harsh dismissal and demotion, could only comment (in Sideways into SIS) as follows: “The real genius of the affair, Major Gill, was also deliberately overlooked. Left to find other employment, he became a radar officer and an expert on captured German equipment. Under the new regime, his name was never mentioned.”

Was there a reason for Trevor-Roper’s coyness over Gill’s treatment, which he also characterised simply as ‘rather shabby’? After all, Gill had been fired without even a formal notification, and then demoted from Major to Captain. Major Cowgill, the offended SIS officer (who had joined SIS only in March 1939, so did not enjoy a reputation of any sort), had repeatedly called for Trevor-Roper, who had been just as complicit in the affair as Gill, to be court-martialled. Yet Trevor-Roper escaped unscathed, even though the head of RSS, Colonel Worlledge, lost his job as well. It is surprising that Trevor-Roper did not provide a more comprehensive coverage of the whole business. In fact he concluded that Cowgill in fact ‘had every right to explode’, as Worlledge had revealed secrets concerning intelligence and security ‘not only to his official contacts in the Armed Services intelligence departments . . . but also (horror of horrors!) to the civilians of the Post Office.’ Perhaps Gill and Worlledge were punished because, as military veterans from WWI, they should have known better. In fact, as will be shown, it was a bit more complicated than that.

One last mysterious incident concerns Gill’s reappearance in April 1942. Despite what Trevor-Roper wrote over fifty years later, Gill’s name was apparently mentioned again, because (as Harrison reports) Trevor-Roper was in contact by letter with F. E. Lindemann (later Lord Cherwell), who was Churchill’s scientific adviser, at a time when Trevor-Roper, disenchanted again with his work in SIS, was looking for other opportunities. As Adam’s Sisman’s biography of Trevor-Roper makes clear, he was highly frustrated over the failure of RSS management (Maltby and Gambier-Parry) to keep technical policy aligned with intelligence aims. He had earlier rather indiscreetly criticised the leaders of RSS, specifically Gambier-Parry, and now told Lindemann that Menzies (chief of SIS) had called him in, and then accused him (Trevor-Roper) ‘of having supplied facts to Gill which he had supplied to you and you to Swinton’. Is that ‘he’ Menzies or Gill? Unlikely the former, as Trevor-Roper would presumably not have been party to information passed by Menzies to Lindemann: Menzies would in that case have been concerned about a breach of security elsewhere.

So if it was Gill supplying facts to Lindemann that got back to Menzies via Swinton of the Security Executive, what could those facts have been about, and on what basis were Gill and Trevor-Roper still in communication over important matters if Gill was by then working in a completely unrelated sphere of the war effort? And why would Gill want to leak secrets to Lindemann? It may be relevant that, at exactly this time, as Dr. Austin informs us, Gill joined the Army Operational Research Group, where he was responsible for investigating advanced aspects of Army field communications, but no details of the exchange have come to light. It sounds very much as if Gill and Trevor-Roper had stayed in touch, as ex-colleagues who had collaborated very productively on the matter of intelligent signals analysis, and that Gill was a man whose reputation had been restored, and had connections with influential persons. Another interesting twist to the story (as related by Sisman) is that when Trevor-Roper made a trip to Ireland in early 1942, i.e. just before the contact with Gill, Colonel Worlledge invited him to his home, Glenwilliam Castle, where ‘over a convivial dinner each outlined to the other what he knew of the takeover of RSS by SIS’. The existence of this conversation hints at untold scheming and plotting. Vivian of SIS was later to use this incident to make the astonishing claim that Trevor-Roper had gone to Dublin to betray the Ultra secret to the Germans, and that he had been ‘motivated by resentment against SIS for its treatment of Worlledge, and of Gill in particular.’ (Vivian was by now unstable: Liddell reports that he suffered a nervous breakdown in June 1942.) Trevor-Roper’s published account of Gill’s dismissal was clearly much more muted than this: he was surely concealing something of substance, but it may have no important connection with the fate and mission of RSS.

Gill’s major contribution to the debate about RSS’s future was his November 19, 1940, paper on the Organisation of RSS. Curry represented the arguments therein (the whole Theseus episode, after which focus was shifted to interception of overseas transmissions) as a clinching argument for RSS’s ‘vitality and value’, and for moving it into MI5, but that judgment appears weak and woolly. The timing of this report suggests it was produced under some pressure, but Gill’s account expresses no concern about the current organisation, or the allocation of work between RSS & GC&CS, and it concludes simply with a modest request for more resources. Yet the report includes a very telling statement concerning Direction Finding: “Any of the residue [i.e. the messages remaining after known ones had been identified] found by D.F. to be outside the country could for the above purpose have been neglected [but were not].”  RSS was successful in tracking those same messages, but, by implication, some unknown messages did originate inside the country. Gill gave, however, no indication of how these were investigated, a statement that should have alarmed MI5’s officers. If anything, the case as he made it was an argument against moving the unit to MI5, contrary to what Curry claimed. As we shall see, the question of territory and ownership would play a strong role in the decision, and MI5, even if the service had an outspoken champion, was on its weakest footing at this stage. The transfer to MI5 of course did not happen, but it did provoke a major debate about where RSS should report. Had Gill performed his job, and was thus no longer needed? Or was his demise just an accident of politics? That question may be unanswerable.

  • Kenworthy and the Secret Interception Unit

SIS was a notoriously secret organisation, but even it had clandestine corners that were not apparently known to all its officers, or even its authorised historian. In Keith Jeffery’s Secret History of MI6 the author informed us that the strategic split between the responsibilities of MI5 and MI6 (SIS) was made on October 1, 1931, when the semi-autonomous unit of the Special Branch, SS1, which consisted of the familiar Guy Liddell and his colleague Hugh Miller, experts in counter-subversion, was peeled off from the Metropolitan Police and handed over to MI5. SIS was also stripped of its domestic intelligence network, the ‘Casuals’, which was causing an embarrassment. This decision apparently simplified and clarified the missions for MI5 and SIS to handle subversion in the Empire and in foreign countries, respectively. “Thus . . . the Secret Intelligence Service and the Security Service took on their modern form and distinct spheres of responsibility which were to survive for at least the next eight years”, Jeffery wrote, with a high degree of authority (p 236).

Yet it was not quite like that. The reader will learn, from Nigel West’s 1986 book, GCHQ, that in 1930, a Commander Kenworthy reported ‘an illicit Comintern circuit operating between a site just outside Moscow and a terraced house in a suburb of London’. (The Moscow location was verified by direction-finders located in Palestine’s Sarafand, in India, and in London, thereby showing that widely dispersed location-finders working in harness could place remote transmitters with an accuracy that could not always be exercised in more confined areas. Such phenomena perplexed security officers like Liddell.) West added that Kenworthy was ‘the controller of the Home Office intercept station at Grove Park, Camberwell’. It might surprise some that the Home Office was involved with interception. Indeed, in West’s later book (2005) on this Comintern project, MASK, the author informs us that ‘GC&CS’s [sic] monitoring station at Grove Park, Camberwell, headed by Commander Kenworthy, first began intercepting Wheeton’s signals in February 1934 . . .’. Aided by the revelation by an MI5 mole of the cipher used, the codebreakers Leslie Lambert and John Tiltman were able to read the traffic until January 1937. By employing the full force of the direction-finding equipment of the Army (Fort Bridgewoods), Navy (Flowerdown) and Air Force (Waddington), the team of technicians were able to locate the members of a worldwide Comintern ring.

Fort Bridgewoods

The intercept station, however, was not run consciously by the Home Office or by GC&CS. It was run clandestinely by the Metropolitan Police. We owe it to a memorandum by Kenworthy himself, available at HW 3/81 at the National Archives, for a richer account of how Special Branch, assisted by both SIS and MI5, kept a watch on traffic that the armed forces declined to surveille. Supported by secret funds, an interception unit was encouraged by its experience in the General Strike (1926) to seek support from SIS in trying to detect foreign diplomatic stations which did not have ‘Berne List’ status (the latter presumably representing official frequencies allocated by international agreement). Kenworthy made it clear that Admiral Sinclair, the chief of SIS, was intimately familiar with what was going on. Remarkably, Kenworthy indicated that the expertise in interception gained by his unit entitled him to attend Y [= Signals Interception] Committee meetings, where the Services ‘looked to him for guidance’. He described his success in locating the illicit Comintern operator in Wimbledon, also showing that he and his colleague Lambert developed a portable direction-finding piece of apparatus that was critical for their mission.

What is intriguing is that The Metropolitan Police was the institution responsible for tracking the increasing volume of diplomatic traffic that appeared in the 1930s. “The Services were however disinclined to intercept Diplomatic (Commercial) Wireless to any extent as it would lead to a curtailment of the examination of their particular Service channels of Foreign Countries, as it became more and more important that encouragement should be given to Police by S.I.S”, Kenworthy wrote. Soon SIS was funding the exercise, as it was difficult to account for the expenses internally, and not long thereafter the new Receiving Station at Denmark Hill was constructed. Some official funding was approved, and made public, in 1938, but SIS maintained a controlling interest in the project. (At the base of one of his many organisational charts, Birch lists the Police Y Station at Denmark Hill as being controlled by the Foreign Office, i.e. SIS’s sponsor!) Now the interest of GC&CS (which reported to Admiral Sinclair, SIS’s chief) was piqued. In 1939 it decided that Commercial traffic should be intercepted as well, requiring a workload that Denmark Hill could not handle. “G.C. & C.S. realised that more facilities were required but unfortunately they had to cloak their activities and could not openly control wireless stations.” Everything that was going on was contrary to the rules of the protocol-oriented GPO. The outcome was that a new interception station was set up at Sandridge, near St. Albans, ‘specially for G.C. & C. S.’. Finally, to tidy up the picture, GC&CS took over the complete Police signals intelligence capability between November 1939 and January 1940, as the summary of the relevant files at the National Archives website informs us. (Regrettably, I have not yet been able to inspect the complete file.)

This whole chapter in British signals intelligence contains some remarkable ironies. The first is that the task of intercepting commercial and diplomatic traffic had devolved to a clandestine unit of Scotland Yard, a fact that appears to have been overlooked by all historians except Frank Birch. (HW 3/81 was not declassified until 2004: Andrew and Jeffery would have had access to it anyway, but chose not to use it.) The second is that SIS was involved in intercepting traffic occurring within the territorial boundaries of the UK, which flagrantly broke the rules that had been set up in 1931 guiding the missions of the two intelligence services. Since one of the main planks of the argument for placing, in early 1941, RSS under SIS’s aegis was the fact that RSS, after the beginning of the war, changed its focus from domestic to international interception, the episode sheds fresh light on the sincerity and professionalism of Sinclair and Menzies. The third irony is that MI5 knew all about this incursion on its turf, but apparently did not raise any protest: Curry mentions, without judgment, that ‘a certain amount of interception work was being done by M.I.6’, referring to the illicit set operated by the Russians. (One of Kenworthy’s paragraphs reads: “A conference took place with S.I.S. and M.I.5. The latter pointed out that strictly speaking the G.P.O. as the Communication Authority were the Department who should tackle these sorts of jobs but for reasons best known to S.I.S. and M.I.5. G.P.O. were not considered a very secure body.”) In early 1941, the Security Service, already weak in its drive and leadership, would have been on insecure footing had it tried to play the territorial card.

The fourth irony is that GC&CS was allowed to enter the interception game at the beginning of the war (the transfer presumably muscled through quickly by Menzies) at a time when Commander Denniston was making vigorous representations about interceptors invading his own domain of cryptography, an action that led to Worlledge and Gill losing their jobs. Denniston was extremely possessive about GC&CS’s ownership of cryptanalysis, even though he and others (according to Birch) accepted that ‘Y generally involved interception, traffic analysis and ‘low-level cryptanalysis’. But Hinsley also records that, in the summer of 1940, Denniston opposed the demand from MI8 (RSS) that its Traffic Analysis staff of 70 officers be transferred to GC&CS (on the basis that Traffic Analysis and cryptanalysis should be done in the same place), on the grounds that ‘his establishment should continue to be a cryptanalytical centre’ (only).

Kenworthy thus moved to GC&CS, worked there during the war, when it became GCHQ, and retired in 1957. Though working for Bletchley Park, he was stationed at Knockholt, where he led the project to intercept German Teleprinter Communications. This was the very important ‘Fish’ set of non-Morse messages, and Kenworthy wrote a report on that activity in 1946. But of enduring interest to this area of research is his achievement in developing, so early, effective handheld location-finding equipment. I have not yet been able to trace the extent to which his inventions passed on to the GPO in wartime, apart from a brief mention by Curry, who stated that Kenworthy’s portable D-F set was tested by MI5, and that ‘some interesting Mobile Unit operations were carried out on connection with this case [the detection of the Comintern transmissions]’. I thus have not been able to determine whether the apparent dilatoriness of the GPO – so frequently demeaned by intelligence officers – was caused by inadequate technology or by official edict.

  • The Transfer to SIS

So was the transfer of RSS to SIS a smooth operation, or was it bedevilled with conflict and controversy? One can learn little from the authorised histories. The History of British Intelligence in the Second World War contains some errors, as well as some very puzzling observations that do not always make sense. Christopher Andrew does not mention the episode at all, or even the mission that MI5 shared with MI8/RSS. You will not find Lt.-Colonel Simpson, Malcolm Frost, the RSS, or even Section B3, in his Index. Keith Jeffery devotes just two sentences of his equally massive book to the adoption by Section VIII of RSS, indicating simply that it occurred ‘on Petrie’s recommendation’. He has nothing to say about Trevor-Roper, and Cowgill receives just a cursory mention. Geoffrey Pidgeon records the event as follows: “In January 1941, Swinton recommended that RSS be handed over to SIS, but this met with fierce opposition throughout the upper echelons of MI5, resulting in a battle that reached the highest levels”.  However, since Pidgeon (like many commentators) appeared to be under the impression that RSS had been run hitherto by MI5, his account may have been coloured. Nigel West, in his 1985 history, MI5, represents the struggle as one more between Menzies, the SIS chief, and Worlledge of RSS than a conflict between SIS and MI5, although West’s somewhat haphazard chronology of events means it is difficult to follow his narrative. He does, however, make the provocative claim that the change-over ‘was, in effect, “C”s (i.e. Menzies’s) final consolidation of his grasp on signals interception, and was only achieved after a closely-fought struggle with MI5’s ‘old guard’, but this interesting thread is not picked up or developed in his history, MI6, which came out two years later. Since Menzies did not assume his leadership of SIS until November 1939, and did not enjoy a reputation as a deep thinker or strategist, West’s opinion comes over as rather startling. I shall return to it later.

Stewart Menzies

So what does the evidence indicate? Birch suggests that several agencies had had their eyes on the prize of domestic interception, namely MI1b, MI5, SIS, the armed services, the police and the Post Office, before the 1938 decision that the War Office should be in charge, and the establishment of RSS. MI5 had a natural interest, because of the mission it shared with the unit, but, as has been explained, was reluctant to plunge in. Lt.-Col. Simpson must have grown frustrated, because he expressed his dissatisfaction with the state of affairs on September 15, 1939, and, according to Curry ‘suggested that the assistance of Colonel xxxxxxxx [name redacted] (an M.I.6. officer) should be sought’. Now, there were not many established Colonels in SIS at that time. Sinclair, mortally ill, was an Admiral, and Colonel Dansey was absent in Switzerland between September and November 1939. Unless Simpson intended to invoke Dansey, not knowing he was abroad, that left Colonel Menzies, head of Section II (military) and Sinclair’s deputy, Colonel Vivian, head of Section V (counter-espionage). Another officer, however, had been promoted to Colonel earlier that year – Richard Gambier-Parry, head of the Communications unit, Section VIII. There is no doubt, given the length of the name redacted, that it is he whom Simpson approached, and the significance of this connection will be explored later. It is not clear why Simpson decided to voice his frustrations at this time, apart from the fact that war had recently been declared. Was he annoyed at the pace of RSS hiring? Or at the shift to tracking overseas transmissions? Or at MI5’s continued reticence to grab the bull by the horns? The fragment from Curry’s report indicates that his ideas had moved on to consider broader issues of signals security, but his plans at that time encompassed a leading role for MI5 as the hub of a wireless intelligence organisation that it must have been reluctant to assume. Perhaps Gambier-Parry was an old ally, and this was a move to invite SIS to step in. But Simpson might have upset his bosses in MI5 during the process.

Richard Gambier-Parry

RSS was in fact moving along reasonably well: the GPO must have been propelled into action, though perhaps reluctantly. It did not think its mission was to build interception stations that would be focusing on detecting traffic originating from overseas. Three new Direction-Finding stations had been set up, and arrangements had been made in August for linking them with telephone lines. Lord Sandhurst was turning the corps of radio amateurs into a more professional body, though perhaps not as quickly as Simpson would have liked. As Nigel West writes: “The operators had to be skilled, discreet and dedicated, so the recruitment process was necessarily slow. By Christmas 1939, the Home South Region boasted only seven VIs (Voluntary Interceptors) on its roll.” Within three months, RSS had recruited fifty VIs, who were tracking 600 sources – all on the other side of the Channel. West reports that the Home South section had produced 1,932 logs by the end of the year, a figure that grew to 3,052 by March 1940. And, by that time, Gill and Trevor-Roper had cracked the Abwehr hand-cipher, and Bletchley Park had agreed to set up a special-purpose cryptographic unit to handle the traffic. RSS’s reputation was on the rise, but its role probably not broadly understood.

At the same time, fierce arguments over policy and organisation were being discussed by members of the Y Committee, which broadly was responsible for interception, traffic analysis, and low-grade cryptography. There were disagreements about the degree to which the needs of the three Services should be shared, or kept separate, but there was also questioning as to why SIS (whose head, Menzies, chaired the meetings) should control proceedings. It took an appeal to Lord Hankey, the ultimate committee man, for a solution, which involved a stronger Y Committee with a full-time chairman, and supporting clerical staff. Frank Birch suggests some of the confusion when he indicates that the news about the interception and decryption of Abwehr traffic in Europe, and the establishment of GC&CS’s ISOS group appeared to come as some surprise to the committee. ‘Officially, all this was no one’s concern’, he wrote, but in May 1940 the Committee gave formal recognition to the extension of RSS’s responsibility to provide preliminary investigation of these groups of signals. Seven months into the war, the Committee was still in reactive mode, instead of setting policy. The full Committee met for the first time not until January 1, 1941.

In the summer of 1940, after Simpson’s departure, Liddell also found a new candidate to lead B3b (Simpson’s unit), one Maurice Frost of the BBC, whom Swinton encouraged Liddell to hire. After initial good impressions, Frost was signed up, and in June 1940, Liddell reported plans for Frost to set up a new branch (the W Branch), instead of reporting to Liddell in B. The decision was made in July, and ‘Tar’ Robertson (who was handling SNOW) was deputed to work for him. But Liddell had to backtrack, and in August the W unit was folded back into B Branch, much to Frost’s annoyance. (Curry’s report states that Frost was Director of the W Division at this time ‘which comprised B.3’. It is probable that Liddell’s journal is more accurate than Curry’s memory on this matter.  MI5 was also notoriously inconsistent in its naming conventions for Branches and Divisions.) Yet Frost was beginning to get under everybody’s skin by this time. Robertson declared he could not work under him, and even Lord Swinton, who had supported Frost’s recruitment, said in late November 1940 that Frost could not stay in MI5. His ambition and untrustworthiness had become intolerable: moreover, he probably did not possess the appropriate skills for such a job. His interest was more in establishing a service to monitor foreign broadcasts.

Matters appeared to come to a disruptive head in September. According to Hinsley, the War Office concluded that its own interception capabilities (of German Air Force Enigma traffic) were not keeping up with GC&CS’s capacity to absorb it. Thus, on Winston’s Churchill’s bidding, Hankey ordered a transfer of an unspecified number of ‘operators’ from RSS to the Services, ‘overruling RSS’s protests’. This was probably a gross misjudgement: the failure to detect the enemy’s movements in the Nazis’ overrun of Europe in the summer of 1940 was due more to an incapacity to analyse and integrate intelligence properly than a paucity of intercepts. That was the insight that Gill and Trevor-Roper had arrived at.  Moreover, the War Office was responsible for MI8, which was where the unit reported. RSS received intercepts from its team of VIs, the permanent stations managed by the Post Office, as well the Armed Forces, the BBC and the cable companies, so simply shifting operators around was not likely to fix the poorly identified problem. Somehow the discoveries that Gill and Trevor-Roper had made about Abwehr communications with agents as the German war machine moved across Europe in the summer of 1940 should have made it to the General Staff, but there was no mechanism for that to happen.

By now, however, MI8 was feeling the pressure. On October 9 it pushed MI5 to take over the RSS unit en bloc, as it needed to concentrate on military matters, clearly not understanding that the work that RSS was doing was much more closely related to the theatre of war than the stated mission of detecting illicit domestic transmissions. But, of course, MI5 did know. Moreover, Brigadier Allen (MI5’s assistant director) went on record as saying that the service was being asked to take over an organisation that was breaking down. MI5 thus still demurred, because of cost and complexity, and because it understood that the current concentration on Abwehr traffic in Europe (and beyond) made the procedural case for the responsibility’s belonging to the Security Service completely tenuous. MI8 and MI5 were at cross-purposes. No doubt the secret but successful execution of an unchartered mission had to be revealed. The publication of Gill’s report in November 1940 thus brought the achievements of RSS into the open, perhaps preventing any further poaching by the Military, but inevitably driving the unit further away from MI5.

MI5 was also experiencing considerable turmoil at the time: even as Vivian of SIS was reminding MI5 officers (via Jasper Harker) of the correct procedures for communicating with SIS, Liddell was lobbying for Vivian to head MI5, so confusing was the current leadership. Lord Swinton, who headed the Security Executive set up by Churchill, had made life difficult for acting Director-General Jasper Harker, and had inserted William Crocker as an awkward co-head of B Division with Liddell. On December 3, 1940, Churchill’s security adviser, Desmond Morton, had told the Premier that MI5 was ‘close to collapse’, but the previous month the Lord President of the Council, John Anderson, had already brought in David Petrie to review its operation. Petrie had in fact been offered the job of Director-General, but declined to accept until he had performed a proper survey of the operation. He did not complete his report until February 13, 1941, but by January 30 he had already recommended to Swinton that SIS take over RSS. Where is the evidence of the struggle of ‘MI5’s old guard’, identified by West? It seems they put up no fight at all.

David Petrie

Yet the same day that Petrie arrived in MI5 to perform his investigation (January 1), Swinton approached the Director of Military Intelligence, Major-General F. H. N. Davidson to discuss the future of RSS. In an exchange that underlined what critical observers might say about the oxymoron of ‘military intelligence’, Davidson was reported to respond that he found RSS and related matters ‘very interesting, very complicated, and a strain on one’s brain’. Maybe this ‘very model of a modern major-general’ was simply overwhelmed, since he had assumed his new post only the previous month. Harrison, having inspected the Davidson papers, observes that Davidson noted in his diary that Swinton was ‘not satisfied that it [RSS] was doing its stuff’. Whether Swinton understood what RSS’s ‘stuff’ was, or consulted Lt.-Colonel Simpson, as a possibly sharper analyst of RSS’s failings, is not recorded. Davidson’s overall contribution is ambiguous: Cavendish-Bentinck, a normally good judge of character, who was the highly successful Chairman of the Joint Intelligence Committee (JIC) for most of the war, recorded that Davidson was ‘a very mediocre officer, with a permanent desire to make our reports fit in with the views of the CIGS [Chiefs of the Imperial General Staff].’ On the other hand, the official history of the JIC makes it clear that Davidson sensibly pressed, in January 1941, for a more integrated view of intelligence to assist the war effort – although he did not include MI5 or SIS in his choice of contributors to the process.

When Worlledge opposed the transfer on February 14 (’vigorously’, as Davies informs us), he also cast aspersions on SIS’s administrative skills, as well as making tactless criticisms of Gambier-Parry’s technical abilities, a mis-step that would later cost him his job. Why Worlledge was so set against SIS’s taking over RSS is puzzling, since it should have been clear to him that MI5 was even less endowed with managerial and technical talent than SIS. Hinsley writes that ‘the MI8 Colonel in any case fervently believed that the Services should control Sigint in time of war’. Was Worlledge perhaps aware of the Metropolitan Police unit, and its mobile detection exercises over the Comintern spies, and harboured some doubts about SIS’s interception policy and strategy?

Maybe Davidson was a fast learner, and had quickly unravelled the complications of RSS. The next day, he questioned Petrie’s decision, pointing out that ‘MI6 is concerned with the transmitting of signals and not their interception or location’, a claim that, as has been shown above, merely indicated that the Director of Military Intelligence did not know the full story of what was going on in the world of interception. Yet Davidson’s preference appeared to be to keep RSS under MI8 control rather than pass it to MI5, echoing his clearly diminished regard for the civilian services. Swinton coolly demolished Davidson’s objections, drawing on his position as supremo of both Intelligence Services to ensure that matters would work out fine, that the necessary committees would be in place to handle overlaps and conflicts, and that more professional training of RSS personnel would address his colleague’s concerns. Davidson was subdued, but not eliminated as a threat. Nigel West informs us that Davidson would later cross swords with Menzies, as he was not happy about the civilian nature of GC&CS, and wanted to wrest control back to the War Office. He believed the Office had not gained the results from interception which it merited for the investment it had made.

Yet another extraordinary step occurred before the eventual decision was made. According to Curry: “Early in 1941 it was suggested that an independent adviser, Mr. Kirke of the B.B.C., should carry out an investigation into R.S.S. organisation from the technical point of view and make recommendations for its future running by M.I.5.” The passive voice disguises an unlikely initiative: that the opinion of a BBC manager, supposedly independent of Frost and his objectives, might have been considered a fair judge of the best home for RSS, with the outcome of the investigation apparently pre-determined, and when in the past year the unit had moved well away from its mission of tracking voice broadcasting, and Frost himself had fallen out of favour, is simply shocking. Unsurprisingly, ‘this proposal aroused considerable opposition’. Curry nevertheless noted that ‘although it was partially carried out’, it resulted in meetings between the Director-General of the Security Service and representatives of SIS. Unsurprisingly, Petrie’s recommendation held. Liddell reported in his diary entry for March 6 that Gambier-Parry of SIS was taking over RSS, and the formal transfer occurred the next day.

  • The Aftermath: RSS under Gambier-Parry

RSS was indeed transferred to the control of Colonel Gambier-Parry in Section VIII of SIS. Gambier-Parry was a larger-than-life character who had been recruited by Sinclair in 1938 to fix the ailing communications systems of SIS and its satellites overseas. Gambier-Parry was an expert on radio: he had worked for the BBC, and for Philco, an American radio company. He had a reputation for being able to get things done, while showing a disdain for any bureaucrats who placed constraints on his will. From most accounts of those who worked for him, he was a popular figure who brought much energy and understanding to the complex challenges facing SIS. He thus embarked on a crash programme of building transmitter-receivers for the locations on the Continent, establishing broadcasting stations in safe places on the UK mainland, and devising the protocols to allow them to communicate securely.

Section VIII was certainly not in the business of interception – overtly, at least. Yet an enigmatic comment by Keith Jeffery in his history of SIS hints at a perhaps clandestine programme that has otherwise escaped the analysts. When Maurice Hankey performed his investigation into SIS at the beginning of 1940, one of the officers he interviewed was Rear-Admiral Godfrey, the Director of Naval Intelligence, seeking his views on the effectiveness of the Secret Intelligence Service. Godfrey was less than enthusiastic about GC&CS, though Jeffery then wrote: “But for Godfrey ‘the one really bright spot’ was the ‘” Y’ side”, in particular the intercepted signals and call signs, which the Admiralty found of the greatest possible use. All praise for this state of affairs’, he added, ‘was due to Colonel Gambier-Parry’. Now Admiral Godfrey was no slouch: he was a well-respected intelligence officer (celebrated for being Ian Fleming’s boss and mentor), and had even been a candidate to replace Admiral Sinclair as head of SIS. It is thus highly unlikely that he would have misunderstood someone else’s contribution as that of Gambier-Parry. This insight therefore does appear to confirm what Nigel West alluded to, namely SIS’s deeper involvement with interception than the authorised histories are prepared to admit.

Guy Liddell knew in March that Gambier-Parry would be taking over RSS, and he was initially optimistic about the changeover, although he recorded in his diary his concern that RSS might now concentrate on ISOS messages solely, to the detriment of MI5’s total interests. Swinton informed the Permanent Under-Secretary at the Foreign Office Alexander Cadogan on March 10, and on March 16 a meeting was held between representatives of MI5, RSS and SIS to discuss responsibilities. Liddell’s diary entry shows that Gambier-Parry was already putting his stamp on the organisation: “It was agreed that G.P. should set up two Y. masts and retain a limited number of skilled V.I.s. At present there were some 450, many of whom were useless and could be returned to store. He would have expert personnel with his Y. masts who would know the ether and be in a position to eliminate anything but the suspicious traffic. Any communication thought to be peculiar would be sent to the W. Analysis Committee and would be co-related and distributed by Cowgill’s organisation. G.P.s organisation would only be responsible for sifting in the first instance genuine traffic from the suspicious.” It seems clear that Gambier-Parry believed the interceptors themselves were capable of deciding what should be investigated, and would be authorized to do so.

In a significant move, Felix Cowgill had replaced Valentine Vivian as head of Section V in January. It was Cowgill who had objected so strongly to Worlledge’s initiative over the Morocco revelation, and for some reason he was given the task of developing a charter for the new RSS. Liddell again wrote an ominous comment on the proposal in his entry for April 10: “It seems to lay far too much emphasis on the interception of the Group traffic and to neglect the possibility of illicit transmissions in this country. We are replying in this sense.” Was someone guiding the novice Cowgill on this issue? Liddell reinforced his concerns in a conversation with Gambier-Parry on May 1, when he urged that he did not want transmissions from the UK ignored. Gambier-Parry gave a very revealing response, echoing the Gill doctrine that traffic had to be two-way, and arguing that ‘thus we have good chance of picking up traffic from abroad’. Gambier-Parry thus appeared to be set out in an unnecessarily dogmatic vein, parroting a policy that he had not crafted himself. Why would he not show greater sensitivity to his customer’s needs? Since the source of previously unidentified short-wave signals could not easily be located, why would Gambier-Parry promote a policy of diminishing efforts at direction-finding on the mainland? It was another indication that, despite the experience from the MASK exercise, non-Abwehr traffic was not going to be considered seriously. Meanwhile, the highly security-conscious Cowgill was already tightening up on the distribution of ISOS material.

The official handover occurred in early May. Gambier-Parry moved swiftly, installing a long-time friend, Major E. H. Maltby, as Controller of RSS. Liddell reported that Army Signals was taking over the responsibilities of the sniffer vans. A new interception station was set up at Hanslope Park, and some select VIs were recruited to become part of a more professional Royal Signals cadre there. Gambier-Parry dismissed Gill in an unprofessional manner, but Worlledge, contrary to some reports, was not fired immediately. He was instead effectively demoted, to work under Cowgill of Section V. Worlledge did not last long there: Dick White reported later that he resigned that summer on a matter of policy. He might have found working for Cowgill intolerable, but it is also quite possible, given his outspoken comments the previous December, that he did maintain grave concerns about the way interception policy was being diverted away from the mission that he had been attempting to execute. As for Trevor-Roper, he escaped dismissal – no doubt because he and Gambier-Parry had enjoyed hunting together with the Whaddon hounds before the war. “In the world of neurotic policemen and timid placemen who rule the secret service, he moves like Falstaff, or some figure from Balzac, if not Rabelais”, wrote the Oxford don of his comic-opera friend. Adam Sisman goes on to record that, after his appointment as head of Section VIII, “Gambier-Parry had seized an opportunity to establish his headquarters at Whaddon Hall, which was not far from Bletchley. There he lived like a colonial governor, with a fleet of camouflaged Packards at his disposal.”

Whaddon Hall in wartime

On May 20, Liddell chaired the first meeting of the Joint Wireless Committee, attended also by Malty, White, Cowgill and Frost. This was a series of fortnightly gatherings that would eventually create deep rifts between the two security services. The first resolution at this meeting ran as follows: “It was agreed that it was the function of the committee to coordinate the mutual interests of S.I.S. and the Security Service in the Radio Security Section [sic: according to Trevor-Roper, ‘Section’ was a temporary name soon abandoned]. It should lay down general directions for the operation of R.S.S. and decide priorities of service to be supplied by R.S.S. to S.I.S. and the Security Services.” It was also resolved to invite Mr. Strachey from GC&CS to become a member, and Captain Trevor-Roper was appointed Secretary. On the provocative and controversial matter of detecting domestic illicit transmissions, the minute for Item 4 read as follows: “It was agreed to proceed with a limited policy of ‘snifting’ in cases where intelligence information gave rise to a reasonable belief that an illicit transmitter existed at any known location in the British Isles. All Sections of the Security Service should be informed of the facilities available but demands should be strictly allotted to those important cases where the position of a wireless set by any individual was considered a genuine possibility. Major Frost would consult with Mr Dick White on the importance of the cases submitted, and the priority to be given to them.” The bland implication here is that some examples of illicit transmissions would be ‘unimportant’. But who would be the judge of that unless the incident were properly investigated?

That same week, at the end of May 1941, agent Sonia of Soviet Military Intelligence sent, from her lodgings in Oxfordshire, her first wireless message from British territory to her masters in Moscow.

*          *          *          *          *          *          *          *          *          *          *          *

So what evidence is there for Nigel West’s claim about SIS’s long-term ambitions to gain control over interception, and that Gambier-Parry’s Communications Section may have been assisting in its objectives? We have the clandestine operation that uncovered the Comintern spies, sponsored by SIS. Lt.-Colonel Simpson may inadvertently have helped the SIS’s cause when he brought Gambier-Parry into the picture in September 1939. That may have provoked SIS into moving on the Denmark Hill operation: the unit was transferred to GC&CS as the disputes over RSS’s future heated up in the winter of 1939-1940. We have the evidence of Admiral Godfrey, who appreciated Gambier-Parry’s valuable contribution to interception and traffic analysis in early 1940. Worlledge is outspoken on his concerns over Gambier-Parry’s and SIS’s suitability for tackling the interception problem thoroughly, and resigns on a point of policy. And SIS’s charter for RSS is oddly delegated to Major Cowgill, who is a relative newcomer to the business, has had no involvement in telecommunications, and does not work for Gambier-Parry. Moreover, Cowgill has recently taken over from Colonel Vivian, who was always bitter enemies with a man who is now his rival as second-in-command at SIS, Colonel Dansey. Dansey will be familiar to readers of Sonia’s Radio, and the most perspicacious of you will recall, from Part 9, that I pointed out an exchange of opinions between Dansey and Gambier-Parry in 1943, which showed conclusively that Dansey maintained a very active interest in clandestine wireless communications. As the saga enters the phase where SIS is in control of RSS, Liddell is soon seen to harbour grave concerns about the purity of SIS’s intentions, and Gambier-Parry gives the impression of voicing a dangerous policy crafted by someone else.  But why would SIS set out so obstructively, not accepting MI5’s requirements, or attending to their legitimate concerns?

Claude Dansey

An observer might ask at this stage: why did the JIC not take a firmer interest in all these negotiations? The committee was in fact still finding its feet after a revitalisation arising from Churchill’s accession to the premiership. MI5 and SIS were not even admitted to the committee until mid-1940, and were normally represented by Brigadiers Allen and Menzies, respectively, who might not have known exactly what was going on, or may not have been certain how much they should disclose. After all, Cavendish-Bentinck, even as Chairman of the JIC, did not know about ULTRA at this time. Yet Hinsley records that the first attempt during the war to involve the JIC in the discussion of Sigint policy and organisation foundered on Menzies’s opposition. This is an extraordinary assertion, given that Menzies, as a newcomer, presumably could not have had much clout, and he would not have been able to display his ULTRA card. As I have shown, the Y Committee, which determined interception priorities, was likewise undergoing a high degree of turmoil at the time. The whole dispersal of policy and practice for interception and intelligence gathering seems a glorious muddle, and then one remembers that glorious muddling-through is the modus operandi of liberal democracies, and the reason they thrive. Halfway through this chapter of RSS’s wartime translocation, the Conservative administration of Chamberlain had been replaced by Churchill’s coalition, with new ministers, new ideas, new appointments. There was in fact a great deal of trust and creative, open discussion between the departments, unlike the fiercely competitive agencies in Hitler’s Germany, or those cowed into indecision under Stalin, with both intelligence groups mainly telling their respective dictators what they wanted to hear.

And, finally, what about the ‘Greek tragedy’ alluded to by John Curry? We recall that this judgment appeared in the official internal history completed by Curry in 1946. Yet in his draft chapter on Illicit Interception dated October 22, 1945, Curry (who was a rather cautious and neurotic individual, as Liddell’s Diaries inform us) came to a very different conclusion. “It is nevertheless true to say that the benefits derived as a result of R.S.S. being under the control, first of the War Office, and secondly of M.I.6. were considerable and the results achieved and the benefits to intelligence work were immense. However, one is left with the feeling that had M.I.5 accepted responsibility for the organisation in 1938 a great deal of the trouble which ultimately arose between R.S.S. and M.I.5. and the ultimate change of command in 1941, would never have arisen and indeed the organisation detecting illicit wireless transmissions would have been just as good, if not better, than the one that ultimately emerged.”

That is a weak and fudgy statement that sounds as if Curry was trying to please too many audiences. Why those multiple ‘ultimates’? Is Curry referring to friction between RSS and MI5 before the ‘ultimate change of command’, or that which occurred afterwards? Was his subjective and unanalytical ‘feeling’ shared by other officers? Why did Curry alone believe that MI5 would have found the right talent and skills to sort out RSS’s house, when its own organisation was in such a mess, and short of managerial talent, and Simpson had resigned? If the SIS control turned out to be a disaster, why did he not say so?

I suspect that the ‘Greek tragedy’ conclusion may have been inserted by Petrie himself. Harrison implies (tacitly) that it might have been the Director-General who doctored Curry’s official history, since he disagreed with Curry’s conclusions, and wanted a firmer statement made on Cowgill’s obstinacies. Harrison, by the way, clearly identifies the ‘Greek tragedy’ as the withholding of ISOS material in April 1942 by Cowgill. Yet that was an Act III episode that was overcome before the finale. I have pointed out before how the circumstances of Petrie’s retirement are finessed by Andrew: I suspect Petrie had discovered some of the nasty smells that derived from a flawed interception policy when he retired in 1946. It is possible that he then realised that a deal between SIS and MI5 had already been in the works when his opinion was sought, one that effectively hamstrung him in his effort to protect the nation from the malign efforts of Soviet spies. Ensuring that his opinion of the whole affair was recorded for posterity was his swan-song.

(I am very grateful to Dr. Brian Austin, for his very helpful comments during the evolution of this article, and to Stan Ames and Bob King for their research contributions and insights. The conclusions made in it, and any errors therein, are mine alone.)

Chapter 3

The Death of a Cambridge Spy

“Warning: Some images may cause distress”  (I reproduce this warning verbatim from the folder on the Abwehr spy Willem ter Braak at the National Archives. When you read down, you will see the reason for the caution.  And the analysis behind this photograph suggests some highly controversial behaviour from British counter-intelligence in WWII, explored here in depth for the first time.)

Ter Braak’s Death Certificate

It has long been an article of faith that British Intelligence controlled all the Abwehr spy networks in Britain during World War II. On July 15, 1942, Lieutenant-Colonel Robertson, identified as B.1a in MI5, reported to the W Board that his organisation controlled all the active networks of spies originally deployed by the Abwehr in Britain. In his book The Double-Cross System, published in 1972, John Masterman boldly declared that ‘we actively ran and controlled the German espionage system in this country’. This opinion was endorsed by Sir Michael Howard in Volume 5 of the authorised History of British Intelligence in the Second World War, issued in 1990, who reported that ‘the Radio Security Service had discovered no uncontrolled agents operating’. Very soon after Robertson’s submission, Colonel J H Bevan, head of the London Controlling Section of the Chiefs of Staff Committee, attended a meeting of the XX Committee. The great deception programme could begin.

The importance of this claim has two dimensions. At the time, it was critical that no undetected network of spies could send reports back to Germany that countered the false information that was going to be transmitted by the double agents, such as a lack of substance behind the reported movements of a dummy army that would turn out to be the most critical aspect of the FORTITUDE deception campaign for the invasion of Europe. Even more critical, perhaps, it was vital that no leakage of information arrived in Germany to suggest that any spies had actually been turned. Historically, it became a matter of pride that the combination of disciplined reception procedures, rapid decisions on the viability of double agents, and interception of Abwehr radio intelligence confirming both the activities of planted agents and the acceptance of their stories, had conspired to eliminate the possibility of any agent’s survival undetected. In particular, the reputation of the Radio Security Service (RSS) for comprehensive interception of illicit transmissions was at stake. This was one of the major stories of British intelligence success in World War II.

Yet doubts occasionally surface about how watertight the processes were. It should be noted that Sir Michael Howard did not assuredly declare that no agents were operating, but merely that the RSS had not discovered any. Guy Liddell, the wartime head of counter-espionage in MI5, also suggested in the summer of 1940 that there might be a few German agents at large.  It would be useful had the authorities provided a comprehensive table of all German agents captured, and what the outcome of their detention was. Howard tells us that sixteen spies were eventually executed, of whom nine met their deaths between December 1940 and December 1941. Volume 4 of the authorised history contains several Appendices that list agents and their fates, but it is not inclusive. Masterman’s Double-Cross System provided a list of thirty-nine double agents, which contains details on each agent’s method of communication, tasks undertaken, as well as the reason for the conclusion of the case, but did not identify them by name. When the magazine After the Battle, in issue Number 11 of 1976, reprinted Masterman’s list, it added some information, such as giving real names, where it could. In his history, MI5, Nigel West extended the number to list forty-seven double agents, a roster that does not include fictitious ‘notional’ agents created to add verisimilitude and vigour to a real agent’s recruitment efforts. After the Battle also offered in-depth profiles of the sixteen who were executed, a list that is confirmed exactly by West (pp 342-343), who also lists two spies executed in Gibraltar.

But were all agents truly detected? And was the outcome solely execution or being turned? The latter dilemma was a subject of intense negotiation between MI5 and Lord Swinton of the Security Executive, who represented Churchill’s preferences. In the heat of the Battle of Britain, Churchill wanted to see more spies executed, as a signal to warn others, and as a show of efficacy to the British public. MI5 was of a more cautious bent, wanting to preserve captured agents as an element to be turned, or a source of information, although it accepted that some spies would have to be executed in order to show the Germans a (partially) successful programme of arrest and conviction, as well as a degree of ruthlessness. The Abwehr would have been perplexed if none had been captured and condemned. Yet there were risks as well. Should a turned spy turn out to be unreliable, or his usefulness to be outworn, a decision on his treatment had to be made. A public trial might expose too many secrets about the process, and if the agent had misled his controllers about his sincerity, he might constitute a serious security risk even if incarcerated. A year later, MI5 had to deal with the realities of deals made, and gone sour, in the case of SUMMER (Gösta Caroli), as Hinsley explains: “In November 1941, however, in discussions held between Swinton, MI5, the DPP and the Attorney General about SUMMER and GANDER, whose careers as double-agents had come to an end, it was agreed that no question of prosecution could arise if MI5 had used an agent or given him a promise: the risk that an agent’s double-cross work would be revealed in court had to be considered; and a promise once given had to be honoured. MI5, to avoid undue publicity, should prepare a statement to be approved by the Home Office before release to the Press through the Ministry of Information.”

It can be seen that MI5 faced a wrenching choice: if a trial and possible execution were not possible, a potentially dangerous agent (especially if he had been exposed to the Double-Cross, or XX, System) would have to be incarcerated and held incommunicado in order to preserve secrecy. Thus uncertainty rests over those agents who evaded capture for any period of time, and over those whose career ended in untidy circumstances, such as Caroli, who in fact broke his side of the bargain, as I shall explain later. The history of these individuals makes the simple conclusions of most accounts of the Double-Cross operation much more complex. This chapter discusses a few of those who fall into those categories, with special attention to the puzzling case of the Dutchman ter Braak (real name Engelbertus Fukken), who was reported to have committed suicide after parachuting into the Buckinghamshire countryside and successfully evading the authorities – including RSS – in the winter of 1940-1941.

The suspicion that the authorised accounts are not complete is reinforced by the occasional comment from Guy Liddell’s Diaries. For example, on August 21, 1940, he wrote: “H. K. BRUIN came over in guise of refugee from Holland, had wireless set he had been using to communicate weather and other information to Germans. Self-confessed agent working for Dr. Rantzau. Is this a shooting case?” Liddell indicated that BRUIN had been active for a while, since one of his goals had been to alert his bosses about British troop movements into Belgium. Yet we never learn how BRUIN was detected, whether an attempt was made to turn him, or whether he simply went to trial. It is an astonishing entry that completely undermines the clean story that has been presented since. On September 14, Liddell also notes that KUHIRT and SCHROEDER are expected to arrive, but that is the only reference to them.

One major assumption that British intelligence made was that the Abwehr was exclusively responsible for placing agents in Britain. In 1981, the journalist and historian Leonard Mosley (no relation to the fascist, Oswald) published a book title The Druid, which claimed that the Sicherheitsdienst (identified by Mosley as the SS, but an abbreviation which probably indicated the German Security Service rather than the familiar SS, the Schutzstaffel, under which the Sicherheitsdienst was originally created), dismayed by the quality of intelligence it was gaining from Admiral Canaris’s Abwehr, in May 1941 parachuted in a spy with ties to Welsh nationalists who survived the war, reporting alongside the set of turned agents. Mosley had been fed with enough leads by his contacts in intelligence to believe that the story was true, but had been hushed up. Yet any substance of truth in his account was overwhelmed by the way his informers embroidered it, and by the many fanciful touches he introduced, with the result that it is very difficult to identify reliable facts among the many fictions. In his 1998 study of bogus memoirs of espionage exploits in WWII, Counterfeit Spies, Nigel West effectively demolished Mosley’s narrative, concluding that ‘most of the book can be traced to three published sources: Ladislas Farago’s The Game of the Foxes; Masterman’s The Double-Cross System in the War of 1939-45 and Popov’s Spy Counterspy.’ [Note: Mosley was an accomplished and careful historian: I have recently read his excellent 1978 biography of the Dulles siblings John Foster, Eleanor and Allen, which also happens to contain some revealing letters to the author by Kim Philby, written from Moscow, as well as Mosley’s absorbing account of the period leading up to the Second World War, On Borrowed Time, published in 1966.]

A Tale of Two Schmidts

A last misconception that has refused to die is the account that appeared in Charles Wighton’s and Gunter Peis’s Hitler’s Spies and Saboteurs, the title used when the book appeared in the USA in 1958. We should recall that this was well before the date (1972) in which any details of the Double-Cross System were made available to the public. Wighton and Peis, claiming to have had access to the diaries of the head of Abwehr Abteilung II (Sabotage and Subversion), General Irwin Lahousen, described in detail some of the exploits of Abwehr spies in Britain. Apart from a chapter that revealed an enormous amount of detail about Arthur Owens (whom we know as SNOW), the authors laid out a convincing account of how two Danish agents had been recruited by the Abwehr in the summer of 1940, and parachuted in to Wiltshire. One, Jorgen Björnson, severely damaged his ankle on landing in a tree, while his companion, Hans Schmidt, came to earth successfully, located his injured colleague, walked into Salisbury for provisions, contacted Hamburg by wireless, and arranged through the Hamburg station for SNOW to set up a sympathetic doctor to attend to Björnson’s ankle. Björnson was soon captured and incarcerated, but Schmidt roamed free, picking up intelligence in southern England. After a breather in Wales to evade the radio monitors, whom Hamburg suspected were closing in on its agent, Schmidt resumed his espionage activity, even found work on a farm, married, and had a child, and continued transmitting his information to Germany until April 1945.

This book does not appear to have been challenged by any authority at the time. After all, despite the authors’ lack of awareness of the Double-Cross project, too many facts were close to the truth, and drawing attention to the activities of these German agents might have allowed some skeletons to escape from the closet before the authorities were ready to share their secrets. Many years later, in the issue of the magazine After the Battle referred to above, the editor and sleuth Winston G. Ramsey listed They Spied on England (the original UK title of Wighton’s and Peis’s book) as a source of information, but made no mention of Björnson and Hans Schmidt – apart from a careless but understandable error of expanding on Masterman’s list of double-cross agents by identifying TATE as Hans Schmidt, when in fact it should be Wulf Schmidt. And herein lies the key to the mystery. There was only one Schmidt.

Yet the story resurfaced in 2017. Last year Bernard O’Connor published a book titled Operation LENA and Hitler’s Plots to Blow Up Britain, an account of Abwehr incursions into British and Irish – and US  –  territory between January 1940 and the end of the war. (Operation LENA was the name given to the undertaking to infiltrate spies and saboteurs to Britain in late 1940 to prepare for and facilitate the imminent invasion of Britain by the German forces.) This volume appears to be a very thoroughly researched book, cataloguing a series of initiatives by the Abwehr to cause havoc, or gain intelligence, in Eire and Great Britain, and it is liberally sprinkled with references to the archives. O’Connor reproduces as fact, however, the whole story of Björnson and Schmidt, commenting only that ‘these two agents are not mentioned in most accounts of the German espionage service’. The author betrays some confusion, however, by maintaining only one entry for ‘Schmidt, Hans/Wulf’, identifying him as TATE, but pointing to two separate passages in his book.

The farrago can probably be explained by the fact that several episodes in the Björnson/Schmidt saga are almost identical copies of events in the Caroli/Schmidt adventure. Both teams travelled from Hamburg to Brussels via Paris, and were delayed by the weather. Both Björnson and Caroli had a dalliance with a servant girl along the way. Björnson was incapacitated in his fall – as was Caroli, whose wireless equipment knocked him out. Hans Schmidt made his way to a nearby town, as did Wulf Schmidt. Both Schmidts arranged for SNOW to meet them, but with Wulf, it was High Wycombe, not Salisbury. (Landing in Wiltshire had been the original goal of the Caroli/Schmidt airdrop, as Nicholas Mosley reported.) SNOW arranged for a sympathetic doctor to treat both Caroli and Björnson. Caroli was arrested, as was Björnson. Both Schmidts were able to roam – apparently freely  – for the remainder of the war, but they both had a clandestine meeting in London with a Japanese diplomat who provided them with more funds. The details of Hans Schmidt’s employment at a farm, and of his marriage, match Wulf’s exactly – except that Wulf had been under the control of the XX Committee.

[Note: The writer Nigel West asserts that the anecdotes about broken ankles, doctors, SNOW’s visit, etc. were invented by MI5 as a smokescreen to explain Caroli’s time of interrogation at Latchmere House.]

How this happened is easier to understand when the circumstances of the authorship are considered. ‘Charles Wighton’ never existed: it was the pen-name of one Jacques Weil, a former Swiss resistance fighter in France whose organisation was subsumed by SOE (Special Operations Executive). Under the same name, he wrote a disguised memoir of himself, titled Pin-Stripe Saboteur, in which he concealed his identity as ‘Simon’ while recounting his adventures in espionage in occupied France. (Nigel West criticises the book for suggesting that the PROSPER network in France was sacrificed for a ‘Machiavellian scheme’ to mislead the enemy about a second front, but does not indicate he knows who Wighton really was.) Gunter Peis was an Austrian journalist, who had met Lahousen after the war. Lahousen was indeed a real figure, who had very significantly provided important evidence against Goering, Ribbentrop and others at the Nuremberg Trials. He had fortuitously escaped punishment for the Stauffenberg plot on Hitler because he had been moved to the Russian Front in 1943, and been severely injured. He did indeed maintain a war diary, which was not available at the time, but, since it is now inspectable on-line, the reader can verify that its entries discuss activities at a very high level. Lahousen had far too wide an area of responsibility, in charge of sabotage (not espionage) across the whole of Europe, to know the details of operatives trained and sent out by the Hamburg Abwehrstelle.

General Lahousen

Wighton and Peis quoted some entries from the Diaries in their story, but they can now be proved to have been faked. What seems clear is that the authors must have used the existence of the Lahousen Diaries as an alibi for a largely reliable source within British Intelligence to tell a surprising story about German espionage in Britain. The source – and Wighton/Weil admits to having a high-level friend in British Intelligence towards the end of Pin-Stripe Saboteur – must have been close enough to the action (or to classified documents) to have been able to relate a sizable amount of information that was true, but which became garbled in the transfer to the author. And if that source knew about the Double-Cross System, he or she withheld that aspect of the story because of the strict embargo that had been placed on all those involved under the Official Secrets Act. (In 1976, Peis repeated some of those initial errors in his book about TATE, The Mirrors of Deception, but he still did not have access to the unreleased files at that time.) Two major conclusions can be drawn from this exercise: i) the RSS was indeed not fooled or evaded by what turned out to be an imaginary duo of Nazi agents; and ii) unreliable sources can easily be elevated to a level of authenticity that they do not deserve once they appear alongside authoritative academic references. (See OfficiallyUnreliable for more on this topic.) O’Connor’s book should be withdrawn.

The Strange Suicide of Ter Braak

The case of J. Willem ter Braak is much more alarming, however. The archival documents on ter Braak (actually classified under his real name, Fukken) would have us believe that the Dutchman parachuted into Britain successfully, was not detected, and was thus not turned, but eventually committed suicide after some months of semi-successful espionage and wireless transmission, followed by a period of rapidly increasing desperation, as his money ran out. As will be shown, this is a highly controversial story, as, if true, it would point to massive failures in security and detection at a time when Britain was supposed to be on highest alert. Yet, if it is not true, what alternative explanation could there be?

On November 3, 1940, a German parachute was found in a field near Haversham, in Buckinghamshire, but the owner was not found, and the search was apparently abandoned after a few weeks. On April 1, 1941, the body of an illegal alien was found in an air-raid shelter in Cambridge, an apparent suicide. The narrative proposed by a superficial examination of the documents in his Kew file runs as follows: MI5 was swiftly able to match the corpse with the person who had landed five months before, and, with the aid of articles found on the agent’s body, and items (including his transceiver) found in a compartment at the Left Luggage Office at Cambridge Railway Station, was able to construct the life that ter Braak had led in the interim. Having evaded capture, he had made his way to Cambridge, acquired a rental accommodation, as well as a bare office premises, and probably broadcast to his controllers in Hamburg until his batteries ran low. He had experienced problems with his food ration cards, but local officialdom had been careless. While pretending to have to leave Cambridge, he had in fact found other rental accommodation in the city, from where he made several excursions to London and to surrounding areas. Having not heard from his Abwehr bosses (possibly because he was not able to get his receiver to work), he was running short of money, and may have asked for help by communicating via traditional mail, using invisible ink and a poste restante address. Having wrapped himself up against the cold, in order to watch for help to be parachuted in that never arrived, he felt abandoned, and shot himself in despair.

What is extraordinary is how the official line has been accepted, even after the release of the files on ter Braak from Kew. For example, the German historian Monika Siedentopf, in her 2014 work, Unternehmen Seelöwe (Operation Sealion), offers one paragraph on ter Braak, merely echoing the conclusion of the authorised historian, Professor Hinsley. She does not appear to have read KV 2/114, the Ter Braak archive, however, as she provides no reference to it in her long list of TNA sources. It is quite extraordinary, given the length of time that this fugitive remained at large, how little attention has been given to him. (The records were declassified nineteen years ago.) Yet several aspects of the case merit very close inspection, namely: 1) MI5’s expectation that a parachutist would arrive; 2) ter Braak’s ability to escape initial attempts to capture him, and remain at large for several months; 3) the deductions made by MI5 concerning his wireless activity; 4) his struggles with his ration-book; 5) the evidence of ter Braak’s movements, and possible involvement in espionage and sabotage; 6) the reaction of MI5 when ter Braak’s body was found, and the subsequent cover-up; and 7) the highly controversial aspects of the victim’s ‘suicide’. I shall now explore each in depth.

Surprise?

In view of the heightened fears about invasion at the time, the recent well-publicised scare about a Fifth Column, the scars from the Battle of Britain, as well as the successful detention of several spies arriving by air and by sea, one might expect the authorities to have been better organised to handle the arrival of further enemy parachutists. Despite the Battle of Britain notionally having been won by then, Guy Liddell, head of B Branch, responsible for counter-espionage, himself wrote on November 15 of ‘one of the worst bombing raids  . . . since the beginning of the war’. The procedures for communicating and following-up on such incidents of infiltration were, moreover, well documented. And yet, when the Haversham parachute was found, the local constabulary ‘forgot’ to inform the Regional Security Liaison Officer responsible. The outcome was that ter Braak managed to escape to Cambridge, about forty miles away, by November 4, and found lodgings there. One might have expected an intense manhunt to have been ordered, but the authorities remained calm. In an almost comical twist, on November 26, three weeks later, Worlledge of RSS suggested to Frost of MI5 that bloodhounds should perhaps be used to help track down the fugitive: two days later, Frost earnestly replies that they were in fact tried, without success.

Complementary to this strange behaviour are the very revealing observations made by MI5 officers. The day that ter Braak’s parachute was found (November 3) Liddell rather drily recorded the details, which merit citation in full: “An enemy parachute landing was reported today. A complete parachute with harness overalls and flying helmet was found neatly folded and placed in a hedge beside a bridle path on Hill Far [sic], Haversham, Bucks. The parachute was wet but the clothing inside dry, and it appears that it may have been dropped during the past two or three days. Inside the parachute was a paper wrapping for chocolate made in Belgium, and a packet containing a white tablet, probably concentrated food. The packet had recently been opened and contents consumed. The parachute had without doubt been used, and the parachutist landed uninjured and is still at large. There is no trace of a crashed aircraft and the parachute was undoubtedly deliberately dropped.” (This entry does not appear in the published extracts of Liddell’s Diaries edited by Nigel West, it should be noted.)

Chocolate wrapper from Brussels found in the parachute

Liddell did not record, however, how he received this news, or how he was able to conclude that the spy had not suffered any contusions in his touching the ground. The message probably came directly from T. A. (‘Tar’) Robertson, then working for Major Frost of W Branch (sometimes called Section W, which had recently been subsumed into B Division), whom the local constabulary had contacted, via the Special Branch. Readers should bear in mind that, as Masterman’s account makes clear (p 100), while the authorised history of MI5 assuredly does not, that the famed B.1a section responsible for managing double-agents, led by Robertson, was not created until June 1941, long after the last LENA agent had landed. (After a showdown with Frost, when Robertson had complained to Liddell that he could no longer work with him, Liddell made the decision on December 12 to transplant Robertson ‘and all his minions’ out of W Branch into an established unit of B Division.) Yet Liddell never questioned why the relevant RSLO (one of the Regional Security Liaison Officers, namely B Division representatives dispersed around Britain, and first in line to investigate possible spy threats) was not informed, or why the message had not arrived through the proper channels. And he never referred to the case again in his diaries until ter Braak’s body was found, an extraordinary example of ‘the dog not barking in the night-time’. Why would it not be of supreme importance to him, and for his chronicle, to record how the hunt for the fugitive progressed?

It is also noteworthy that the proceedings of meetings of the RSLOs and the Security Executive (the supervisory body installed by Churchill to manage domestic intelligence) in the period October 1940 to April 1941 focus almost exclusively on what the procedures should be for processing captured agents, not on how resources should be deployed to tracking down undetected agents whose traces of arrival have been found. It is very telling that, at a meeting a few weeks before ter Braak’s body was found, the RSLOs engineered a change in the communications procedures, which now required that the Police Constables first inform them, who would in turn let W Branch know of a suspected agent. What is also intriguing is that the notes supplied for White’s address at the meeting (on February 18) specifically refer to ‘the Haversham parachutist’, indicating that he was still at large, but nowhere does this highly provocative state of affairs appear to engage the attention of the participants.

What should have been fresh in the minds of the RSLOs, however, was the case of SUMMER (Gösta Caroli), a supposedly turned agent who had tried, on January 13, to kill his guard and flee the country, being arrested at Ely, not far from Cambridge. (I have not yet discovered any record of what they were told of the affair.) Masterman later hinted that the RSLOs must have been in the know, since he wrote that SUMMER was apprehended ‘after some anxious hours in which we had been compelled to warn the appropriate authorities over half England to set a watch for the fugitive’ – a pattern of behaviour in marked contrast to the lethargy over ter Braak. Masterman went on to write: “His escape, had it succeeded, would indeed have wrecked all our schemes, but as things were no harm was done – not even to the strangled guard, who was the richer for a stimulating experience (and a good story) at the expense of some small temporary inconvenience.” Masterman’s levity was misplaced: if the guard had indeed talked carelessly about the event, the outcome might have been just as calamitous. I shall return to SUMMER’s fate later.

A full chapter – a thesis, even – could be written on this aspect of British security measures at the time. One of the most unprofessional, almost scandalous, aspects of the affair is that the role of Malcolm Frost and W Branch has been completely excised from the authorised history, as if Robertson always worked for the not yet created B1a. Neither Malcom Frost nor W Branch (nor even the RSLO organisation) appears in the index of Sir Christopher Andrew’s authorised history of MI5, Defending the Realm. Yet, at a time (November 1940) when Frost was being scorned by such as Swinton for his ego and his ambition, and he was apparently driving Robertson to distraction, the Director of Military Intelligence, Beaumont-Nesbitt, was writing to Frost, almost as an equal, to suggest that the military authorities should be given the responsibility for handling suspected spies after their arrest. It was a very puzzling relationship.

The focus returns to Liddell, since his Diaries are such a central source of the story. In the middle of November, instead of instantly organising pursuit of the dangerous quarry, he started plans for the formation of the famous XX Committee, and then had to deal with the sacking of Jane Archer, which in his journals he ascribed to her extended derisory comments about the MI5 head-in-waiting, Jasper Harker. (In April 1940, Jane Archer had been taken out of her vital role as lead in Soviet counter-espionage to design the RSLO group, and then manage the team of RSLOs.) Yet the timetable offered by the ter Braak archive lays open a completely new interpretation. Had Jane Archer perhaps challenged Liddell’s methods of undermining her authority through Section W’s continued bypassing of the RSLOs, who seem to have respected her skills very highly? Archer knew that Liddell could not be held totally responsible for the dysfunction in MI5, as an ungainly organisation had been forced upon the service by Swinton. In August 1940, however, Frost’s W Branch had been moved under Liddell’s ‘B’ organisation, and tensions between Frost’s group and the rest of ‘B’ were slow to be resolved.

The presumed cause for Archer’s sacking comes solely from Liddell himself, but in his Diary (in a passage also not published by Nigel West) he refers to a contentious conference on October 31 between himself, Archer and Frost. It would not be surprising if the highly capable Archer, perhaps still smarting about her removal from Communist work, had in fact challenged Liddell quite robustly over the way the whole RSLO infrastructure was being undermined, and over Liddell’s inability to control Frost, and that she thus forced a rupture. The timing of this meeting occurs just before ter Braak’s arrival: if Liddell was expecting him, as he admitted, could it be that he and Archer disagreed about a level of secrecy required in dealing with the Abwehr spy, that would have entailed excluding the RSLOs? I have found no documentary evidence of this, but it seems highly possible that he could not convince her otherwise, leading to a dissension that brought on her forced resignation. Liddell may then have made recourse to an external enduring motif to explain her dismissal. In addition, Frost was certainly a problem: on December 6, Robertson declared he could no longer work for him, and Liddell, who had been so enthusiastic about his ‘man from the B.B.C.’ a few months before, must have been having second thoughts.

Thus, after the heated events of the winter, and the arrival of David Petrie (who, after a few weeks of intense analysis, was officially appointed as the new Director-General in mid-February 1941) to shake things up, it is with some degree of astonishment that the ter Braak case suddenly reappears in Liddell’s Diary. We read in his entry for April 1, 1941, that he has, the same day the body was found, already been informed of the discovery, and of the verdict of suicide (before any inquest, one should note). The local police force had very speedily undertaken their investigations, being able to inform Liddell that ‘he had lived in Cambridge for about four months’ (an incorrect calculation, as it happens), having arrived with a small suitcase and parcel. Liddell concluded his entry as follows: “On form I should say he was undoubtedly a parachutist, and probably one whom we expected at that time.” (In a more public forum, however, he was much more guarded: at a meeting with the RSLOs in Oxford a few days later, he merely commented that the person discovered was ‘probably’ the parachutist from November.)

So why would Liddell not have pointed out that fact – that ter Braak had been expected – when the parachute was found? And why might MI5 have expected his arrival? Ter Braak was, according to most accounts, not in fact linked with the Operation LENA spies (e.g. Caroli and Wulf Schmidt), so the story has been encouraged that MI5 would not have learned about his role from Caroli (SUMMER) or Wulf Schmidt (TATE). After the war, in June 1946, when MI5 interrogated Abwehr officials, and showed them ter Braak’s photograph, neither Nikolaus Ritter (head of Abwehr I in Hamburg), nor Jules Boeckel (who replaced Ritter in March 1941) could identify who he was. Ritter, who denied that ter Braak was a LENA agent perhaps a bit too assertively, said his wireless apparatus was not designed for summer use (unlike the LENA agents), and Boeckel suggested that he might have been managed by Brussels, which would make excellent sense, given the origin of the chocolate wrapping that Liddell had referred to. (Though one has to question such an obvious clue: why would an ostentatious wrapper be packed into the parachute, and were parachutes not supposed to be buried or burned on arrival, rather than being ‘neatly folded’?) Thus two scenarios present themselves:  1) The RSS had picked up Abwehr messages, with the resulting decrypts pointing to an imminent drop by the Brussels station. (I have not yet been able to inspect the detailed source records at Kew. And we recall Liddell’s comments about KUHIRT and SCHROEDER.) 2) Ter Braak was indeed a LENA agent, and Liddell had been warned of his arrival by TATE, but for some reason all the Abwehr officers after the war wanted to disown any connection with him. That would not be surprising if the Churchill assassination plot (see below) were indeed true. Abwehr officers, in any case, gained a well-deserved reputation for not speaking, under interrogation, openly and directly about their wartime experiences.

Finally, we have the evidence of Robertson. On November 15, this gallant officer had had to apologise to Michael Ryde, the RSLO in Reading, stating that the blame for the failure in communications lay with the Chief Constable of Buckinghamshire, who should have reported the matter to the RSLO rather than alerting adjacent police forces. The same day, Robertson wrote to Ronald Haylor, the RSLO for the adjacent region of Nottingham, to make a similar apology, explaining that he did not hear about the parachutist until November 3 (i.e. the same day that Liddell learned of it), but enigmatically added that he did not recommend ‘using the parachutist for our own purposes, so I think it would be advisable to lay on as wide a hunt as is possible’. Is this not an extraordinary careless way of representing MI5’s intentions, and expressing the necessity of speeding up isolation of the spy? First of all, Robertson obliquely admitted that MI5 had ulterior purposes in considering the exploitation of parachutists (for the emerging Double-Cross System), and implicitly that the RSLOs knew of this project, but at the same time indicated that efforts to track down spies could be restrained – and no doubt were being held back – in the cause of presumed monitoring of their activities. He very significantly echoed this policy when, after Jakobs had parachuted in on January 31, 1941, Liddell noted in his Diary that Robertson wanted to give him [Jakobs] a run ‘in order that we may find out exactly how much these people can ascertain if they are left to themselves’.

Yet it was a very risky and slapdash judgment to make about the potential of a double agent when MI5 had not even identified or interrogated the suspect, and did not know how dedicated a Nazi he was. As soon as every agent was captured, he or she should have been sent to Camp 020 at Latchmere House in Ham, for interrogation by Colonel R. W. G. Stevens, but such a consideration appeared far from the minds of Liddell and Robertson in this instance. On November 18, Robertson wrote two more significant letters. The first was to Colonel Wethered, the RSLO for Birmingham, following up on a telephone call with him the previous day. It is clear from this long letter that Wethered knew nothing about the parachutist before then. The RSLOs had not been informed. The same day, Robertson had an important insight, showing all the wiliness of the veteran officer’s knowledge about fugitives, and wrote to Worlledge of RSS in the following terms: “This brings one to the conclusion that the parachutist, if he is at large, must have gone to some hiding-place.” Indeed, for that is what fugitives do the world over – unless they are allowed to survive in broad daylight. Suddenly, a pattern for ter Braak’s bewildering ability to stay on the run emerges.

Ter Braak’s Wireless

What is critical to this story is the use of wireless: ter Braak had a working set, and RSS was supposed to be supremely well-equipped to deal with incidences of illicit transmissions from within the nation’s borders (see earlier Chapters 1 and 2 in this saga). Thus an inspection of what RSS and related units did or did not do, and how MI5 responded to their actions, is highly important. The analysis by MI5 officers of ter Braak’s wireless activity can be seen to take place in three stages. There is an initial assessment necessary to provide the political cover for informing the government of what happened, completed by May 1941. That is followed by a deeper internal review later that year. The last stage occurs after the war, when Abwehr officers have been interrogated about the case.

After the wireless apparatus had been discovered in the left luggage office on the day after the corpse was found, Herbert Hart (B2.b, the Oxford academic, who the following October was to marry Jenifer Fischer-Williams, the Soviet spy and probable betrayer of Walter Krivitsky to Guy Burgess), on April 11 sent it to the SIS for examination. Marriott (B2.a) recorded on May 6 that he had received the report from SIS, which had, strangely, been ‘missing’ for a while. The report included the statement that the apparatus was ‘identical in design to one taken from enemy agents off coast of Scotland, though not so complete.’ It added: ‘The separate H.T. and L.T. battery completely run down [sic], while H.T. batteries in suitcase show 195 volts instead of normal 270.’ The conclusion was that the set had been used considerably

Hart’s official May 7 report to Dick White, head of B.1, responsible for ‘Espionage’, offers a different conclusion, however: “Ter BRAAK’s wireless transmitting set has been examined and the expert opinion is that it had probably been used in the effort to establish contact, but it is impossible to say whether the effort was successful or not.” This last observation was an embellishment by Hart, as SIS ventured no opinion on that unverifiable truth. Moreover, that was not the only opinion Hart inserted. He also described the set as follows, saying that the search at Cambridge Railway Station revealed ‘a brown Moroccan leather case containing a W/T set similar in all respects to sets brought to this country by enemy agents who had been dropped by parachute and captured’. This statement is false in many respects. There had been only four admitted enemy agents who had been dropped by parachute – TATE, SUMMER, GIRAFFE (Graf) and GANDER (Geysen). The last three brought with them a device that contained a transmitter only. The SIS report referred to the three agents who had been arrested in Scotland on September 30, Drucke, Walti, and Eriksen, but they had not arrived by parachute. And there is a combination of naivety and excessive detail in Hart’s report as well. He offers a conclusion that ‘there are strong grounds for thinking that Ter BRAAK was, in fact, the parachutist whose parachute was discovered on 3.11.40 at 12.00 hrs at Hill Farm, Haversham, Bucks.’, when, as has been shown above, Liddell had come firmly to that conclusion the day that ter Braak’s body was found. Hart also provides some details about ter Braak’s ‘self-inflicted wound’ that should perhaps have been kept under wraps.

A reason that this report might have been considered unsuitable for wider dissemination is that, while White’s forwarding of it appears in the ter Braak files at TNA (KV 2/114), Hart’s report is absent. It can, however, be found in a folder concerned with ‘immobilisation and arrest of enemy agents’ (KV 4/406). Thus Hart’s misrepresentations might have been deemed unsuitable for inclusion, as they both contradicted facts elsewhere, and told too much: his memorandum somehow managed to escape to another file. Were Liddell and White withholding information from Hart, and letting him blunder on in the dark? That would be highly unlikely, given Hart’s deep involvement in the case, and the sensitivity of his mission. It seems much more probable that he was guided to write a report that emphasized the similarity between ter Braak’s case and those of other agents, and that for some reason MI5 wanted ter Braak’s possible wireless activity to be minimized. After all, it would not help much if the report – to be sent to all the RSLOs – indicated that RSS had not been succeeded in performing its job properly.

MI5 picked up the investigation again in the autumn. Hart had provided some details about ter Braak’s movements with suitcases, gathered from his landladies, that indicated that the agent had concealed the wireless set from both of them. By this time MI5 officers have started to theorise more profoundly. Now under the new leadership of Petrie, a fresh organisation is in place: head of B1.a is ‘Robertson, ‘Special Agents’, and Hart is now alongside him as B1.b ‘Espionage Special Sources’. (Dick White is now head of B.1, while Frost now reports to Liddell as head of B.3, responsible overall for ‘Communications’, which includes B.3A, ‘Censorship and Reception Analysis’, B.3B, ‘Illicit Wireless Investigation and Liaison with RSS’, and B.3C, ‘Lights and Pigeons’.) On September 10, Gwyer in B1.a issued a long report. It informs us that, during both his periods of lodging, from early November to late March, ter Braak never left his accommodation overnight. He had rented an office above the agents who had found his rental properties, but apparently visited it only two or three times, and was never seen taking a suitcase with him when he left in the mornings. Gwyer’s conclusion was that he could not accept the view (apparently promoted by Fl. Lieutenant Cholmondeley of B.2a, the officer who later cooked up the Operation Mincemeat scheme with Ewen Montagu) that ter Braak failed to communicate by wireless, pointing out that he had one aerial suitable for night-time transmission. (And the SIS report indicated that he was given two crystals, and thus had two frequencies to use, the lower one being necessary for night-time use.) “We know that he was only equipped with one aerial which was suitable for the frequency he was instructed to use at night, but we also know  . . . that the only time that he was alone with his wireless set, apparently, was at night. He could not have transmitted during the day, as he could not have left lodgings with a wireless set in a suitcase without somebody noticing this fact. Equally, he could not have used his office for transmissions if he was there so infrequently.”

Gwyer owned up to a high degree of precision in his estimates. He stated that ter Braak’s batteries, ‘exhausted at time of his death’ would not have lasted more than two months, although how he came to that conclusion without knowing how frequently, or for how long, the apparatus had been used is not explained. Ter Braak was able to replace his low tension battery, but not the high tension one, and thus had to purchase an accumulator. His conclusion was that ter Braak had probably transmitted successfully up until about Christmas 1940, after which he communicated in secret ink, using a poste restante address. Gwyer assumed that he had asked for a new battery through this medium, and was expecting delivery from a new agent. When none appeared, he killed himself. Gwyer’s colleague, R. T. Reed, added that ‘contacts during the day would have been impossible, because of aerial provided and difficult to avoid suspicion or capture if seen’, but expert judgment (in the person of Dr. Brian Austin) has informed me that adapting an aerial for the night-time frequency should have been a straightforward task. In Dr. Austin’s words: “Ter Braak’s single piece of aerial wire, of some unknown length, could have been made to work at maximum efficiency on both the frequencies allocated to him.”

Perhaps ter Braak had not been properly trained for such adaptations. In his interrogation after the war, Lahousen (the officer who headed Sabotage and Subversion) criticised the Abwehr’s radio ‘expert’, Rasehorn, for being ‘not very well qualified’, asserting that ‘his wireless connexions did not work’. Thus there could have been deficiencies in understanding in both Abwehr and MI5 camps. Indeed Reed’s expertise must also be questioned. On July 24, he had written a report on wireless telegraphy which stated: “I tested the set of ter BRAAK in communication with Radlett. The aerial that he was given is satisfactory on 4508 kcs, but will not work effectively on 5435 kcs., which is his day frequency. According to the file of Ter BRAAK, it would seem that he never tried to get into communication during the night time and only tried during the day time.” Did Reed adjust the aerial? It is not clear. Yet how did anyone know about ter Braak’s failure, since other testimony points to the fact that ter Braak did not even take his set out during the day? The evidence provided elsewhere is that ter Braak could have communicated only in the night-time.

It is only at this stage that the RSS appears to become seriously involved, which is quite astonishing. Back in November 1940, RSS was certainly deploying its interception capabilities. A report from W.2 (Robertson, working for Frost, whose main interest at that time was investigating ground communications with aircraft) on November 5 declares that the unit was hunting the spy, with promising indications. “From the intercepts produced by R.S.S., it seems likely that there is at least one wireless set being operated in the country. Every effort is of course [sic] being made to locate this.” Dr. Austin has estimated that, at the frequencies used by ter Braak, a successful Direction-Finding station would have had to be as close as 6 miles away for his ground wave to be picked up – a calculation that might suggest that Lt.-Col. Simpson’s planned dispersions (see Chapter 2) were not dense enough. The nearest station appears to have been one at Steeple Bumstead, about twenty miles south of Cambridge, but, with a known enemy agent on the loose in the area, one might perhaps expect a more flexible campaign to have been undertaken to track down the suspect.

The same day (November 5), Robertson reports that all the intercepts appearing in RSS’s weekly report have been identified. And a week later, Hinchley-Cooke in B.13 receives a report from Robertson that ‘so far neither the Bucks Police nor R.S.S. have been able to find any trace of the man.’ One could reflect that the parachutist might not have considered helping the authorities by remaining in the small county of Buckinghamshire, but the current record then goes quiet. Did MI5 not really care what happened after that? So, much later, during this post mortem, when Robertson asks Hughes of B3.b (the liaison between MI5 and RSS), on September 11, 1941, to enquire of RSS ‘whether any station heard on 4508 kcs or possibly 5435, in period November 4 1940 to January 31 1941, between 2100 and 0900 GMT using a 3 letter call of the type others used having a circular code, i.e. LNP, GIK, etc’, one’s first reaction is, even if the exact frequencies and callsigns were not known then, why would such an intense action have not been performed at the time of the hunt? And why would Robertson and Hughes not ask whether any traffic had been intercepted from Hamburg during the time of ter Braak’s fugitive status?

The description of the exchange that followed deserves quoting in full. Thus Hughes to Robertson on September 14, 1941: “With regard to your note of the 11th Sept. concerning the possible workings in connection with the Cambridge ‘suicide’ case, the Controller, R.S.S. replies that they are doing what they can with extracted records, but that the original traffic of that period has been destroyed. It is not possible for RSS to keep all original logs for any great length of time as the volume would be entirely unmanageable, so I am afraid there are not likely to be any useful results of this inquiry. Is there any other line of attack you would like me to suggest to RSS? RSS are hoping to be able to make records of any QZZ traffic which they pick up during this week. Only part of the text is likely to be recorded, but I think this will be satisfactory for your purpose.” And two days later, Reed writes: “I spoke with Major Morton Evans yesterday and asked if it were possible to keep a careful radio check on the East-Anglia area. We decided that this was technically impracticable. The only line to take was that if any station was suspected of working in the East Anglia area – by the results of the direction finding stations – special attention should be paid to it. Should any such station be intercepted, Major Morton-Evans will see that this is done.”

Group of ex-RSS members and supporters, Bletchley Park, 2013
Bob King is seated, far right. Stanley Ames and Dr. Brian Austin, standing, far left.

Over the years, this puzzling interchange has tested many members of RSS. At one of the assemblies of ex-RSS interceptors and officers, at which enthusiasts also take part, in 2015, with ter Braak’s records then open at last, a group discussed how RSS could have avoided picking up his signals. After all, they should have had enough density of voluntary interceptors to simply pick up the groundwaves. Bob King, who had been working at Arkley, the location to which all logs were sent, at the time, offered the following possible reasons (which I reproduce here in their original form):

  1. He did not transmit.
  2. If he did his procedure was not suspicious.
  3. His calls were very short and with a non-resonant aerial very weak.
  4. No one was listening at the right time on the right frequency. (Not so likely).
  5. He was heard and his logs arrived at Arkley but not fitting any group was marked in the books as ‘Suspect’ and awaiting further reports.
  6. He was reported to Arkley, and log readers were told to ignore [signals] believing it to be an agent already covered.

What Bob King overlooked in this analysis was the detection of signals from ter Braak’s controllers in Hamburg. According to the confident doctrine of Walter Gill, whose report was written just after ter Braak arrived, the absence of agents in the UK could be determined by the fact that no incoming transmissions were intercepted by the RSS’s major stations. Yet, if ter Braak had a receiver, messages would surely have been sent to him – at least in the first few weeks. Thus, one must conclude that RSS picked up those signals (even if ter Braak was unable to), but that MI5 chose not to act on that discovery. In light of the fact that MI5’s leaders admitted that an unidentified German agent had parachuted in with a wireless apparatus of some kind, Gill’s policy appears at best bizarre, and at worst, simply irresponsible.

Interrogations with Abwehr officers after the war did not disclose much else, although now new officers in B1.a were still asking the same questions. Joan Paine of B1.a in September 1946 still acts as if she does not know whether ter Braak and the Buckinghamshire parachutist are the same person, which makes it appear as if she had been deliberately kept in the dark. Her colleague Warrec expresses the same uncertainty. But the Abwehr officers either do not know any more, or pretend they do not. Karl Krazer, of the Hamburg unit, asserted that the last LENA mission took place in September 1940, and thus ter Braak could not have been part of it. Moreover, he echoes Ritter’s claim that LENA agents all had ‘summer-time’ transmitters, unlike ter Braak, thus providing an opinion contrary to what Hart had said. Both Krazer and Richter recommended speaking to Sensburg of Brussels. The file on Sensburg is dominated by his time as head of the Athens Abwehrstelle, whither he was transferred from Brussels in May 1941. He was surely still in charge when ter Braak’s crisis occurred.

The documents also confirm that a Brussels-based agent named Fackenheim was close to being parachuted into Britain, but was confounded by the weather. Yet Fackenheim was returned to Germany in the autumn of 1940, so he could not have been around in April 1941. Sensburg also gave final instructions to Waldberg, Kieboom, Pons and Meier, who were sent to the Kent Coast by trawler early in September, all of whom except Pons being subsequently executed, so his role in handling ter Braak sounds very probable, even though Sensburg does not list ter Braak as a member of the LENA operation. (Interestingly, Jakobs and Richter, both of whom were sent out later than ter Braak, are listed as LENA agents by Sensburg, thus contradicting Krazer. Again, these Abwehr officers may not have been entirely honest.) But the opportunity to interrogate Sensburg on that matter seems to have been missed. A few extracts from ULTRA intercepts describe some of his actions, but nothing concerning spies sent to Britain. Perhaps senior MI5 officers were hoping the whole issue would die a gradual death.

Ter Braak’s Mission and Movements

So what was ter Braak doing in Cambridge? It might seem an unlikely city to conduct espionage, although several military airfields were in reasonably close reach. One theory, promoted by Winston G. Ramsey in After the Battle, suggests that the spy was sent, by SS General Walter Schreckenbach, on a special mission to assassinate Churchill, and it describes how ter Braak was trailed, his lodgings searched while he was out, and found to contain ‘a wireless transmitter, detonators, a Luger pistol, a file on Mr. Churchill’s movements and three crudely forged Dutch passports’. The writer provides no source for this fantasy, which sounds like a crudely conceived smokescreen. (The assassination plot is echoed by Charles Whiting, in his 2000 book Hitler’s Secret War, though whether he used Ramsey as a source, or shared Ramsey’s informant, is not clear. Whiting is of the school that considers that ter Braak was indeed one of Ritter’s LENA agents, while he repeats Ramsey’s more credible account of how ter Braak’s body was found. He does not list Nikolaus Ritter in his Acknowledgments, but it is evident from his text that Ritter became a friend after the war, and was a major source of his information, thus undermining Ritter’s testimony to his interrogators.) It would have been highly unlikely for the SS to select an Abwehr operative for such a scheme; it would not be good practice to try to mix sabotage/destruction with espionage (although it did sometimes happen), as sabotage is noisy, and gains attention, while espionage is quiet and clandestine. An assassin would not need a wireless, just a rapid means of escape. The pistol found by ter Braak’s body was a Browning, not a Luger. And why ter Braak would languish in Cambridge for five months if charged with such a mission cannot be answered.

Ter Braak’s passport

Another explanation, which surfaced when a flurry of memories was published in the Cambridge Evening News in January and February 1975, following the After the Battle investigations, was that ter Braak had come to blackmail prominent émigré academics living in England into cooperating with the enemy, presumably by threatening their relatives. “One person on ter Braak’s list”, the account read, “was professor of Law at Clare College, Kurt Lipstein, who had left relatives behind in Germany when he came to England in 1936”. But the leads appear not to have been followed up. A perhaps more convincing explanation was that ter Braak was guiding German aircraft to military targets in the area, and that that ‘Army Intelligence knew that signals were being given during raids and troop movements in the Cherry Hinton area’ (which was part of Cambridge). A major example provided was a raid by Dornier 17s as tanks were being unloaded in the Cambridge marshalling yard, with eleven fatalities as a result. Yet the date given was February 24, 1941, when ter Braak’s radio was supposed to be no longer working. Can we rely on that? Or was he indeed still broadcasting at that time, and was that a fact that MI5 chose to conceal when it realised it might have had blood on its hands by allowing him to remain active? Did MI5 then attempt retrospectively to ‘silence’ ter Braak from December?

The National Registration Card of ter Braak

Yet Cambridge was indeed his objective. The false and very clumsy identity-card he was provided with in Brussels (see photograph) gives a non-existent address of ‘Oxford Street, Cambridge’. It has one person’s handwriting in places where that of an official and the owner should appear. The number of the residence was entered after the name of the street, in continental style. The Christian name appeared, wrongly, before the surname, and the card was machine-folded, not hand-folded. (These failures were provocatively used at RSLO training in February 1941.) Thus we have to face the facts that the spy apparently landed on October 31 or November 1 (since enemy aircraft were spotted both those nights, and the parachute was wet from rains since) and was able to extricate himself from his parachute, conceal it neatly, pick up his two suitcases, and somehow walk the forty miles from Milton Keynes without being spotted, or gain transport from some unsuspecting or abetting agency, not reaching Cambridge until November 4. The parachute was found at noon on November 3: this was at a time when the head of counter-intelligence at MI5 declared the service was expecting him, and a ‘thorough search of all woods, buildings’ was in process and that ‘enquiries are being made at all shops, cafes, hotels, railway stations, etc.’. One report, from November 12, says that the parachute was found near the house of a man who was under suspicion, but no more is said of this gentleman, or whether he was able to help ter Braak on his journey to Cambridge. Thus we have to conclude that Ter Braak either arrived in Cambridge with at least three days’ stubble on his chin, and a wet overcoat, having slept in the open, with a suitcase on either arm, yet failed to provoke any attention, or found shelter with some sympathizer. He then successfully installed himself with Mr. and Mrs. Sennitt of 58 St Barnabas Road, Cambridge, after gaining their address from a  rental agency.

Ter Braak’s movements overall were erratic, and showed no focused pattern of activity. They were partly re-creatable because the spy rather enigmatically kept all his used bus-tickets, and thus his journeys around the countryside, and once to London, could be traced. But, since he never spent a night away from Cambridge until his final foray into the night, they are not really germane to the story. MI5 knew, from interrogations of other agents, that reporting on the success of bombing-raids was a major part of their mission, so it is likely that ter Braak was involved in such activity. What is far more intriguing is how he managed to deceive the authorities for so long. Most of the recorded saga refers to his travails with his ration-book, but the failure of everyone (rental agency, or landlord and landlady) to make even a cursory inspection of his identity card, is dumbfounding. What is noteworthy about the card, apart from its obvious forgery to the eye, was the fact that it contained a serial number that had been provided to the Abwehr by SNOW, thus showing that, even if the Hamburg Abwehrstelle did not know about ter Braak, Nikolaus Ritter was clearly passing on seemingly valid numbers to be shared by the Brussels Abwehrstelle. Evidence of such cooperation could of course reinforce the theory that ter Braak was indeed one of the LENA team. Only when the Food Office in Feltham looked up the ID card number supplied by ter Braak did it realise, very tardily, that the number was one issued to a Mr. Burton. When challenged on explaining this, ter Braak panicked and left his lodgings, saying he was leaving Cambridge, but in fact he only moved to a different accommodation in the city.

What occurred with ter Braak’s National Registration Card and ration-book would come back to haunt MI5. The card found on ter Braak’s body had the number BFAB 318-1 (see image). When ter Braak’s landlord, Mr. Sennitt, dutifully went to see the Assistant Aliens Officer of the Cambridge Borough Police after ter Braak’s arrival, in order to report the Dutchman’s presence, the officer told Sennitt to get hold of a copy book and get the Dutchman to write his name, and date of arrival in it. As Hart reported: ‘the Assistant Aliens Officer concluded his remarks to Mr. Sennitt by saying: “Don’t you worry, the fellow will be along shortly soon to report himself.”’ Hart added that ‘ter Braak, of course, did nothing of the sort and nothing further was done by the Assistant Aliens Officer’. That was quite an extraordinary oversight by the Officer, but why Sennitt – or even the rental agency, Haslop & Co. – had not thought to ask for the alien’s identity papers on first encountering him, is also worthy of comment. Were all regional towns, cities and their establishments not on high alert for detection of an alien parachutist? Experiences with other dubious-looking strangers, such as the three spies who landed in Scotland by boat (Drucke, Walti, and Eriksen – see above), show that inspection of ID cards was the first task to be undertaken.

That hurdle passed, ter Braak had to provide a ration-book to his landlady, so she could buy provisions for him. Yet the forged one he supplied was seen to have expired in July 1940. Here the narrative diverges. Hart’s memorandum indicates that that fact was ‘soon’ discovered by ter Braak, although the word ‘soon’ is enigmatically written in by hand, replacing the word ‘not’, a paradox, and maybe a subconscious error, that Hart did not attempt to explain. The hidden notion may be revealed in the report by the Cambridge Borough Police, dated April 24, 1941, where the officer wrote that Mr. Sennitt agreed to visit the Food Office in Cambridge to obtain an emergency ration card, while ter Braak wrote off for a new one. Again, why no one asked how ter Braak had managed to survive beforehand with an out-of-date ration book is a question unraised by all concerned. It should be recalled that November 1940 was the peak of the Blitz, when Hitler was trying to strangle Britain to death: nearly 7,000 civilians died from German bombing that month, and offences against the use of ration-books were publicly listed to deter abuse.

Here again, Hart distorted the truth. In his memorandum, he wrote that ter Braak applied on November 28, 1940 to the Food Office appropriate to the number on his Identity Card (i.e. Feltham Food Office) for a new book, ‘explaining away the fact that his book had become out of date and had not been used on the ground that he had been living in Cafes and Hotels’. Yet the only address on the Identity card was ‘Cambridge, 7 Oxford Street’, and MI5 officer Gwyer’s report of September 10, 1941 expressed puzzlement as to how ter Braak knew to write to an office in Feltham. “He himself never visited the Food Office, his emergency cards being drawn by his landlord. It is true that the Food Office may have told the landlord which was the correct office for ter Braak to apply to, but the Police Report of April 21st states ‘so far as Mrs. Sennitt (the landlady) is aware, Braak sent the old book to either Cambridge Street Road or Terrace, London’”. The issue is mysteriously left unresolved, though later in his report Gwyer refers to Mrs. Sennitt as ‘an incompetent witness’. Mrs Sennitt persisted in questioning ter Braak each week about the receipt of his new ration book, and each time, on receiving a negative answer, she would acquire another temporary one.

Ter Braak’s application for a new ration book

I reproduce ter Braak’s application here: the date is clear; a Food Officer has entered the wording of ter Braak’s explanation for survival without a book, and also entered his National Registration number, as we can confirm. The Ration Book number is printed – CA 567132. (Hart mistakenly listed it as CA 567123, which might explain some later confusion. The Cambridge Borough Police had by April 4, 1941 verified that that number had been issued by the Feltham Food Office to one William Widhers, a civil servant employed by the Prison Commissioners. This was the same number that had been provided on August 9, 1940, to MI5, in the name of Burton, and was passed on, nominally through SNOW, as a safe number for the Abwehr to use. MI5 then tried to contact Burton to clear up the duplication, but he was away, and the error had to be ascribed to ‘mistakes by the registration authorities’.) The timing after the receipt of ter Braak’s submission is not precise in Hart’s account: he wrote that, ‘on receipt of ter Braak’s application,’ Feltham Food Office looked up the number of the ID, found that it had been issued to a Mr. Burton, and thus sent the official form R.G. 32 to ter Braak, asking for full particulars. At some stage ter Braak must have sent in his R.G. 32, as a letter from Feltham, dated January 25, asks him to confirm that he no longer wants a new book, and that the previous request was a mistake.  Ter Braak’s reply of January 28, 1941 can be seen in the accompanying image. Either the Feltham office had been very sluggish in sending out the original R.G.32, or it had been very lenient in not demanding the prompt return of the form. Again, no questions appear to have been raised about the efficiency of the process, or the lack of consultation between the Feltham and Cambridge Food Offices.

One can now understand why ter Braak probably relaxed up until the end of January. He had followed the protocols, submitted his application, continued to receive temporary ration books (though to the consternation and amazement of his landlady), and could carry on with whatever he was doing. Yet the lie he gives in his letter about the Dutch Emergency Committee, and its supplying him with a new ration book ‘a few weeks ago’ with the result that he no longer needs a new one, does betray some desperation. Moreover, the normal number of weekly emergency ration books that was allowed to be issued was six: the Food Office in Cambridge had been lax enough to extend it to twelve in ter Braak’s case. Eventually its patience ran out – on January 30, just before ter Braak’s letter was received in Feltham, The Cambridge Office informed Mrs Sennitt that ter Braak was required to pay a visit to the Food Officer.

At this stage, ter Braak panicked. “He appeared agitated’, said Mrs Sennitt later. He told his landlady that he had to go to London, and would have to quit his premises. Only then did Mrs Sennitt notice his second suitcase (containing his wireless set), which he put in the taxi taking him to the railway station. Yet, by a remarkable coincidence, Mr Sennitt, who suddenly realized that he had to follow him to regain the front-door key, could not find ter Braak when he explored all the carriages in the London train waiting to leave. And a few days later, Mrs Sennitt bumped into ter Braak on the street in Cambridge, at which he explained that he had had to return for a few days.

In fact, ter Braak, again using the rental agency, had arrived the same day he left the Sennitts, on January 31, with his two suitcases, at 11 Montague Road, the residence of Miss Rosina Greenwood, to begin a new rental. Part of the arrangement was that he did not require feeding (apart from toast and tea at breakfast), so Miss Greenwood had no need to ask her new lodger for a ration book. Yet the indolence of the authorities is perplexing. Why on earth would Mr and Mrs Sennitt, who had watched a lodger with dubious papers suddenly come under close examination from the Food Office, one who apparently never followed up with the Aliens Officer, and then observed him displaying erratic and mendacious behavior in suddenly quitting his lodgings and making spurious claims about going to London, yet not boarding the train, but taking a mysterious suitcase with him which they had not seen before, not think they should perhaps report such behaviour to the authorities? Had they perhaps been primed to stay silent about the whole business?

Yet ter Braak managed to stay undetected for another two months. He continued his perambulations; he made a visit to London; in mid-March he went to Peterborough for the day. He visited his office above the agents for the last time on February 12. Obviously it was important for him to stay in Cambridge, although one might judge that he could have been safer settling down in another city, with a different Food Office. Why did he think he could remain safe? He left the house each day, but, according to Miss Greenwood, never took either of his suitcases with him. And then, according to one report, a fortnight before he left for the last time, in mid-March, he told his landlady that he would be leaving in two weeks. Gwyer’s report, however, states that he left his lodgings on March 29, without settling the bill, but taking with him nearly all his luggage, and returning the key to his landlady. He told her he would be back on April 5. Incredibly, apart from an aside where he tells us that ter Braak asked, on March 20, a fellow lodger who was employed at Lloyd’s Bank to cash some dollar notes for him (an account that is incidentally undermined by the subsequent police report which states that ter Braak asked Mr Sennit, his first landlord, to exchange some dollars, which duly occurred at Lloyd’s Bank), Hart’s memorandum just devotes one sentence to ter Braak’s time at 11 Montague Road, and does not discuss the circumstances of his departure at all.

The ‘Suicide’

The investigation into ter Braak’s death was beset with contradiction and controversy immediately his body was found. It was perhaps a bit hasty of Liddell to record confidently in his diary on that day, before any official police report or inquest the following assertions: “He had evidently shot himself and had been dead some 36 hours. His Dutch papers were out of order and did not show any authority to land. He had lived in Cambridge for about four months [thus echoing the erroneous Police calculation]. He had arrived about 4th November with a small suitcase and parcel.” Such a conclusion would have required the Cambridge Police to have traced ter Braak’s residence through two rentals in the space of a few hours. Yet that unit did not deliver its report on finding the body until April 4, the day of the coroner’s inquest. Amazingly, that report declares that the first landlady’s address was determined only through ascertaining that the suit that ter Braak was wearing had been bought at the Fifty Shilling Tailors in Petty Cury, and the purchaser had given his name and address of 58 Barnabas Road, Cambridge. The report continues: “Detective-Inspector Ernest Bird went with Captain Hughes and Detective Sergeant Robinson to that address, and interviewed the occupier.” The officers then had to make ‘continued enquiries’ to trace the suitcases in the Left Luggage Office, as well as tracking down and interviewing ter Braak’s second landlady. This constituted an impressive piece of work by the Cambridge constabulary, in order to be prepared for the April 4 inquest.

Liddell appeared to be on top of the case, and have paranormal insight, as he recorded the following in his April 3 entry: “There is no doubt that he was the parachutist who was reported to have come down near Bletchley. We have obtained his wireless set which was in the cloak-room of Cambridge Railway station.” It appears he had been kept informed by Robertson, who in another memorandum written on April 2, indicates that Mr. Hughes, the Regional Officer at Cambridge, had spoken to him on the telephone the previous evening, informing him that the police had spoken to the first landlady (only). Hughes called again on that morning (April 2) to let Robertson know that they had found ter Braak’s possessions at the station. But at that time they had not yet interviewed Miss Greenwood: Liddell should not have been able to know for certain that ter Braak had resided in Cambridge ‘for about four months’ as he stated in his April 1 diary entry. He clearly anticipated the error from the April 4 report.

In any case, MI5 was preparing single-mindedly for a quick and clean inquest. As early as April 2, Brigadier-General Jasper Harker, Deputy Director-General, wrote a letter to the coroner W.R. Wallis, pointing out how delicate the case was, and how grateful he was to Wallis for consenting to take the case in camera. ‘We have in our possession information which satisfies us beyond any doubt that the deceased was an enemy agent’, he wrote, an extraordinarily premature claim if it had been based on materials found on ter Braak’s body, and definitely not yet in the possession of MI5. Harker continued: “You will appreciate that it is of paramount importance that no report of any kind should be published with reference to the proceedings before you”, adding that the Home Office had approved MI5’s request for secrecy. All went off very smoothly. Dixon was able to report to Dick White that ‘I am pleased to say that the Coroner’s inquest went off very successfully, Brigadier Harker’s letter helping in no small degree.” The target of Jane Archer’s lampoonery had come through.

Yet the facts of the death should have provoked some deep questions. The police report said that the body had first been found by an electrician who had entered the air-raid shelter to complete electric installation at 11 am on the 1st April. “He had at once telephoned the police. Nothing had been disturbed by him.” Police photographers then took pictures of the body, which was then removed: a list of possessions found on the body was made. Yet when After the Battle published the results of its investigation in 1976, it told how its researcher had interviewed Mrs. Alice Stutley, who had been Air Raid Shelter Marshal for her area in 1941. She told her interviewer that she had been walking her dog in the park, Christ’s Pieces, where six air-raid shelters stood. At about 9 a.m. a small boy came up to her and said that there was a dead man in the second shelter. She at first took the observation as an April Fool’s Day joke, but, when the lad persisted, inspected the site herself, confirmed that a corpse was there, and informed the police, who likewise thought she was jesting. When he finally agreed, she showed him the body. She heard nothing more until that evening, when she received a knock on the door, and two unidentified men asked her if she was the one who had found the body. ‘When she said she was they warned her in no uncertain terms to keep it to herself and not to say a word to anybody”.

Christ’s Pieces, Cambridge

Yet they were too late. Apparently, the next day the Cambridge News contained details of the discovery (according to After the Battle), but they were withdrawn after an urgent call from the Ministry of Information, the government department responsible for censorship. In the 1974 interview, Mrs Stutley told the investigator that the fact that the name of the dead man was ter Braak and that he was a German spy was ‘common knowledge amongst the older residents of the area’. And small boys talk.

But now we come to the most arresting part of the story. I present the photographs of the corpse here – the pictures that the authorities might cause ‘some distress’. In his words, Mrs Stutley also told the investigator the following: “She found the body of a man, dressed in a dark overcoat, horn-rim glasses and black homberg (sic) hat, wedged tightly underneath one of the fixed slatted seats. She could see no pistol but it was evident that he had been shot in the head.” Unless Mrs Stutley had some strangely defective vision, which allowed her to see accurately the details of ter Braak’s body but not the weapon a foot in front of it, or had some inexplicable reason to lie, her evidence would suggest that the weapon was not present when she inspected the corpse, but must have been later planted next to it.

Ter Braak’s Corpse

Apart from that, one cannot look at the sad photograph without thinking: how could a man who committed suicide manage to wedge himself underneath a seat after he shot himself? And, as the police report informs us, ‘a bullet wound was traced above the left ear’. Even an amateur criminologist could point out that, in order for a suicide to shoot himself in the left side of his temple, he would have to be left-handed. Was that so? Did a graphologist ever study the writing in his letter to the Feltham Food Office (shown below), or ask his landladies whether they noted how he wrote? To me, the slant of the words in his letter suggests the script of a right-hander, but should that not have received expert analysis? Was the weapon inspected for any fingerprints? The report tells us that it was a Browning Automatic Revolver, No 468225, Calibre 7F/M55. Did the bullet-hole match the calibre of the pistol? And was the weapon traceable? It certainly looks like a Browning High Power, manufactured in Belgium under German control from 1940: did MI5 record whether such a device was found on other agents?  As I noted earlier, the report on Gösta Caroli tells us that he carried an unnamed German revolver, and the After the Battle story, in its more fanciful exposé of ter Braak’s room being searched, relates that a Luger was found in his possessions. Whatever the real facts were, the whole process of the post-mortem shows an extraordinary slackness and naivety. Perhaps the authorities though they would get away with it, as they believed the photographs would never come to light, and that they had instantly hushed the source of any dissent and rumour that might shed light on the true circumstances of ter Braak’s death.

Ter Braak’s Letter to Feltham Food Office

Certainly, there were circumstances that might have led ter Braak to consider suicide: he was running out of money, the Food Office was tightening up on his papers, and he may have felt abandoned if and when he ventured out at night to make contact with someone who could provide him with money, and no help came. Why did he cover himself in newspapers, no doubt to keep warm, if he had already considered killing himself as an option? Did he consider all hope gone when he failed to make contact with another parachutist? That seems unlikely, as none was scheduled to help him: Richter was dropped later to verify whether TATE had been turned, not to help ter Braak. Did he receive guidance to make a rendez-vous with SNOW, perhaps, and was then ambushed? Was another agent actually scheduled to parachute in, like Fackenheim, but was frustrated by the weather? These questions may remain unanswerable unless further documents are released.

Yet indications that the death was not actually a suicide can be seen in later memoranda concerning the case. On September 11, 1941, while Robertson was charged in carrying out a deeper investigation, he sent a message to Hughes of B.3.b (responsible for liaison with RSS) that included the sentence that I quoted earlier: “This investigation is in connection with the Cambridge ‘suicide’ and while I believe you may have had some general enquiries made I think that these new suggestions may narrow down our field of search considerably and prove more fruitful.” The appearance of the word ‘suicide’ in quotation-marks is a clear expression that the word should not be taken literally. Three days later, Hughes echoes and reinforces the notion, showing that he is another in the know: “With regard to your note of the 11th Sept. concerning the possible workings in connection with the Cambridge ‘suicide’ case, the Controller, R.S.S. replies that they are doing what they can with extracted records, but that the original traffic of that period has been destroyed.” But the coroner had given his verdict. Why should the MI5 officers question it? The weeders of the archive were as careless as Robertson and Hughes were in their missives.

Thus, if ter Braak was not a suicide, who was responsible? Was he murdered by a person or persons unknown, with MI5 deciding that it would be too dangerous to bring his case into the open? The reactions of Liddell and others after the incident suggest not. The only other explanation is that MI5 engineered an extrajudicial killing since it had no reasonable alternative.

MI5’s Blunder

“It is not altogether fanciful to speculate how much more happy and more useful his career might have been if he [ter Braak] could have fallen into the hands of the Security Service and become a double agent.” (Sir John Masterman, in The Double-Cross System, Chapter 3)

A possible sequence of events leading to such a decision could run as follows. When it learned of the imminent arrival of another agent, MI5, instead of making the immediate step of preparing to arrest him, and interrogate him in order to determine whether he could be turned, or would have to be executed, decided on a new ploy. It would watch the agent’s actions and movements, to try to discover exactly what his mission was, the substance of the messages he passed back to Germany, and whether he had any contacts in the country. It is possible that this strategy was promoted by Frost, opposed by Archer and perhaps Robertson, but gained a measure of support from Liddell, who managed to sell the proposal upwards to Harker and Swinton. This strategy meant that a select number of agencies (one or two RSLOs, Police, rental office, landlords, Food Offices, etc.) would have to be brought into the subterfuge. No doubt ter Braak’s messages were picked up by RSS, and analysed, but a stop was put on them. For a while, the project progressed smoothly. But then matters started to get out of control. The spy’s wireless stopped functioning. The problems with the ration cards began to extend to a broader base of officials, which meant that it was becoming increasingly difficult to allow him to survive unchallenged. Bombing casualties could perhaps be attributed to him. Yet there was also pressure from the RSLOs for more disclosure, and greater control. How would they react if they learned that they had not been taken into their bosses’ confidence, and that MI5 senior officers had tried to deceive them? So something had be done with ter Braak.  But what?

It was too late to try to turn him, as, even if he had become compliant, the Abwehr would wonder why an agent who was desperately running out of battery power and money had suddenly come to life again. MI5 could not now arrange a trial, as it would bring the spotlight on to the circumstances as to how the country’s security defences had failed by allowing a parachutist to evade dragnets, identity checks, ration-book requirements, and police surveillance for so many months. When the story got out, it would make MI5 a laughing-stock, and heads would surely roll. So the obvious answer was to eliminate ter Braak. Masterman’s observation (above) now appears cynical rather than sympathetic.

There could have been a precedent for such an action a short time before. As I outlined above, Gösta Caroli (SUMMER) tried to strangle his minder on January 13, 1941, and then attempted an escape with a canoe on a motor-cycle towards the North Sea. He was soon captured, but he had clearly broken his commitments, and would have been a constant danger because of what he knew and where his true allegiances still lay. (He had attempted suicide on October 11, 1940.) The conventional story is that Caroli was incarcerated (at Huntercombe, 020R, a reserve camp for Latchmere House), and then repatriated to Sweden after the war. Yet this account is suspiciously contradictory, as Nigel West, in Seven Spies Who Changed the World, tells us that Caroli was interned at Camp WX on the isle of Man.  Elsewhere, West writes in MI5: “SUMMER was deported to Sweden at the end of hostilities and found a job in an agricultural seed firm near Malmo. His health deteriorated gradually and he died in 1975, having spent his final ten years in a wheelchair.” Yet Hinsley’s official history has no trace of SUMMER after his arrest, and his file at TNA stops abruptly at the same point. If the Double-Cross System had to remain a secret until 1972, why would the British have allowed a renegade from it to return to his homeland to talk about it.

The prolonged incarceration and repatriation might well be a ‘legend’ created by MI5, especially given the confusion that MI5 officers may have generated through the conflicting accounts that they leaked to journalists and historians. For other writers point to a more grisly end. In his book The Druid, Nicholas Mosley presents the following commentary: “He [Cassio = SUMMER = Caroli] was indeed dead, but not by suicide. After his capture, the English had put him on trial for espionage and at Birmingham prison, in the first week of February, 1941, he had been hanged. Every old lag in the jail was talking about it. It was one of the executions the British had carefully not made public, in case someone in Germany wondered exactly what had happened to Cassio between his landing in Britain and his death four months later. But you couldn’t stop warders and released prisoners from talking.” And, in a more serious work, Time-Life’s The Secret War (1981), the author, Francis Russell, writes: ‘For his effort, Summer met the traditional fate of the spy – execution’. Unfortunately, it is not clear what the source for this claim is: Russell acknowledges a number of distinguished intelligence experts who assisted him, but none of his statements is individually attributed.   (The Swedish biographers of TATE, Tommy Jonason & Simon Olsson, attribute to Nigel West the fact that Caroli was imprisoned at Huntercombe after the Isle of Man, but West in fact makes no such claim, as I have recently discovered, having acquired the book. Mysteriously, given their nationality, Jonason and Olsson offer no details of Caroli’s supposed return to Sweden, listing in their Bibliography only an unpublished manuscript by themselves on the short-lived double agent.  Their only photographs of him are sourced to a Claes Caroli, but are from Gösta’s pre-arrest period.)

Gosta Caroli

Hinsley does indeed refer to SUMMER in terms suggesting he was still alive in November 1941, when Swinton, the chief of the Security Executive, held discussions with MI5, the Director of Public Prosecutions, and the Attorney-General, on what should be done with double-agents whose career had come to an end, specifically about SUMMER and GANDER. But that minute should perhaps not be trusted: it was fearfully late to be discussing what to do with SUMMER when he had tried to abscond to Germany ten months beforehand: he was either incarcerated or dead by then. Ironically, in March 1941, MI5 recorded the opinion that executions were wasteful, and that ‘intelligence should have precedence over blood-letting’. Hinsley notes that MI5 was ‘also fearful of the damage to security that might follow from the fact that, as was inevitable when a spy was put on trial, the Press demanded maximum publicity about the case’.

What appears as solid evidence to support the idea that Caroli could have been murdered is the evidence from Liddell’s Diary. On January 16, 1941 (i.e. immediately after Caroli’s desperate escape attempt, which Liddell recorded graphically the day that it occurred), Liddell noted: “We had a long discussion this morning about SUMMER’s future and that of the other people with whom he has been associated. We have all come to the conclusion that somehow or other SUMMER must be eliminated.” And on February 3, he added, concerning how parachutists should be handled: “It was agreed that V.V. [Valentine Vivian, of SIS) and I should make representations to Swinton that bumping off should be the exception rather than the rule.” This crude language, not suggestive of a formal trial and sentencing, indicates that an exception had probably already been made. A week later, he discussed the matter with the Director of Military Intelligence, Major-General Davidson, who appeared to support the notion that a case for ‘bumping off’ agents should be prepared by Liddell and Vivian. The archives show that executions would be the order of the day if an invasion occurred, but these are the clearest indications that extrajudicial killing was considered acceptable when captured spies could not yet be considered part of an enemy military force.

And is it possible that MI5 was influenced by the case of Krivitsky, the Soviet defector whom Liddell and others had interrogated just a year beforehand? On February 10, just as the whole ter Braak saga was winding down to its embarrassing conclusion, Krivitsky was found dead in a Washington hotel, very probably a victim of a murder by Stalin’s Special Tasks force made to look like a suicide. If these were extraordinary times, and the risks to the security and success of the whole Double-Cross operation could be multiplied by careless talk from a bitter ex-double agent in prison, maybe the ‘wet business’ (mokrie dyela) of the NKVD should be imitated. For MI5 had indeed dug itself a large hole in its management of ter Braak, and an obvious ‘suicide’, and quick and secret inquest would be the cleanest solution to its quandary.

Lastly, MI5 had some unfinished business to attend to. It had to account for the intolerable leaks in security that had allowed ter Braak to remain undetected, and ensure the correct disciplinary action was undertaken, and it had to make discrete arrangements for the disposal of ter Braak’s body. Lord Swinton wanted a full inquiry. But this was obviously a time for mercy. On May 7, 1941, Dick White, on behalf of B.2, drafted a letter to be sent to the RSLOs that drew attention to the lessons learned from the exercise, and identified the failings of the Cambridge officer involved. Liddell approved it, but added that he was very anxious that the RSLOs ‘should not make use of it in such a way as to pillory the Cambridge police’. David Petrie reinforced this motion of tolerance, and wrote a letter to Swinton, in which he explained that it was ‘the human element’ that failed, and repeated the need not to ‘pillory’ the Cambridge police.  The eventual cover memorandum clearly put the blame ‘almost wholly on the slackness of the Assistant Aliens Officer’ but the clause that ‘It is not, however desired that the conduct of the Cambridge Police should be singled out for condemnation’ was underlined in the letter. Thus the unfortunate office – who must surely have been receiving guidance from his superiors – was not unjustly dealt with, and the much greater failings of the security infrastructure, including the bewildering inefficiency of the RSS to detect illicit wireless signals, were conveniently overlooked.

As for the disposal of ter Braak, the archive is silent after the inquest, but investigative journalism thirty-three years later, as reported in After the Battle, unveiled an eerie tale of switched bodies and curious reporters trying to chase down the burial. In summary, John O’Hannan, a funeral director at the time, was asked by the police to move the body from the shelter, and sworn to silence, but he believed that one of his assistants informed the Press, whose members pestered him. Instructed to bury the body in an unmarked grave when the Coroner’s certificate was issued, he switched the body with that of another suicide in order to distract reporters, and moved the corpse to the church at Great Shelford, three miles south of Cambridge, where he had made arrangements with the vicar. “At 9:00 a.m. on April 7, Ter Braak was buried, attended only by the Vicar, the gravedigger, Mr Duisly the clerk and Mr. O’Hannan.” Recently, the town of Great Shelford has researched ter Braak’s death, and given him public recognition. (see https://www.cambridge-news.co.uk/news/cambridge-news/nazi-spy-who-getting-gravestone-13257650)

Ter Braak’s Grave at Great Shelford

In summary, MI5’s collusion in ter Braak’s fate can be described as follows:

  • Liddell betrayed the fact that he knew of ter Braak’s impending arrival
  • A probable clash between Archer and Liddell over policy was concealed
  • Robertson disclosed the fact that ter Braak was being considered for XX work
  • The procedures for informing RSLOs were bypassed
  • Ter Braak, unlike other agents, was not pursued vigorously
  • Ter Braak’s identity card was not checked
  • Ter Braak’s transmissions were ignored
  • MI5 officers made imaginative assessments of ter Braak’s radio usage
  • Ter Braak’s problems with ration cards were suppressed
  • MI5 did not discuss ter Braak’s fugitive problems with RSLOs
  • Ter Braak’s suspicious movements were ignored
  • The finding of ter Braak’s corpse was mismanaged and its state misrepresented
  • Liddell accepted too hastily the report of suicide
  • The Cambridge Police Report was anticipated by Liddell and Harker
  • MI5 showed secretive and inappropriate haste over the inquest
  • MI5 tried to silence news reports
  • The authorities showed indulgence over serious breaches of security protocol
  • Petrie and Swinton tried to draw attention away from the errors of Cambridge Police
  • Robertson and others referred elliptically to a ‘suicide’
  • Hart was encouraged to write a mendacious report

Conclusions

MI5’s problem was that it conceived, despite opposition, an imaginative but risky plan, and then failed to think through the consequences.  If it tried to manipulate and monitor an enemy agent, what happened when that agent outlived his usefulness? Great Britain was not a totalitarian state, with arbitrary powers of arrest and execution: the rule of law applied.  At the time, one can understand, perhaps, why its officers felt ashamed at such a lapse in ‘fair play’ in dealing so savagely with a dangerous element whom they had toyed with. Hence the enormous cover-up. Yet, in looking back from seventy-seven years ahead, one could wonder what other choice they had. The deed paled in comparison with what the NKVD and Gestapo were doing at the time – or even with the murders and executions that some SOE agents would soon be performing in France.

The confidentiality of the Double-Cross System – even though MI5’s officers did not then understand how vital it would turn out to be for the winning of the war – would surely have been at stake if it had ever become public that Britain’s Security Service had actually acted as a puppet-master for German spies airlifted into Britain. If they had simply recommended to Special Branch that ter Braak be arrested, even if he were eventually executed, it would have been impossible to keep the lid on the whole exercise. Yet the notion of extrajudicial – even legal – execution, was apparently such an anathema to the culture of Britain’s counterespionage officers (despite Liddell’s casual aside about a ‘shooting-case’) that they must have had some deep misgivings over the whole episode. Liddell’s admissions in his Diaries show that such decisions were not taken lightly, and required approval at higher levels. They learned their lesson, however: no more monitored agents were allowed to roam at large, and the next two who arrived, Jakobs (who parachuted in on January 31, 1941) and Richter (May 13), went fairly promptly to the firing-range and the gallows respectively.

The Abwehr’s lack of interest in the ter Braak case is puzzling. No doubt the news of the ‘suicide’ reached it in some way at the time, as Ritter and Co. would have needed to know of a convincing explanation as to why their agent had stopped communicating, but did not appear to have been arrested. Maybe SNOW was able to pass on a ‘rumour’. Yet why, after the war, Abwehr officers claimed no knowledge of who he was, both Ritter in Hamburg and Sensburg in Brussels, even though ter Braak was given an identity card number provided by SNOW, and the obvious (too obvious?) evidence of activity in Brussels was provided with the parachute paraphernalia, remains a conundrum. Perhaps ter Braak really had been on a mission of such dastardliness that the interrogated officers believed it might hurt their chances if they admitted to knowledge of it. On the other hand, their interrogators did not really press them on the subject. That question may have to remain a mystery.

As for the RSS, it was vindicated. No doubt it did pick up all the signals that ter Braak sent before his batteries ran out, but, when they were sent to the headquarters at Arkley, the monitors were guided not to be concerned about them, to the extent of using direction-finding equipment to locate them precisely, or even sending out sniffer-vans. The messages themselves, however, were no doubt closely inspected, and probably deciphered, since ter Braak would have been using a cipher-wheel similar to that issued to TATE. At the same time, the fact that an Abwehr agent had been able to evade British detection-finding equipment for many weeks would have been a convenient message for MI5 to have acknowledged by its counterparts as it tried to maintain the illusion of a flawed system of radio interception in the country. And that, of course, is the most significant underlying discovery of this whole research exercise, the saga of the ‘Mystery of the Undetected Radios’.

So was this irregular outcome the ‘Greek tragedy’ hinted at in Curry’s report at the end of the war? Probably not. It all happened before the transfer of RSS to SIS, and, if there had been an aberrant directive to Arkley on the investigation of illicit domestic traffic, it would have been engineered by MI5. The reporting structure of RSS was at that time irrelevant. SIS can in no way be held responsible for what happened, unless a malign influence on MI5’s policy towards foreign agents was somehow exercised by such as Claude Dansey. (We should perhaps recall that Dansey had, with MI5’s connivance, recently engineered Soviet spy SONIA’s British citizenship and passport, giving her the right to leave Switzerland and enter the UK as an observed agent.) Such a hypothesis, however, would push speculation too far into undocumented conspiracy theory. We are left, however, with the bewildering story of Malcolm Frost’s being airbrushed out of history, like one of Stalin’s commissars.

Finally, the archive itself. For those researchers who claim that archives alone can tell us what really happened, this exercise has an ambiguous message. The whole ter Braak/Fukken dossier has an unreal air about it, as if it had been carefully selected and crafted to display a solid and incontrovertible story. Superficially, it does just that, with the memoranda about missed opportunities, and flawed systems, and inattentive watchdogs, and all-too-human failure, embellished by realistic accounts of inquests, exchanges with bureaucrats, photographs of a corpse, and statements by experts into subjects such as wireless messaging. Yet a deeper inspection reveals all the contradictions, anomalies and misrepresentations that are bound to occur when a broad conspiracy is under way, and too many persons are involved. Maybe we should treat the release of the whole file as a challenge by MI5 and the Home Office, and whatever other authorities were involved. Maybe they intended to say: “Look. Enough time has passed. The clues are here. Work out what really happened. Revealing the secret really no longer affects the security of the country, or the reputation of MI5. All the participants are now dead. Go ahead.”

On the other hand, it could simply be that so much institutional memory has been lost that no one understands the complications behind these fragmentary documents any more.                                 © Antony Percy, 2018

(I am again grateful to Dr. Brian Austin for his advice on wireless matters. The opinions here are my own. Readers may be interested in a fictionalised account of ter Braak’s time in Cambridge by Tony Rowland, titled Traitor Lodger German Spy, now available in paperback at https://www.amazon.com/Traitor-Lodger-German-Tony-Rowland/dp/1912309475/ref=sr_1_1?ie=UTF8&qid=1537892509&sr=8-1&keywords=tony+rowland+spy  I have deliberately delayed reading this work until I completed this month’s posting.)

An Update on ter Braak (October 31, 2018)

I have a professional researcher and collaborator in London, Dr. Kevin Jones (whom I have not yet had the pleasure of meeting) who is an expert at navigating the indexes at Kew, and who on my behalf inspects undigitised files at the National Archives, namely those that I cannot acquire and download. He has recently been delving around some of the lesser known folders that have a bearing on the ter Braak case, and I wanted to bring some aspects of these to readers’ attention to provide an update to the analysis outlined in last month’s blog.

One of the challenging dimensions of ter Braak’s life as a fugitive is the level of political compliance in the plan to let him roam, and then to eliminate him. It is clear from Swinton’s endorsement of the initiative for a speedy inquest, and his approval of the decision not to engage in recriminations over the Cambridge Police Force, that he was either party to the original decision, or had been convinced of the need for extraordinary measures when matters started to run out of control. Given the speed with which actions progressed after ter Braak’s death, it is more probable that he had approved the whole operation in advance.

But how much did Churchill know? Swinton had been brought in by Churchill to oversee MI5, and parts of SIS, after Vernon Kell had been dismissed, since Churchill was not confident in MI5’s ability to defend the country against the ‘Fifth Column’ menace that he then believed had been a prime factor in the defeat of the France and the Low Countries. Swinton had not been a universally popular choice at the House of Commons, but, when the Fifth Column ‘menace’ was shown by August 1940 to have been illusory, Swinton’s supervision of MI5, and mission to help Jasper Harker, the acting Director-General, to rebuild the service, continued, and his focus on subversion shifted to the arrival of the Abwehr agents. The files PREM 3/418/1 and 2 show records pertaining to the establishment of Swinton’s Committee, the Home Defence (Security) Executive, and correspondence on enemy agents between Swinton and the Prime Minister.

On September 10, Churchill made a request at Cabinet for a report on information obtained from enemy agents in the UK. The records show that he was informed about the declared mission of the four agents who landed in Kent (three of whom were executed) and of Gösta Caroli (who was successfully ‘turned’ – for a while). Reports from the interrogations indicated that the spies believed they were the advance guard of an invasion that was to follow in a couple of weeks. When Swinton reported, on October 4, on the spies who landed by boat in Scotland, however, he showed that Churchill already knew about ‘Agent 5 and Agent 6 who are being used successfully in deception operations already’. The list provided to Churchill has yet been found, but is highly noteworthy that it cannot be the same list that MI5 used for RSLO training (KV 4/407), since ter Braak appeared there as Agent 5, but he had not yet arrived when this memorandum was written!

Thus we have proof that Churchill knew about the emerging Double-Cross operation as early as October 1940, if not sooner. This all goes against the grain of what the authorised historians tell us. In his recent book, The Secret World, Christopher Andrew suggests that the first report given by MI5 to Churchill on the actions of the XX Committee did not occur until March 26, 1943. “It was an instant success with the Prime Minister. Churchill wrote on it in red ink: ‘deeply interesting’”, writes Andrew. This is rather hard to believe: that Churchill, with his massive interest in intelligence matters, and having been made aware of Nazi agents being used for deception in the autumn of 1940, would let the matter drop for two-and-half years.

In addition, it would seem that Swinton was passing Churchill a longer list than was being maintained by MI5. Given what I have written about overlooked spies not appearing in the official records, it would be fascinating to learn what names had been given to the Prime Minister at this stage. The archive is confoundingly sparse at this point. On October 31, Swinton advised Churchill of the arrival of the spies Lund, Edvardsen and Joost, but the narrative stops on November 2, just after ter Braak has landed. Is that significant? The search for other revealing items that might fill out this story continues.

At this stage, one can only speculate what went on between Swinton and Churchill. Since ter Braak was not a captured spy, perhaps Swinton would have interpreted his guidance literally, and decided to conceal the project from his boss. What would Churchill have thought of an armed agent running loose in the Cambridge area? Might he have approved of the plan to monitor his activities in order to learn more? Knowing his expressed desire at this time to see more spies executed, however, it is more likely that he would have cancelled the project, and have ter Braak hauled in. I just hope that some other records are found that shed light on this intriguing dynamic.

I also made some changes to the September text in the light of a re-discovery of passages in Guy Liddell’s Diaries, to which Dr. Giselle Jakobs had pointed me. I had read these a long time ago when I was not focussed on the LENA agents. They show very clearly that, just after Caroli’s recapture, Liddell discussed very seriously with his superiors (and Valentine Vivian in SIS) the possibility of ‘elimination’ or ‘bumping off’ of recalcitrant German agents. This is not the language of judicial trial and possible execution. Yet Caroli’s possible career after incarceration is plagued with contradictions, a matter to which I shall eventually have to return. In the meantime, please see Dr. Jakobs’s website at http://www.josefjakobs.info/ for a recent posting on Caroli,

Lastly (for this month, anyway!), is the fate of Jasper Harker. In last month’s blog, I had begun to cast doubt on Guy Liddell’s declared rationale for Jane Archer’s dismissal, namely that she had ridiculed Harker one time too many. Liddell reports the sacking on November 18, 1940, and, two days later, suggests that she should speak to David Petrie. Clearly, by that time Petrie has already been invited to take over the direction of MI5, so Harker’s fate was effectively sealed at that time. Ironically, Swinton was one of the few (apart from the disgraced Vernon Kell) who had supported Harker, and saw it as part of his mission to help him with the reconstruction of MI5. What PREM 3/418/1 shows is that, as early as August 29, 1940, Desmond Morton (an ex-MI5 officer, and Churchill’s right-hand man on intelligence) was telling Churchill of the multiple criticisms of Harker from within MI5, and reported that he was ‘a weak man’. Given the military circumstances, and the pressures, it seems that Swinton (whose judgment of character was not good, as is shown by his endorsements of Joseph Ball and William Crocker) was slow in realising that Harker was not up to the job, while Jane Archer – alongside multiple other officers in MI5 – had come to the conclusion much earlier that he was dragging down morale. It casts even more doubt on the reason for Archer’s forced departure, and, if a meeting with Petrie could not salvage her employment in MI5, suggests that there were deeper reasons for the parting of the ways.