SIS grew spectacularly during the war. This is reflected in a summary of SIS’s wartime accounts drawn up by Paymaster Sykes in June 1945. Sykes broke down the total sums into three main sub-divisions: ‘S.S.’ (secret service), GC&CS (signals intelligence) and ‘Section 8’ (Gambier-Parry’s technical branch). In 1940-1 the total expenditure was just over a million pounds, with ‘S.S.’ spending £889,000 (as opposed to £117,000 in the mid-1930s), GC&CS £79,000 and Section VIII £111,000. Three years later secret service spending had increased more than threefold to £2,828,000 and Section VIII’s fivefold to £569,000, but GC&CS’s had risen over sixteenfold, to £1,302,000. Numbers of personnel had increased commensurately. In April 1939 there had been forty-two officers in SIS (of whom fifteen were special operations – Section D, later hived off into SOE) and fifty-five secretaries. At the beginning of 1944 SIS had 837 people working at headquarters: 307 officers and 530 secretaries, as well as an unknown number of other support staff, including janitors, cooks and cleaners. Section VIII and the Radio Security Service had 190 officers and 4,783 other ranks (which included wireless operators abroad), while GC&CS had a total of 7,847. As for SIS personnel abroad (excluding the Mediterranean Command, for which figures were not available), there were 253 officers and 322 secretaries.1 These figures do not include numbers of agents, which are effectively incalculable. In 1946, for example, Menzies, evidently including anyone who served even in the most minor capacity, claimed that during the war ‘we had 25,000 French agents working with S.I.S. and the D.G.E.R.’ (the French secret service).2
The growing size and responsibility of SIS continued to put pressure on its administrative structures. Following the 1942 reforms, when the service Deputy Directors had been brought in and the production side tidied up, further changes were made in the spring of 1943. Partly to improve liaison with the War Office, and partly to ease his administrative load, Menzies had his old friend General Sir James Marshall-Cornwall transferred from SOE to SIS and appointed Assistant CSS on 29 March. The intention was that he should understudy Claude Dansey (who had been redesignated Vice CSS in October 1942), and act for him in his absence. Marshall-Cornwall, who had served in the Intelligence Corps during the First World War and had been military attaché in Berlin from 1928 to 1932, remained as ACSS for the rest of the war but his time at SIS does not, on the whole, seem to have been particularly happy. It was apparently intended that he should relieve the sixty-seven-year-old Dansey of some of his burdens, assuming responsibility for his territorial concerns across Europe. He took over the supervision of Mediterranean stations, but Dansey held on to his French operations. In the careful words of Robert Cecil (who succeeded Patrick Reilly as Menzies’s personal assistant at the end of September 1943), writing long afterwards: although ‘showing signs of wear’, ‘“Uncle Claude” remained sufficiently himself to resist intrusion into the secrets of the networks he was developing in France in anticipation of D-Day’.3 Marshall-Cornwall, too, had a tendency to bypass existing arrangements, which upset the occasionally somewhat Byzantine SIS chains of command. Both John Bruce Lockhart in Bari and Harold Gibson in Istanbul grumbled about his ‘interference’. The former told Cuthbert Bowlby that Marshall-Cornwall’s ‘complete ignorance and misrepresentation of the facts’ had caused ‘indignation and resentment’ and he objected strongly to Marshall-Cornwall’s ‘practice of addressing or repeating insulting and entirely ignorant letters to my officers without consulting you or me first’.
At the same time as Marshall-Cornwall’s appointment, Valentine Vivian’s designation was changed from DCSS to Deputy Director/SP. Although in practice his duties did not change, and he remained in charge of security and counter-intelligence, the new title, sharing that of Deputy Director with the three service appointees, could widely be interpreted as an effective demotion. Combined with Marshall-Cornwall’s appointment as Assistant Chief, and Dansey’s confirmation as Vice Chief, this change undermined any claims Vivian might have continued to nurse for the position as Menzies’s number two. A more unambiguously positive appointment in March 1943 was that of Commander Kenneth Cohen to the new position of Chief Staff Officer, Training, which was designed to make training more comprehensive and systematic than hitherto. The administrative side at Head Office was further reinforced in September when Menzies, responding to concerns which colleagues such as Frank Foley had felt, created a new post of Deputy Director Administration who was to be responsible for ‘recruitment and administration of all personnel, officers and secretaries; works, buildings and transport; welfare; finance . . . and the general running of the machine except in regard to Section VIII’ which was to ‘remain autonomous in the above respects’. To fill the post, which he held for the rest of the war, Menzies brought in Air Commodore Harald Peake, an able businessman (albeit also an Old Etonian) who had been serving as Director of Air Force Welfare. As with the case of Vivian, this appointment marked another change from the old order, as Peake supplanted Rex Howard, who had been chief of staff since before the war. Howard’s own post was ‘put into abeyance’, and he was reduced to being chief staff officer to Air Commodore Peake.4
Over the turn of the year 1943-4 the experiment of combining armed service representation with geographical responsibilities was abandoned. Of the three service Deputy Directors, only Bill Cordeaux was adjudged to have been a complete success. Menzies separated the two functions, creating a series of ‘controllerates’ covering wide geographical areas, and with primary responsibility for production. Cordeaux was made Controller Northern Areas (CNA), Cohen Controller Western Europe (CWE) and Marshall-Cornwall (whom Menzies consistently supported) Controller Mediterranean (CMed). A Controller Far East (CFE) position was later created to which Dick Ellis was appointed. Payne and Beddington continued as Deputy Director/Air and /Army respectively, while Commander Christopher Arnold-Forster became Deputy Director/Navy, which he combined with being one of Menzies’s personal assistants.
The administrative changes in SIS are also reflected in the increasing formalisation of arrangements at Broadway. The small Head Office community which had existed before the war had been run by Sinclair as if it were a family business. Individual officers had ready access to the Chief and no one worried very much about the niceties of hierarchy or the proper channels. Menzies continued this style of management, but it became increasingly difficult to carry on as the Service grew and the pressure of work expanded exponentially. Robert Cecil recalled that, without any prescribed arrangements for individuals to communicate with the Chief, ‘queues formed outside his room at the end of the corridor, imperfectly controlled by lights, which showed whether or not he was occupied’.5 This complete free-for-all was unsustainable, and although Menzies remained remarkably accessible, by October 1943 an appointments system was in place and staff had been circulated with a list of the times (9.45-10.45 a.m. and 2.45-3.45 p.m.) when the Chief was not available to see people.
Like all other wartime government offices, Broadway was overcrowded and scruffy. A secretary who joined in September 1943 thought that, whenever it was that the office had been built (actually in 1924), she did ‘not think it had ever been refurbished, or redecorated’. The floors were ‘covered in worn lino, quite dangerous in places, walls were a mucky grey/ white/cream, and the rooms were lit with bare light bulbs: only senior personnel were allowed to have desk lamps’. The offices were so dark that at the back of the building, which looked into a courtyard, there were what she remembered as ‘long panels of silver painted wood’ placed outside the windows and slanted in order to reflect light in. She herself worked with three other people in a ‘large gloomy room in the front of the building, grottily furnished’. When she first started, her hours were 9.30 to 6.30 six days a week, ‘but we could choose which day of the week we wanted off’, and she was allowed one week’s holiday for every three months worked. A great many of the staff were ‘serving officers who wore uniform (except on a Saturday, when they all wore tweeds)’.
This secretary worked in the Finance Section under Paymaster Sykes, ‘a fearsome gentleman in naval uniform, very deaf ’, and some of her early duties concerned staff salaries, another area where the pressures of war forced rationalisation and change. Before the war the practice in both SIS and MI5 had been for salaries to be paid free of income tax. In his second report on the Secret Service (which focused primarily on MI5) Lord Hankey raised the matter and suggested that, while it was probably unwise to alter the basis of existing salaries, the situation might be regularised for future appointments, with salaries subject to tax. In August 1940 Sir Horace Wilson, Permanent Secretary at the Treasury, told Menzies that there was really ‘no option’ but to adopt this proposal. Menzies was appalled, describing it as a ‘drastic change’. Members of SIS, he told Wilson, were ‘just as anxious as anybody else to make their proper contribution to the war expenditure’, but SIS salaries were so comparatively low that in order to make up for the deduction of income tax ‘a very considerable additional payment would be necessary’. It was still the case, moreover, that there was no pension provision for SIS officers. Security was also a consideration. Menzies was concerned that if tax deductions were made from Service salaries, and even if the information was restricted to only a few tax officials, the potential leakage of information about SIS personnel would be ‘deplorable and fraught with the utmost danger’. But the huge expansion of wartime staff, with many individuals being seconded from the armed services, made it difficult to sustain the traditional, idiosyncratic system. The Treasury had their way and from April 1942 all new appointees paid tax. But everyone, even the Chief, was still paid in cash, and the paybooks were all written in pencil, ‘ready to be rubbed out if anyone tried to prove there was an SIS’.
Training and communications
Commander Cohen’s appointment as Chief Staff Officer, Training, in the spring of 1943 marked another move towards professionalisation. The war years saw the beginnings of a more coherent and co-ordinated, if rather modest, training regime. Up to this time, with the exception of codes and signals, training had been decentralised and varied considerably from section to section. The sort of thing that might be done is illustrated by the bespoke training programme Rex Howard devised in summer 1941 for the novelist Graham Greene, due to go to Freetown, Sierra Leone, in November. Greene was detailed to visit Sections I-IV (Political, Air, Naval and Army respectively), V (Counter-espionage) and VI (Economic), who would be asked to describe their work to him, and also their requirements from the West Africa area. Section V would instruct him in the ‘general question of C.E., enemy methods, hints on security and telephones’. Greene would be instructed in codes and wireless procedure. Howard noted that, ‘if possible’, Greene should see an agent being trained and also be instructed in secret inks. He would need to read ‘Instructions to Representatives’ and ‘Notes for S.I.S. officers’, along with the instructions given to the current representative at Lagos. Howard himself would meet Greene for discussion ‘regarding S.I.S. Organisation at home and abroad. Relations with other Government Departments. Objectives, methods and the question of personal cover. Symbols for Departments and Sections. Post-boxes, cut-outs, agent doubles, [and] provocateurs.’ Greene would then see Frank Foley for instruction regarding the training of agents, and other officers for ‘discussion on censorship’, telephonic communications and security. Finally, Greene would be introduced to the Deputy Chief, Valentine Vivian. Beyond this specific SIS training, in order for the rather unmilitary Greene to maintain cover as an army officer, he was also sent to a general course for Intelligence Corps officers at Oriel College, Oxford, for the sole purpose ‘of being given the most elementary instruction in soldiering, so that he could wear battledress without embarrassment’.
However comprehensive and detailed individual programmes like this might be, it was clearly an inefficient way of providing training for a rapidly expanding Service. Among the signs of the new approach to training and tradecraft was a memorandum prepared by the Army Section in February 1942 on ‘S.I.S. Stations abroad: their operations and special tasks’. Compiled at Rex Howard’s request, this embodied general principles, ‘based on personal experience’. That the training arrangements were becoming much more systematic is illustrated by a ‘P.9 [Norwegian Section] Agents’ Training Sheet’ from 1943 which set out a list of twenty-one separate possible training requirements, including ‘Security’ (at the top), through various types of communications, ‘Principles S.I.S.’, ‘Arrest information’, ‘Pistol’, ‘Unarmed combat’, to ‘First Aid’, ‘Photography’ and ‘Parachute’. Against each of these were boxes for completion under four headings: ‘Whose responsibility’, ‘Training commenced’, ‘Training completed’ and ‘Remarks’.6
Commander Cohen’s new Training Section began in a modest way. The first officers’ course, run in the summer of 1943, had only three students, but there was a steady expansion. Ten officers attended the second course in September, and by this stage the section were helping overseas stations with training instructions and manuals. Individual officers could now be directed to specific programmes, as was Graham Greene’s fellow writer Malcolm Muggeridge, in August 1943. ‘P.11’, the head of the Africa Section, suggested that, ‘time permitting’, Muggeridge, home from Lourenço Marques and expected to return there (though in the end he did not), ‘should do a short refresher course with our Training Section to bring his ideas up-to-date. The “Tradecraft” lectures in particular should prove helpful.’ The Training Section was keen to expand its activities and when in April 1944 it responded to a request from Major Frederick Jempson of the Belgian Section to train an individual agent ‘in the best means of conveying documents through occupied territory’, especially ‘the best methods of sealing and packing films so that we could tell on arrival whether or not they had been tampered with en route’, it reflected on including the instruction on general training courses. In November 1944 Menzies told Sir Alexander Cadogan at the Foreign Office that the new training arrangements would help Service esprit de corps. In order to overcome problems of compartmentalisation arising from ‘an enforced separation of some of the Sections, solely due to housing difficulties’, he had ‘introduced a training system which gives numerous officers of the S.I.S. a wide insight into its various functions’. He was also currently ‘experimenting with a similar course for selected members of the female staff’.7
Communications were at the heart of SIS wartime operations. Gathering intelligence was difficult enough, but the whole exercise was vitiated if the information acquired could not be got back to London, or wherever it was of use. Section VIII’s outstanding achievement in developing and refining radio transmitters and receivers made an indispensable contribution, but at the sharp end it was up to individual men and women to operate the equipment in often very hazardous circumstances. A letter to Head Office in March 1943 from ‘Magpie’, a British radio operator working in the Lyons area in France, described conditions on the ground:
For the moment I have to change places after each transmission, with only one set this means that it has to be carried from place to place in order to maintain daily contacts. Owing to the absence of other means of transport this has to be done mostly on foot. Last Sunday I had to walk 9 miles carrying MED [his radio set] and the aerial, you can guess what a sport that is. Incidentally, the handle of the case is not strong enough, mine broke twice. I can tell you it is not very easy to carry without a handle.
Magpie proposed that, to avoid the risky moving of a radio set from place to place, more sets should be provided, which could be left ‘in different parts of or outside the town’. Balancing Magpie’s complaints (which considering the life-and-death circumstances in which he worked seem rather mild) was appreciation for the support he got from London. ‘I must thank Home Station’, he wrote, ‘for their work during the last months, contacts always having been made very rapidly. You may be sure that all of us here do our utmost to make these contacts a success.’ Commenting on the letter, Kenneth Cohen noted that Magpie’s network had been more or less on the run during the previous two months and suggested that his difficulties bore out the case ‘for more W/T sets in each group’. London was able to respond, and after Magpie had been in the field for six months there were thirty radio sets working in his network, for which he recruited and trained the operators locally.
Apart from the vagaries of weather, air pressure, power supply, local security arrangements and so on, reliable wireless communication also depended on the most scrupulous attention to detail in the preparation of signals, as is illustrated by one case in January 1944. Section VIII broadcast a signal for a Free French agent code-named ‘Bonita Granville’ informing him to be ready for a message at 1300 the next day. A signal of this sort was generally sent blind to a number of agents without acknowledgement by them, and London had no means of knowing for certain that the message had been received until actual contact was made at the time specified. Agents, however, also made regular contact at pre-arranged times and, as it happened, this agent’s next scheduled contact was at 1200 the following day, when, after passing other traffic, the signal about contact at 1300 as requested by P.1 (the section which ran Free French networks) was repeated, and acknowledged by the agent. But at 1300 the agent did not respond to London’s call. At 1400, however, the Free French communications training section based at Ealing in west London (who regularly monitored the traffic with France) telephoned Section VIII to say that Bonita Granville was calling London. Contact was made and a further message was sent asking him ‘to listen to us at 2000 the same night’. The agent replied asking: ‘check and repeat your encoding’. London did so, but the agent twice more asked them to ‘check and repeat’. ‘After this it is said that the agent in exasperation made the signal “Merde, merde, merde!!” and closed his station.’
What had gone wrong? Gambier-Parry himself investigated the case, personally coding and recoding the messages, which ‘came out correctly as sent by the operator’. Only when he compared the coding substitution table used in France with the master copy in London did he discover that, ‘owing to the misplacement of one figure in one line of typing out the copy’, a coding error had resulted in the first message asking Bonita to call at 1400 (which he had done), and the second asking him to call at ‘3000’, which he naturally could make no sense of. A mistake of this sort could have had extremely serious consequences. Gambier-Parry’s investigation of the matter revealed the care rightly lavished on the system. The master substitution tables – each of which had ten vertical and horizontal columns, and a hundred different versions of which were used – were typed in the office at Whaddon Hall and a carbon copy on flimsy paper given to the agent. But the tables were ‘retyped on tough paper for the station where they receive a considerable amount of handling from day to day’. Gambier-Parry reported that the mistake had occurred in the re-typing, ‘despite the fact that the invariable practice is to have the retyped copy checked by two people independent of the person who has typed it’. He had identified the original typist (who made the copy six months previously) and ‘admonished him for his mistake’, though he argued to Menzies that it was ‘a human error of which the chances of recurrence are extremely remote’.
Secret writing was a tried and tested method of communication, but it, too, could be problematic. In late 1942 Cyril Cheshire, head of the Stockholm station, had a job persuading a Danish agent, ‘161’, who was ‘naturally of a timid nature’, to take back to Denmark some new tablets for making ‘invisible’ ink. The tablets were supplied in a Swedish box ‘which had previously contained stomach pills. I assured 161 that in case of need he could prove that these pills were for his own personal use and could demonstrate this by taking one.’ But, he continued, ‘I have no idea whether this will have any ill effect on this valuable agent, but I am hoping for the best.’ Back in Broadway a colleague noted, ‘not toxic to any great degree’, and Cheshire was reassured that the ‘developer, although slightly toxic, would not harm him in smallish doses’. While ‘the whole packet might put him under for a bit’, it was ‘unlikely to bring about his demise’.
Some of the many ingenious types of concealing devices prepared for the use of SIS and its agents.
The conveyance of reports by agents or couriers out of occupied Europe was considered by Bill Cordeaux in a review of SIS operations after the war. Most documents were ‘photographically reduced in size’ and various concealing devices were used. Reflecting that ‘ingenious concealing devices are a favourite subject with writers of spy stories’, Cordeaux remarked that ‘in practice there is not a great deal to say on the subject. Once a person is suspected of carrying secret papers, it is almost certain that a rigorous search will discover them.’ Concealing devices, therefore, were ‘normally only designed to evade a routine examination’, for which ‘a quite simple device is adequate’. Among this type, ‘the well tried ones of hollow hairbrushes, nail brushes and tooth brushes, the backs of pocket mirrors, fountain pens, pencils, the soles of shoes, cakes of soap, tooth paste and shaving cream were all used’. The most popular were creamy substances, such as toothpaste and shaving cream. Reports, reduced to ‘about a square centimetre’, would be wrapped in a condom and buried in the centre of the pot of cream concerned. In February 1944 John Bruce Lockhart in Italy reported using animals ‘with false shoes, tails, horns etc’ to get matter through controlled areas. ‘A shorn sheep, for example,’ he wrote, ‘if well fitted with a woollen skin sewn onto light gauze can conceal under it on the flat of its back a number of largish documents.’ He also claimed to know of a case where a false horse penis had improbably been employed. ‘Contrary to popular belief,’ asserted Cordeaux, ‘suitable parts of the human body, such as the inside of the mouth, hollow teeth, the aenus [sic] and the vagina were not often used.’ In one instance, however, a sleeping-car attendant on the Trieste-Lausanne train brought reports from a network in north Italy to Switzerland ‘wrapped in greaseproof cloth’ and concealed in his anus. The fastidious Cordeaux remarked that the transference of ‘reports from the agents to our representatives, and their subsequent transcription into readable form are just another example of the varied and not uniformly pleasant nature of the duties of an S.I.S. representative’.
Beyond the signals intelligence and technical branches of the Service, the department which grew most dramatically during the war was the counter-intelligence or security Section V. Before the outbreak of the war it had comprised three officers: Valentine Vivian, who had run the section since its creation in 1925, an assistant, and Colonel Felix Cowgill, who began work in March 1939, apparently ‘on the understanding that he might in due course expect to succeed Vivian, aged 52, as the anti-Communist expert’.8Cowgill, the son of a missionary, came from the Indian Police where he had served as personal assistant to Sir David Petrie when he had been Director of the Delhi Intelligence Bureau. It had become apparent in the 1930s that the Germans were targeting SIS networks abroad and before the war two Section V officers were posted overseas: one to Brussels and Rodney Dennys, who had been sent to The Hague in 1937. Dennys later claimed that ‘after Cowgill’s arrival he, rather than Vivian, became the driving force’ of the section and ‘it was almost entirely due to his initiative, drive and abilities that Section V not only grew rapidly in size and efficiency, but also was the one H.O. section that really had a finger on the pulse of its sub-organisations in the various Theatres of war’.
Allowing for some special pleading on the part of a former Section V officer about the excellence of his branch, its expansion was certainly impressive. By 1945 it had 163 officers, three times the size of any other single Head Office branch, apart from Gambier-Parry’s mammoth Section VIII. In addition, Section IX, which was created in May 1943 to take on the study of Communism and Soviet espionage (which the focus on Germany had caused to be largely neglected), numbered thirty-two officers by Victory in Europe Day, making a grand total of 195 SIS counter-intelligence officers at the end of the war. The growth of Section V was stimulated not only by its counter-intelligence work, but also by the fact that from the end of 1940 Menzies entrusted Cowgill with the security of GC&CS signals intelligence material which was beginning to come on stream and which came to be known as ISOS. Having its own officers in stations abroad with their own cypher communications, this section became something of a service within the Service, arousing some resentment of its position and the resources it absorbed. The high seriousness, moreover, with which Cowgill took his responsibilities over the precious ISOS (and related) material, which was vital both for counter-intelligence and later deception operations, and the extremely tight restrictions which he imposed on its distribution, caused some friction, both within the Service and in its relations with other bodies, MI5 above all.
The expansion of Section V was initially assisted by its sister service. In June 1941, for example, MI5 approved an SIS request to release staff for the section, though not without some internal dissent. ‘Its work is so essential to our Service’, minuted one MI5 officer, ‘that . . . we are fully justified in making this sacrifice.’9 There was, however, a strong and growing feeling in MI5 that SIS had no business trespassing in what was essentially MI5’s primary concern. On 17 April 1942, Sir David Petrie, now head of MI5, while assuring Menzies that he wrote ‘with the single purpose of doing better a most important job’, made a bid to take over Section V altogether. He sent Menzies a paper arguing for the amalgamation of Section V with B Division of MI5, which was responsible for investigating all threats to security. His case was founded on ‘the basic desirability, or indeed, necessity, of handling a single subject as a whole instead of in two parts. The German espionage organisation’, he declared, ‘does not recognise our artificial divisions of a home and a foreign field, but operates without regard to geographical or other boundaries.’ There was currently ‘an immense duplication of effort . . . most particularly apparent with respect to ISOS’. Nor, he argued, ‘can there be any real doubt as to which of the two organisations, S.S. [Security Service] and S.I.S., should belong the united whole’. It would be MI5, ‘who are the executive body responsible for arresting and prosecuting enemy agents directly’. Petrie further proposed that MI5 should in future have direct liaison with GC&CS (hitherto arranged through Section V) and full access to all ISOS counter-intelligence material. MI5, furthermore, would direct the intelligence-related work of the Radio Security Service (which had been handed over to SIS on Petrie’s recommendation only a year before) and would retain the direction and planning of all double-agent cases under the control of the XX Committee.10
Unsurprisingly, this proposal did not go down at all well in SIS. Vivian prepared a draft reply (which does not survive in the archive) which he did not expect Menzies ‘to approve as it stands, since I have “dipped my pen in vitriol”’. Menzies’s reply of 11 May, while not vitriolic, was a detailed, frank and robust rejection of Petrie’s case. He started with an exaggerated definition of the purport of Petrie’s proposal: ‘that Section V. of S.I.S. should be absorbed into M.I.5, [and] that you yourself as the latter’s Director General should assume exclusive control of the Section V. functions’. Petrie, furthermore, and here Menzies quoted from the MI5 paper, would have ‘the right to communicate information and to issue directives as to what was required for the proper discharge of security functions everywhere’. He cast severe doubt on MI5’s competence to carry out delicate counter-intelligence operations in foreign territories of which they had no experience. Neither he nor the Foreign Office would agree to the transfer of Section V since the intelligence it circulated was of far wider than simply security significance. It was ‘unsound’, moreover, that Menzies, ‘as the Director of the G. C. & C. S., and being acutely aware of the vital importance to the national interest of preserving the secrecy of its material as a whole, should relinquish responsibility for the treatment of any portion of its output’, or ‘any measure of control over XX agents abroad’, where SIS understood operational risks and conditions better than MI5. Past proposals for the amalgamation of SIS and MI5 (for example in 1927 and 1931), he argued, had had ‘the virtue of logic, which the present one has not’. Besides, he concluded, the dangers of separation had been exaggerated and ‘good liaison and good will should prove the solvent’.
Petrie did not take this lying down. His reply on 5 June, which had a slightly sarcastic tone, corrected the inaccuracy of Menzies’s definition of his proposals: he had not suggested that he should assume ‘exclusive’ control of Section V functions, but was merely advocating the ‘unified direction of counter-espionage as a subject’. He introduced some shrewd new arguments. Until a year before, he wrote, Section V had consisted ‘of only a handful of officers, and was in very deep water’. As a consequence it had drawn ‘considerably’ on MI5 staff: ‘Are we to accept it then that they alone are capable of handling the products of C.E. intelligence by some esoteric method outside our comprehension and competence?’ He observed that both SIS and SOE had ‘for months past been suffering serious losses of agents on the continent’ because of German penetration, and asserted that SIS was ‘not producing enough C.E. material because of “operational” claims’, and was ‘falling down over “operational” [activity] because of your neglect of C.E., so that the vicious circle is complete’. He acknowledged that amalgamation of SIS and MI5 had been rejected in the past and said that ‘speaking for myself ’ he did not favour complete amalgamation, on the (faintly odd and certainly debatable) grounds that he did ‘not believe that as a race we show to any advantage in handling big organisations’. He assured Menzies that MI5 had ‘no wish to usurp your authority over the G. C. & C. S.’, though he ‘should have thought that since ISOS intercept material is of such great significance to my Department . . . you could have expected us to have quite as catholic an appreciation of the need of safeguarding it as anyone else’. Petrie concluded by telling Menzies that he intended to take the matter up with Lord Swinton, to whom he reported as head of the Security Executive.
Petrie’s appeal to higher authority had what was presumably the desired effect in bringing Menzies round to a more accommodating position. ‘The letters you and I have exchanged . . .’, he wrote on 28 June, ‘have certainly cleared the air, and if I may say so allowed us to blow off a little steam.’ Menzies had taken the wise course of having an ‘informal talk’ with Petrie, and, having given ‘very careful thought’ to the matter, now conceded ‘that the general basis of your original proposals is logical’. Although he gracefully accepted the need for ‘a single unified body responsible for studying the activities of the enemy secret services and for co-ordinating and directing action to counter them’, he still could not accept that ‘this unified Contre-Espionage body should be divorced from my organisation and incorporated into M.I.5’. What he proposed instead was a joint section. Considerable discussion ensued into the autumn of 1942 over the details of how this might be arranged, including sharp differences of opinion about where this new body should be located, Menzies wanting it to be at the Section V base in St Albans, Petrie insisting that it should be in central London. But no joint section was ever created and by October the two services had formally agreed merely to hold regular bi-weekly liaison meetings, and even this appears to have petered out after a few months.
Yet both Petrie and Menzies had agreed on the principle of close cooperation, and Menzies’s change of attitude seems to have been important in bringing to pass what he had predicted in his first letter to Petrie, that ‘good liaison and good will should prove the solvent’. At a practical level, moreover, Section V and B Division officers actually appear to have been able to co-operate on a more or less satisfactory basis all along, despite occasional bouts of agency rivalry. Robert Cecil, for example, cited ‘first hand evidence [albeit from unspecified sources] that MI5 had full access to all relevant ISOS from the very beginning and that at the working level co-operation was good, as indeed the fruitful outcome testifies’. Relations between SIS and MI5 certainly improved markedly from late 1942, especially after Section V did move from St Albans to Ryder Street in St James’s in July 1943, which greatly facilitated personal contacts between the two agencies.11
Stewart Menzies to Peter Loxley of the Foreign Office refusing a request to transfer Kim Philby, later exposed as a Soviet agent, out of SIS. This letter demonstrates how trusted and valued Philby was by his unsuspecting Chief.
By 1943 Section V was organised into six main territorial sections – Kim Philby heading the especially important Iberian one – each responsible for counter-intelligence in specific areas across the world. There were also four specialist sections: one processing and analysing GC&CS material; another devoted to double-agent operations; a third which liaised on censorship with the various agencies concerned; and the last dealing with vetting (an important wartime innovation) and arranging facilities for agents to enter and leave the United Kingdom. The GC&CS section used ISOS to help assemble the details of enemy intelligence organisations and individuals which were collated in the purple primers. Dossiers of German intelligence personnel were prepared for the use of the armies in the field. As regards double agents, broadly speaking those run in the United Kingdom and from British military bases abroad were an MI5 responsibility, while those operating in foreign countries came under SIS. In practice, it was a combined SIS-MI5 joint effort within the machinery of the XX Committee in London, the XXX Committee in Cairo, and the XXXX Committee, based first in Algiers and later Caserta in Italy. Each of these committees had broadly the same tasks, among which was the provision of ‘foodstuff’ or ‘chickenfeed’ – information – for the double agents concerned to pass on to the enemy, and the degree to which this information might be false, true or ‘true-ish’ was a matter of very careful judgment. The committees also had to co-ordinate double-agent activity in their areas, implement deception schemes and, by studying the questionnaires given by the enemy to double agents (among other things), prepare appreciations of enemy plans and intentions.
Reflecting on the performance of SIS during 1942, F. H. Hinsley noted the Service’s marked improvement across the board, but especially so in the Middle East and Mediterranean. By the end of the year, he wrote, ‘the SIS was at last accepted as “a serious intelligence agency”’ in this theatre.12 Cuthbert Bowlby and John Teague were always invited by General Sir Harold Alexander (C-in-C Middle East) to his very exclusive morning meetings and they were granted ready access to the Cabinet-level Minister of State in the Middle East, the Australian R. G. Casey. SIS’s position was certainly boosted by the increasing volume of signals intelligence which it was able to supply, but this was reinforced by valuable human intelligence, especially from Tunisia, which became the last bastion of the Axis forces in North Africa, as Anglo-American forces advanced from the west following the Operation Torch landings in November 1942, and British Commonwealth forces approached from the east through Libya.
From November 1942 the Dick Jones network in Tunisia supplied a mass of intelligence. Jones, who had begun to organise among fellow prisoners in the Tunis Civil Prison, was able to start almost immediately following his release on 14 November. Using an existing SIS radio which had been kept hidden in Tunis, he made contact with Tony Morris, SIS head of station in Malta. On 20 November Morris reported to London that Jones had ‘organized group of about 20 (repeat 20) de Gaullists who left prison with him’. His ‘gang’ consisted of ‘determined men’ who had arms and sabotage material ‘and can assure results on receipt of prompt reply. Many others’, he added, ‘ready to join.’ Within three hours London had replied: ‘Under no circumstances’ was Jones to ‘participate in sabotage . . . Please impress upon 78 [Jones] that he must confine himself to his duty of producing information which is of vital importance at the present juncture.’ But they also asked if the ‘gang’ might be taken under SOE’s wing, so long as Jones himself was not involved or compromised in any way. When Morris reported that there was no SOE organisation in Tunis, London softened its position. ‘Notwithstanding vital importance of maintaining intelligence work’, it directed, sabotage ‘by persons disassociated from our own activities’ would ‘certainly be welcome’. In this instance, moreover, it was prepared to leave the initiative to Jones and his colleagues in Tunisia: ‘While active operations are in progress, we appreciate here that in last resort men on the spot must be best judge of how to assist in defeating enemy.’
Reviewing the work of the Jones group, one agent summarised their activity as providing ‘general information’ on the strength, organisation, armament and location of enemy troops and supply dumps; sea and air traffic, including ‘tonnage and nature of cargoes landed’; and the ‘results of bombing, damage and effect on the morale of the native population in particular’. The detail provided can be illustrated by a report sent through the newly established SIS station at Algiers in late November 1942: ‘Petrol dumps as follows: on quay opposite maritime station clock tower Bizerta, also in wood at road Ernk to Fishery and Sidi Ahmed Road at end Sebra Bay, also at road junction two (rpt. two) kilometres southwest Bizerta-Mateur-Ferryville railways junction’. Jones himself is credited with having first reported by wireless the arrival of the monster fifty-six-ton German Tiger tanks later the same month, and his was the most productive of five groups operating in Tunisia (involving in all some fifty men and women). Their peak period was the first and second weeks of January 1943, when their reporting ‘embraced practically every branch of enemy activity, including enemy intentions and French political tendencies. Military, air and shipping intelligence supplied was voluminous - indeed, for reasons of security,’ reported Morris afterwards, ‘it was necessary to curb the almost continuous flow of messages.’ At one stage, in fact, the Malta station was decoding up to thirty reports a day. The productivity of these Tunis agents helped to raise SIS’s reputation with the military. ‘After a very cold initial reception,’ one officer told Kenneth Cohen in February 1944, ‘the information from our Tunis groups became so operationally valuable that First Army were literally hanging on our daily signals to them.’
The extent and enthusiasm of the Jones network led to security lapses, which were exploited by Inspector Marty, a Gestapo collaborator in the French police. A series of arrests in January 1943 led the police to Jean Coggia, a twenty-five-year-old former medical student who had been part of André Mounier’s network and subsequently worked alongside Jones. Arrested in Bizerta, Coggia was taken to Tunis for interrogation. Here, ‘while awaiting the arrival of the notorious Marty’, though still hand-cuffed, he managed to escape and rejoin Jones. The following month, as they endeavoured to cross over to Allied-occupied territory, Jones and Coggia, both disguised as Arabs, were caught by enemy soldiers after an exchange of fire which left Coggia with a bad wound in his shoulder. The prisoners were brought to Tunis on 23 February ‘and Coggia was hurried to Marty’s office, first-aid for his wound being refused’. An Italian officer, who was described as being attached to the SS, recounted to a Gaullist policeman what happened then: ‘although exhausted and suffering from loss of blood, Coggia maintained a complete sangfroid, and during a long night of interrogation and torture, resolutely refused to reveal anything. Indeed, he so taunted Marty and his fellow inspectors, that the blackguardly Marty finally lost all control and finished him off by shooting him in the neck with a revolver.’ The Italian said that a German SS officer present ‘was unable to stomach the scene and finally left, declaring that these were “Russian methods”!’. Jones, who managed to persuade his captors that he was in fact a British officer, was taken to Germany where, with the help of a sympathetic Abwehr major, he eventually gained formal prisoner-of-war status. After spending some time in a prison cell next to the celebrated anti-Nazi Pastor Dietrich Bonhoeffer, he ended up at the famous Colditz Castle.13
The group survived and reorganised after the loss of Jones and Coggia. Despite further arrests in March 1943, they continued to supply Malta with vital military information until ‘Allen’, an Englishman who had taken over the leadership, was himself caught on 15 April. He had been denounced by an Italian who lived in the flat below, and when the police came to arrest him, Allen jumped out of a second-floor window, but broke his leg and was quickly stopped. As reported afterwards, ‘at the time of his arrest, his real identity was unknown, and he might have got away with his story of being a Spanish petty thief, had it not been for a slight carelessness in his disguise’. He had dyed his hair black, but the dye had not been renewed, ‘with the result that the roots disclosed the natural red shade, while the ends remained jet black’. Following the Axis surrender in Tunisia in May 1943, Morris summed up the performance of the Dick Jones team as displaying ‘high grade “morale”’ but ‘low grade security’. The ‘whole “esprit”’ of the group was ‘based on great daring and disregard of danger, and these admirable and lovable traits brought them many recruits. On the other hand, this same daring (often amounting to rashness), together with a taste for good fellowship and good living, were largely responsible for their final undoing . . . In the type of work in which they were engaged, over-confidence is as great a pitfall as faintheartedness. ’ Morris drew ‘two outstanding lessons’ from the experience: ‘(a) what can be achieved under brilliant leadership, and (b) the inevitable results of over-confidence and lack of security precautions’.
Elsewhere, the price of discovery was equally grim. After Axis troops had occupied Corsica in November 1942, a number of SIS agents were infiltrated on to the island from North Africa. In April 1943, while waiting to rendezvous with a submarine, two agents from group ‘Auburn’ were arrested by Italian troops. One of the men, a British-born Belgian merchant navy officer called Guy Verstraete (cover-name ‘Vernuge’), who had unfortunately been carrying plans of the island defences as well as some English-language intelligence reports, was very badly treated by his Italian captors. ‘In spite of three months of the severest interrogation and the foulest tortures,’ reported a French colleague, ‘he never gave anything away.’ For a fortnight he was forced to use his chamber-pot as a container for food. His right leg was broken at the ankle, fingernails and toenails were torn out and cigarettes stubbed on his chest. He was sentenced to death by firing-squad in July. A report received in London the following October said that he placed his right hand on his heart and told the soldiers to ‘aim well’. Then, raising his left hand and declaring ‘Vive la Grande Bretagne,’ ‘he died a brave man [“il est mort en brave”]’.14
The Torch invasion of French North Africa precipitated not only the despatch of German troops to Tunisia, but also their entry into unoccupied France. This fatally undermined the Vichy regime and its hold (such as it was) over French overseas territories. Over the next few months there was an awkward period of uncertainty over who would assume the leadership of what had now become known as the Fighting French. Churchill had thrown his support behind Charles de Gaulle in 1940, but the United States, which had recognised the Vichy government and whose first major commitment to the war in the west was in North Africa, regarded him with suspicion. After Admiral Jean Darlan, the Vichy-appointed Commander-in-Chief of French armed forces, called a ceasefire on 11 November 1942, the United States commander, Dwight D. Eisenhower, had him confirmed as the political head of French North Africa, a move which disgusted the Gaullists. In the wings, moreover, was General Henri Giraud, who, having been taken prisoner at the fall of France, had escaped from Germany to Vichy in the spring of 1942 and appeared (especially among Americans) to represent an acceptably anti-Axis French leader.
Algeria was the first part of metropolitan France (as it was regarded by the French) to be liberated from Axis control, and after the ceasefire the political and military focus shifted to Algiers where Eisenhower established his headquarters. Giraud, having been smuggled with SIS help out of southern France on a British submarine, arrived on 9 November. By 12 November SIS had established a presence there under the code-name ‘Orange’, and Menzies in London pressed the unit for news: ‘in particular we wish to know nature of Giraud’s relations with Eisenhower, his attitude towards Darlan and the latter’s general orientation’. Two days later Algiers signalled that the Americans were backing Darlan as they hoped that through him Vichy supporters could be brought round to the Allied cause. Meanwhile Giraud was ‘somewhat unwillingly’ to be kept in the background. His staff were reported as being ‘fully aware grave danger [to Giraud’s ambitions] if Darlan can continue to convince 48-landers [the Americans] his value’. On 17 November, having managed to get out of France, the Deuxième Bureau chief, Louis Rivet, and his air intelligence specialist, Georges Ronin, arrived. Menzies asked Orange to assure Rivet ‘that his arrival is really heartening news’. Evidently anxious to keep the French liaison exclusive to SIS, Valentine Vivian told Algiers that he was sending out Winterbotham, head of the Air Section at Broadway, to meet Ronin. ‘Try and persuade him and Rivet’, he continued, ‘to keep clear of S.O.E. and other newly formed intelligence services’ (perhaps a reference to OSS, which had also set up a station in Algiers). ‘We trust they will stick to old friends who are, as always, ready to help them in every direction.’ So anxious was Menzies to keep the French out of SOE’s clutches that in January 1943 he asked for Rivet to be given an oral, personal message informing him that since ‘our previous association’ (before the fall of France) ‘sabotage’ had been ‘divorced from me’. A senior officer of SOE was to visit North Africa shortly ‘and I would like you to know that they have nothing repeat nothing to do with S.R. [intelligence]. I forbid the use of my agents or organisation for this purpose as disaster would inevitably follow.’
Ronin, who asserted that the Deuxième Bureau still had ‘considerable possibilities’ in France and the Low Countries, suggested that they might establish a liaison station in London and, following Winterbotham’s visit to Algiers, Ronin and Paul Paillole came to London for ten days over Christmas 1942. Here they met most of the SIS high command. Paillole recalled Dansey as being ‘rather strange’ (‘un curieux personnage’), but added that Menzies, who spoke with an air of authority, treated Dansey with respect. The two Frenchmen also met de Gaulle’s intelligence chief, André Dewavrin, amicably enough it seems, and had some preliminary discussion about co-ordinating activities.15 Menzies’s ‘ultimate hope’ was ‘to get one combined bureau amalgamated with [the] fighting French’. He recognised, however, that ‘this will not be easily attained’, and indeed co-operation between the various French organisations remained highly conjectural for some time. In the meantime SIS had to work with Dewavrin’s BCRA (Bureau Central de Renseignements et d’Action) and Rivet’s set-up largely as separate entities. The assassination of Admiral Darlan by a maverick anti-Nazi French royalist on Christmas Eve 1942 simplified the political situation and during 1943 de Gaulle gradually gained the upper hand. Giraud faded from the picture, and by the winter of 1943 Rivet’s organisation had been merged into the Free French BCRA.
All through 1942 coverage of Italy proved a very hard nut to crack. An Italian Jew recruited in Tunisia by André Mounier was given wireless training in Malta and landed on the west coast of Sicily by submarine. When he began transmitting, ‘we discovered by the check question method [by which the operator could use a prearranged answer to show he was under duress] that he had come under enemy control’. Rather than break off contact, Dudley Clarke used him as part of his brilliantly successful deception operation before Operation Torch, sending the agent ‘questions and warnings’ which implied that Sicily was to be the target rather than North Africa. A second operation involved the landing by submarine of two Italian ex-prisoners-of-war on the coast near Livorno in Tuscany. They had claimed to have anti-Fascist friends and that they would be able to pose as soldiers on leave, for which they were provided with false documents. But, as their case-officer recalled, ‘they were a very happy go lucky pair and we never heard of them again’, rather confirming Cuthbert Bowlby’s prediction in January 1941 that recruiting agents from among captured Italian servicemen might ‘merely be treated as a beneficial repatriation scheme’.16
By the spring of 1943, when it was clear that the defeat of Axis forces in North Africa was only a matter of time (the capitulation came on 13 May), pressure to get agents into Italy stepped up sharply. On 18 April London instructed Algiers that ‘considering the grave importance of southern Italy, you must take that as your first target’. But Algiers replied ten days later that the prospects were not very promising. ‘All our experience so far’, they signalled, ‘emphasises difficulty in finding suitable Italians willing to enter Italy or islands.’ They had consulted the Deuxième Bureau ‘who admit their prospects almost zero’. Sicily and the southern mainland would now be under intensive security, which would make it especially hazardous for operations. ‘Trained Italian agents have always been and are likely to continue to be so scarce that their quick loss in hot areas would appear uneconomic.’ The Auburn network in Corsica, which had been asked to attempt the penetration of Sardinia, was ‘now virtually wiped out’. But there was one hopeful note. Algiers viewed recruiting prospects in Tunis ‘with relative optimism’, as after its capture ‘truly patriotic Italians may take more stock in our victory. A few such should be worth many traitors or cranks.’
Early in May Menzies sent out his newly appointed Assistant Chief, Sir James Marshall-Cornwall, to review the situation on the spot. Perhaps the most important recommendation which Marshall-Cornwall made was the posting of an extremely able young officer, John Bruce Lockhart, to help out in Tunis. A St Andrews-educated schoolmaster in peacetime, a nephew of Robert Bruce Lockhart who had served in Russia during the Revolution and a keen player of real (court) tennis, John Bruce Lockhart had been recruited to SIS only in January 1942. Before becoming involved with counter-intelligence and double-agent operations in the Middle East, he had spent some time in Cairo working on German order of battle and (as he recalled years later) ‘six glorious weeks’ with ‘a Buick, a bag of gold and a heart full of hope’ in the Caucasus organising stay-behind teams in case the Germans penetrated through the mountains.
In June 1943, reflecting the shifting geography of the war, Menzies moved Bowlby and the regional SIS headquarters from Cairo to Algiers, where he was instructed to ‘ensure all liaison with British and Allied Staffs’. This, as is so often the case, was easier said than done. The French, having established their seat of government in Algiers, resented having to refer operational requirements back to London. The Americans were inclined to act entirely off their own bat, and in early July Bowlby worried that, without proper liaison with SIS, OSS would ‘undoubtedly turn to S.O.E.’ for assistance in running their own agents. He reported that the intelligence authorities at the Allied headquarters in Algiers were ‘most anxious’ for closer co-operation between OSS, SOE and ‘other interested parties’, such as Dudley Clarke’s A Force deception team, counter-intelligence and SIS. ‘My experience of working with the Americans in Mideast’, added Bowlby, ‘is that one must at any rate give the appearance of co-operating with them completely . . . To hold back in any manner is a game at which they are likely to be more adept than we, and I feel we should lose by it.’ While this observation has an air of cynical calculation, Bowlby also stressed the overall importance of co-operation. ‘I should like to be certain’, he signalled, ‘it is clearly understood in H.O. [Head Office] that all operational and intelligence activities in North Africa are carried out on a complete fifty-fifty basis. It is most unfortunate that an organisation such as ours should, by unwillingness to co-operate, fail to fall in with this general spirit so carefully fostered by the Commanders in Chief in North Africa.’ In July, Menzies, having spoken in London to Colonel Donovan of OSS, emphasised the ‘necessity for harmonious collaboration with Americans and of pooling our resources, training and transport’, and instructed Bowlby to set up a co-ordinating committee (with French representation) to deal with the problem.
The maintenance of good Allied co-operation was undoubtedly intended to reinforce SIS’s enhanced reputation, which was itself underpinned by the high quality of operational intelligence the Service had been able to supply from Tunisia. Although an SIS group (with the cover name of ‘No. 1 Intelligence Unit’) was attached to Alexander’s 15th Army Group, events moved too quickly following the invasion of Sicily on 9-10 July 1943 for it to replicate SIS’s performance in North Africa, even had it been able to draw on any existing networks of agents. Enemy resistance ceased in Sicily on 17 August, and already by the end of July John Teague in Cairo was proposing that SIS should aim to be ‘ahead of operations. We should’, he said, ‘look over Italy to Austria.’ As soon as a station was established on the Italian mainland, it should make the penetration of Austria and south Germany ‘one of its primary duties’.
On 23 August Menzies raised the role of SIS with Alexander. The ‘original intention’ was that it ‘might be useful for providing tactical information by sending Agents through the line’, but this had not happened and Menzies’s opinion now was ‘that proper S.I.S. function is the provision of Intelligence on a long term basis provided by agents working on the enemy’s lines of communications’. Alexander agreed. ‘Up to present in Sicily’, he cabled on 29 August, ‘S.I.S. unit has not been able to provide any reports of tactical importance and in my opinion it is not likely that they will be able to do so in future except in the case of prolonged static conditions.’ He recommended, therefore, that SIS units be ‘retained under centralised control for work of strategical interest’.
After Allied forces had landed on the Italian mainland at Reggio and Salerno, and Marshal Badoglio’s government (which had ousted Mussolini in July) capitulated on 8 September 1943, hopes were high that the Germans might withdraw, at least to the north of the country. But the German commander, Field Marshal Kesselring, resolved to resist, and the Italian campaign became a costly slogging match as the Allied armies inched their way up the peninsula from late 1943 and through the following year. During this campaign SIS, in addition to developing its own organisation in Italy, ran a second set-up in conjunction with the Italian secret service, SIM (Servizio Informazione Militare), now working for the Allied cause. One early, and perhaps disturbing, result of this was the news from SIM officers in October that their cryptographic section had previously broken the ‘diplomatic cyphers’ of United States military attachés, the ‘Russian confidential cypher’, and those of Turkey, Yugoslavia, Greece, Portugal, Egypt, Brazil and Belgium. The results had been ‘passed to Germans and Hungarians’ who could ‘now presumably decypher all above cyphers’.
In October 1943 Bruce Lockhart was posted to take charge of No. 1 Intelligence Unit, by this time based at Bari on the southern Adriatic coast of Italy. He found the unit in bad shape. Agent recruitment, training and security were all very lax, and he reported (in a letter which Bowlby described as ‘helpful and honest . . . a rare combination in this racket!’) that it was ‘common gossip among the rougher element in the town that I.S.L.D. is the easiest way of making soft money (what A.C.F. [the African Coastal Flotilla which provided sea operations support] refer to as “our repatriation scheme”)’. Bruce Lockhart rapidly began to sort out the unit and build it up into a very efficient operation, assisted by ‘five or six SIS pre-war secretaries . . . who were absolutely first-class’. He was notably good at managing the personnel under his command, and keenly aware that sustained hard work had to be balanced with moments of recreation, as illustrated by one response to a request for equipment. ‘Hope to have a rugger ball for you shortly,’ signalled London, ‘and some tennis balls. Please remember these are not easily come by. We do our best.’17
Under Bruce Lockhart’s command Bari became the base for SIS operations, first into German-occupied northern Italy, and then also into the Balkans and Central Europe. From late 1943 various schemes were mooted for the penetration of ‘southern Germany’, effectively, in fact, Austria, which had been integrated into the Reich since 1938. One involved the recruitment of Austrian and Italian prisoners-of-war in North Africa and their repatriation by means of an existing escape route through Spanish Morocco, Spain and France. Another suggested recruiting agents among partisans in northern Croatia and smuggling them into Lower Styria. Two-man agent/wireless-operator teams would be dropped in blind to establish ‘advanced bases’ (one was proposed to be in ‘the Brenner Pass area’; another on the Slovene-Austrian border) to which agents would report. Nothing came of this ambitious scheme, and in February 1944 Bruce Lockhart, while conceding that his section (‘44200’) dedicated to penetrating Germany was under-performing, noted the real difficulties with which it was faced. There was a shortage of air support; blind-dropping agents was ‘useless’; some of the proposed agents were unsuitable. One, indeed, could ‘not possibly pass as Italian due to Prussian accent’. Bruce Lockhart thought that better results might be obtained by infiltrating foreign labourers and specialists. One of his colleagues complained about the poor quality of the Austrians he had been given. ‘The supply of even half-way suitable types is strictly limited,’ he observed, and, even if he sent some off, he doubted if he would ‘ever hear from them again’. ‘Leaving aside all wishful thinking’, and drawing a striking parallel, he argued that SIS was ‘in much the same position, even at this stage of the war, as the 12-land [German] service when they are looking for 22-landers [Britons] to work for them in 22-land’. SIS of course got ‘plenty of rats that leave the sinking ship, and we shall get more, but precious few are prepared to go back to gnaw another hole in her bottom’. Not much was achieved until the spring of 1944 when two largely self-contained sections were set up. The first, for the penetration of northern Italy, was under Major Brian Ashford-Russell and the second, for Yugoslavia and Central Europe, under Major James Millar.
During the Second half of the war SIS in several places dealt directly (and productively) with Communists. Across the Balkans both SIS and SOE found themselves working against Axis forces with Communist resistance and partisan groups. But the Western Allies were also supporting and mobilising centrist and right-wing elements in occupied Europe, and as victory began to seem increasingly likely, tensions also began to emerge between domestic political groups, looking to winning not only the war but also the peace that would follow. Furthermore, as Soviet armies began to roll back the Germans and move towards Eastern Europe, concerns were raised, both in London and by the men on the spot, as to the political consequences of this advance, in both the short and the long term.
Greece was a case in point, where resistance coalesced into two main groups: the rightist National Democratic Hellenic League, EDES, broadly aligned with the Greek government-in-exile, and the Communist-backed National Liberation Front, EAM, with its military wing, ELAS, the National Popular Liberation Army. British engagement with these groupings, at least in 1942-3, depended less on any perceived political leanings than on their efficiency in fighting the enemy. Responding in August 1942 to Greek complaints about the ‘British Secret Services’ (which included both SIS and SOE) consorting with elements ‘hostile to the King and the Government’, for example, SIS recognised that it was ‘only natural’ that SOE’s ‘operations are connected with subversive elements’. If ‘they are using Communists, which we have every reason to believe is correct, they have chosen these men because they are more active against the Axis than the other elements in the country’. At this stage SIS’s main Greek operation was based in Smyrna, where Lieutenant Commander Rees was running boats into Greece, supporting escape lines, but not producing much intelligence from the Greek mainland. All through 1942 there were concerns about the lack of information. Menzies (as always) was being pressed for military intelligence by the service departments, while the Foreign Office wanted better reporting on the admittedly complicated and confused political situation. In Cairo an office was established with the main function of penetrating Greece, not only from North Africa but also through Smyrna. Matters improved in 1943 when (among others) a network of agents run by Rees – called ‘S’ sources – began to report. ‘S.41’ was recruited in September 1942 and sent military and naval information from Salonika, much of it based on train-watching. Another agent, ‘S.30’, who had previously worked for the Italians against the Turks, was regarded as ‘a most valuable source’ who in 1943 alone, up to his arrest in July, had supplied thirty-nine reports on the Dodecanese, including air, army and naval information which had received ‘very favourable criticism’ from the service ministries.
In 1943 some SIS officers were dropped into Epirus in north-western Greece, primarily to report on German order of battle in Greece. One of these was Nigel Clive, who had been recruited by the Service in December 1941 and had served as assistant to the Baghdad representative, before engineering a move to Cairo in a successful effort to get a more active job. Equipped with a money belt containing a hundred gold sovereigns and under the pseudonym ‘Jim Russell’, he parachuted into Greece in December 1943 along with a wireless operator and a regular Greek army officer to help with liaison. Clive’s duties were not particularly covert. Formally attached to SOE’s Allied Military Mission at EDES headquarters, he served in British battledress and lived openly among the partisans in the Greek mountains. As Clive claimed afterwards, one unusual circumstance of his posting was that his SIS predecessor, an alleged Greek-American called Costa Lawrence, had been shot dead by an SOE officer, an Irishman named Spike Moran. Although Clive was well aware of the intermittent tensions between SIS and SOE, this seemed to take things too far, and it emerged that Moran had believed Lawrence was a traitor and thought that he was about to be betrayed to the Germans. Moran himself and his SOE colleagues were actually very helpful to Clive, who was able to collect a great deal of information about the German forces, finding Greek interpreters working for the enemy to be especially willing to assist. Clive remained in Epirus until the German withdrawal from Greece in the autumn of 1944, and from April 1944 he began also to report on the political competition between the rival Greek resistance groups, in effect predicting the civil war which erupted at the end of the year.18
Turkey, meanwhile, remained a lively centre of intelligence activities during the second half of the war. The enemy scored a notable success with the famous spy ‘Cicero’, Elyesa Bazna, Albanian valet to the British ambassador in Ankara, Sir Hughe Knatchbull-Hugessen, whose sloppy attitude to security enabled Bazna in 1943 systematically to photograph highly classified documents and sell them to the Germans. SIS were involved in the investigation after the leakage of secret information from the Ankara embassy was revealed both by signals intelligence and by an OSS agent in the German Foreign Ministry, but the full story did not emerge until after the war.19 At the same time as the Cicero crisis, SIS struck an outstanding blow against the Abwehr by recruiting and organising the defection of Dr Erich Vermehren and his wife Elisabeth. Vermehren, an anti-Nazi Roman Catholic who had won a Rhodes Scholarship to Oxford just before the war (but had been prevented from taking it up on account of his refusal to join the Nazi youth organisation at school), had been assistant to Paul Leverkühn, the senior Abwehr officer in Turkey, since October 1942. His wife was a cousin of Franz von Papen, the German ambassador in Ankara, and had family connections with Adam von Trott zu Solz, a leader of the July 1944 plot to assassinate Hitler.
The Vermehrens’ case officer was Nicholas Elliott, who had been Section V representative in Istanbul since the spring of 1942. An assessment after the war recorded that Elliott handled the Vermehrens with ‘consummate skill and sympathy, but with just the necessary touch of firmness’. Vermehren had got in touch with the British assistant military attaché in Istanbul, who passed him on to Elliott. The first contact was made on 18 January 1944 and on 21 January Elliott told Felix Cowgill in London that he had recruited Vermehren, whom he described as a ‘highly strung, cultivated, self-confident, extremely clever, logical-minded, slightly precious young German of good family’. Vermehren, he said, was willing to work for SIS as he was ‘intensely anti-Nazi on religious grounds’, and also because his wife and her family had been persecuted by the Nazis. Vermehren had produced ‘a quantity of detailed information’ about the Abwehr organisation in Turkey, which had ‘fully convinced’ the SIS station of his bona fides. Over a four-day period in late January, Vermehren managed to bring out important Abwehr files for SIS to photograph, including an organogram of the complete Abwehr set-up in Istanbul, and he was able to give SIS details of current Abwehr operations in Turkey and the Middle East.
On 25 January Elliott was tipped off by a contact in the Turkish secret police that they knew Vermehren had been in touch with the British. Reasoning that it would not be long, therefore, before the Germans got wind of this, with Turkish assistance he arranged for the Vermehrens and two other anti-Nazi Abwehr officers to be smuggled out to Cairo through the SIS station at Smyrna. Because of Vermehren’s supreme access, his defection completely compromised and demoralised Axis espionage throughout the region. Leverkühn, who was reported to be in ‘a hell of a flap’ over Vermehren’s flight, was recalled in disgrace to Berlin. Apart from its impact on German operations in the Near and Middle East, the Vermehren defection had a very significant impact on the struggle for intelligence supremacy in Germany between the Abwehr and Himmler’s Sicherheitsdienst, which ended with the latter absorbing the former. It was also one of the factors leading to Canaris’s downfall. As Michael Howard has concluded, the entire German intelligence service ‘was thrown into a state of confusion just at the moment, in the early summer of 1944, when its efficient functioning was vital to the survival of the Third Reich’.20
In Bucharest the productive head agent, Nannygoat, continued to provide a variety of intelligence. In the summer of 1942 he was asked to confirm the identification of five Italian divisions deployed in Romania. The following year he supplied a ‘news reel film purporting to show fortifications on French coast’ and reported that a sub-agent claimed to ‘know German engineers willing [to] desert bringing new bomber range 3,000 kilometres’. ‘Shall I make further enquiries,’ he asked in July 1943. ‘This is probably the Heinkel 177,’ replied London: ‘ZA [Air Ministry] very interested. Please go ahead.’21 Nannygoat’s contact seems not to have delivered the wavering German engineers, but the exchanges between Head Office and Istanbul demonstrate the range of information which any one agent might provide.
Nannygoat’s network also illustrates the polyglot nature of wartime intelligence-gathering. He himself had originally been a French agent, and his own organisation included several different nationalities. This, of course, could bring risks. In November 1943 he acquired some Polish wireless operators, three of whom were arrested the following February ‘due to discovery in Warsaw of Polish organisation working to Bucharest’. By the spring of 1944 Nannygoat (no left-winger he) was wiring ‘for instructions as to what course to pursue in the event of Russians arriving in Bucharest before British’. ‘In principle,’ replied London, ‘we would like as many agents as possible to remain undisclosed to Russians but feel that those who are likely to be arrested should be protected by us.’ In June 1944, however, ‘after denouncement by two Frenchmen’, Nannygoat was put under house arrest by the Romanian authorities, who suspected him of working for the British. But in the confused Romanian politics of the moment, with the Germans in retreat and the Russians on their doorstep, ‘working for the British’ could be an advantage. As reported afterwards, Nannygoat secured an interview with Colonel Traian Borcescu, head of Romanian counter-intelligence. He admitted that he ‘was head of a British Intelligence Service in 14-land [Romania] and warned him that if he and his collaborators were not released Borcescu and his colleagues would have to suffer for it’. This ‘produced an electric effect’. Nannygoat and some of his agents were released and put on a plane for Istanbul on condition that he would if possible try to secure British help for Romania.