by Alastair Denniston- Code-breaking From Room 40 to Berkeley Street
***
By the end of 1941, the Diplomatic Section at BP, which AGD directly controlled, had grown to eighty staff which included sixty cryptanalysts (twenty-nine Senior and Junior Assistants, including temporaries and thirty-one Linguists or Clerical Assistants), and the remainder typists and other ancillary staff. The main country sections were Italian, Japanese, Near Eastern, Chinese, Balkan, Portuguese and a French Section enhanced by eight staff to work on Vichy government ciphers. In all, they worked on the traffic of twenty-six countries, received around 100,000 telegrams, read 70,000 and translated and circulated 8,495.110 At least an additional fifty staff were required but working space was not available. So there was little change to the Diplomatic Country Sections apart from a small increase to the Italian and French Sections, bringing the establishment up to 100. Early in the year liaison was established with US Diplomatic Agencies and later a joint effort began to break and exploit German diplomatic traffic.
The America Army Security Agency (ASA) had more staff working on Italian diplomatic traffic, but GC&CS had more experience and knowledge, so it could provide them with codes more fully recovered than theirs along with information about enciphering tables. The Americans in due course provided help in recovering keys and had at their disposal more machine tabulating equipment. Regular exchanges of traffic registers, wanted material and code recoveries were in force by the beginning of 1942 and continued until the end of the war.
After the Americans had provided GC&CS with at least one replica ‘Purple’ machine in February 1941, there was joint collaboration in recovering the current settings of the machine. This work required both a specialised team of cryptanalysts and a greater number of linguist/translators. Once Japan entered the war, many former consular and diplomatic officials returned to Britain and were drafted in to help with the task.
Up until the Japanese attack on Pearl Harbor on 7 December 1941, GC&CS’s Diplomatic Section had been reading US State Department cipher messages. On 25 February 1942, in a personal letter to Roosevelt, Churchill wrote:
My Dear Mr. President,
One night when we talked late, you spoke of the importance of our cipher people getting into close touch with yours. I shall be very ready to put any expert you care to nominate in touch with my technicians. Ciphers for our two Navies have been and are continually a matter for frank discussion between our two Services. But diplomatic and military ciphers are of equal importance, and we appear to know nothing officially of your versions of these. Some time ago, however, our experts claimed to have discovered the system and constructed some tables used by your Diplomatic Corps. From the moment when we became allies, I gave instructions that this work should cease. However, danger of our enemies having achieved a measure of success cannot, I am advised, be dismissed. I shall be grateful of you will handle this matter entirely yourself, and if possible burn this letter when you have read it. The whole subject is secret in a degree, which affects the safety of both our countries. The fewest possible people should know.
I take advantage of the Ambassador’s homeward journey to send you this by his hand, to be delivered into yours personally.111
While collaboration between Britain and the US was going well on the diplomatic front, AGD was becoming concerned that the Americans were going back on their promises to him in August. On 23 December, he wrote to Washington with his concerns:
In a telegram from War Department, A. 16 of 18th December they raise the question of investigating the German Air-Army cypher. During my visit it was agreed that we should be responsible for this investigation and that when USA were in real need of this work we should invite their party to join ours.
Could you find out if their views on this procedure have changed and if they wish to begin their own investigations now? It is devoutly hoped by all here that any such investigations will not interfere with their progress on Japanese work for which we count on them.
We could send by bag two days of traffic and certain keys (all information on this being sent in cypher) but bag communication is very slow.
Could you also find out if they are intercepting any of this material? I am sending by bag material for German air to ground traffic.112
This development prompted AGD ‘to organise, without delay, an interservice Japanese Section comprising all phases of Japanese work, ie, Diplomatic, Naval, Military and Air’.113
By the end of 1941, the pressure on AGD was growing with political battles continually being fought with the Service ministries and internal staffing and administration problems a constant concern. Inter-Service difficulties were also ever present and according to Birch:
Throughout the history of Sigint in the Second World War there is noticeable the pursuit of satisfactory compromises between two pairs of alternative and mutually incompatible ideas: - Service versus inter-service structure of Sigint and centralisation versus decentralisation of cryptanalysis. In these early days, whenever the inter-service pooling of some Sigint function was mooted, one or other of the Services was apt to discover that its needs were different from those of the other two. When inter-service coordination of D/F was proposed, ‘the Navy and Air Force had to deal with rapidly moving targets, whereas the stations the Army would try and place were either fixed or capable of only a small movement from day to day’.114‘ In the matter of ‘Operational Intelligence’, as we shall see, the Admiralty was ‘different’, in that it was ‘an executive organisation’ and therefore needs an O.I.C. inside itself, whereas the Army and Air Force need rapid dissemination from the source (GC&CS) to their executive commands.115 So, too, when an Air Section was set up in GC&CS in 1936, Air Ministry felt that it had been ‘formed for a somewhat different purpose to the other two’.116
In January 1942, DNI John Godfrey wrote:
BP has grown in a haphazard way out of small body of research workers to a heterogeneous establishment numbering some 1,500 persons. I am urging ‘C’ with some success to put his house in order and to tackle the administrative problems involved, and what I hope to aim at is that there should be a Senior Officer in administrative charge of the whole establishment, who will look after administration, security, welfare, feeding, housing, etc. … leaving Denniston and the technicians to pursue their highly important specialities.117
The problem according to Birch was that ‘the Y expansion programme was out-of-date before it was fulfilled, and the Y communications plan was carried out only partially and tardily’. Furthermore, criticisms of GC&CS were driven by a paradox: ‘The Services controlled Y, and GC&CS, under CSS, controlled cryptography, but Y and cryptography, although separate, were inseparable.’
The Army seemed to play the leading role in the battle for Sigint control. Enigma cryptanalysis was outside the jurisdiction of No. 4 IS, MI8’s Military Section at GC&CS. The Enigma decrypts went from Hut 3 to MI14 in the War Office via MI6/SIS, bypassing MI8. Enigma traffic analysis was the function of MI8’s No. 6IS but the major part of it was still at Beaumanor due to lack of accommodation at BP. While MI8 was responsible for all Army interception including Enigma traffic, it seemed sensible that interception policy should be driven by General Staff requirements. However, interception was, in practice, arranged by GC&CS and there was nothing in place for any sort of control by the General Staff.
DMI made a proposal in autumn 1941 concerning the operational control of Sigint. The Directors of Military Intelligence met weekly before seeing the Chiefs of Staff, and Sigint policy for the week could be set at that meeting. The Directors thought that they should be supplied with weekly reports and returns from GC&CS through the German Sections at the War Office and Air Ministry, supplemented by W/T reports from MI8.118 However, the proposal was challenged by Welchman on behalf of Hut 6 and Hut 8:
The operational control of the sets allotted for E work is an hour-tohour business which should be left in the hands of experts, provided of course that these experts follow the general policy laid down by higher authority and report any emergency
action that they have taken. This operational control could not be done efficiently by a body of men meeting only once a week and provided only with MI8 returns on which to base their decisions.119
Welchman’s views were supported by the Y Board at their meeting of 24 September,120 and there was no support from the other two Services. Naval Sigint in GC&CS worked well with NID and all naval decrypts and intelligence, either current or research, were passed by Naval Section to the Admiralty. As de Grey wrote ‘The friction over responsibilities and divisions of labour which wasted so much time in the other Services, never occurred in the relations of GC&CS and the Admiralty.’121 The Air Ministry also had few complaints, since almost the whole of the huge output of the Hut 6/Hut 3 operation concerned the GAF. There was, however, a very general Service dissatisfaction with GC&CS, based partly on the separation of Y and cryptography, the position of CSS between the Services and GC&CS, overall shortages of staff, equipment and facilities, and on what appeared to be managerial weaknesses at BP.
As Birch says:
It had been intended to use Bletchley Park for a cryptanalytic bureau and, in the interests of security, to keep the bureau as small as possible. A great deal of the ensuing trouble may be ascribed to the retention of a parochial attitude long after the march of events had proved the inevitability of the development of ‘the Park’ into a global Sigint centre. The difficulties of recruiting civilians were common to many departments, but in GC&CS they were aggravated by abnormally low rates of pay and bad living conditions in the first years, and later by the operation of the call-up of women. Anomalies arising from the employment of many Service and civilian grades on substantially the same work needed constant reconciliation. The provision of working quarters, billets, meals and transport lagged increasingly behind requirements. At last, in July 1941, it was agreed as a matter of GC&CS policy to build two-storey brick buildings instead of the temporary wooden huts hitherto provided, but pending their construction, the limit of accommodation had been reached by the end of that year, and by the time they were ready for occupation, they were inadequate for the further expansion which had meanwhile become necessary.
The JCC had been set up in April 1941 to handle the general administration of BP. In effect, the administrative staff of CSS took over control of administration.122
According to historian Christopher Andrew, ‘Without the expertise painstakingly built up at GC&CS on minimal resources by Denniston between the wars, Bletchley Park’s wartime triumphs would have been impossible.’ However, now others argued that AGD did not have a great vision, that he only wanted a restricted remit for GC&CS, i.e. cryptography, not the wider role of Sigint and all that entailed. The reality was that the two were inexorably bound together, but the Service Ministries insisted that it was their job to determine what intelligence was useful and what wasn’t as well as evaluating it. If they sensed that they were being second guessed by the civilians at GC&CS, they would move to take over the cryptography role as well, making GC&CS a small research section. So AGD insisted, at least in any document that he produced, that the role of GC&CS should be restricted to cryptography pure and simple. In reality, he knew full well that their work was indivisible from other Sigint activities. His strategy would eventually win BP the right to be in total command of Sigint.
In early January 1942, Menzies, having received the report from his independent investigator, General K.J. Martin and in agreement with the DMI, invited Brigadier W.L. van Cutsem, formerly DDMI (I), ‘To report to the Director and Head of GC&CS on any means of improving the military information derived from the cryptographic work and its flow to the military authorities interested. The report is only to deal with existing machinery, without suggesting any major change.’123
The Brigadier decided ‘to extend the inquiry beyond the purely military side, in order to gain an insight into the organisation as a whole and to compare the systems in the Air and Naval Sections with that in the Military Section, with a view to seeing what features in the Naval and Air Sections might profitably be adopted in the Military’. He also decided ‘to enquire into the administration of the GC&CS, for … it is evident that administration plays a direct and important part of the efficiency of the work carried out’. His report was presented on 30 January at a special meeting of the Director, ACSI and Travis.124 Menzies acted quickly and, by 3 February, issued instructions for a radical reorganisation of GC&CS:
Reorganisation GC&Cs
With the ever increasing work, I have found it necessary to carry out a reorganisation of the GC&CS.
The posts of Head of the GC&CS and Deputy Head of the GC&CS have been abolished and the work of the GC&CS will henceforth be divided into two parts:
Civil
Services
Commander Denniston and Commander Travis are appointed Deputy Directors to control the Civil and Services’ sides respectively, with the titles of Deputy Director (C) and Deputy Director (S).
Civil Side:
The Deputy Director (C) (Commander Denniston) with headquarters at Wavendon, will control this side of the work, i.e. the Diplomatic and Commercial Sections.
Mr. Earnshaw-Smith is appointed Assistant to Commander Denniston, with the title Assistant Director (C).
Service Side:
The Deputy Director (S) (Commander Travis), with headquarters at B/P, will control the Service Sections, including Hut 3, ISOS and ISK [Intelligence Services, Knox] , and will exercise general control at B/P and Elmers.
Mr. N. de Grey is appointed Assistant to Commander Travis with the title Assistant Director (S).
Colonel Tiltman, in addition to his duties as Commandant No. 4 IS, will act as Chief Cryptographer and will take charge of the Research Section.
The Deputy Director (S) will represent the GC&CS at the Y Board and for the time being at the Y Committee. Colonel Tiltman will be responsible for Liaison with the FECB and the USA Bureaux. Mr. Cooper will be responsible for Liaison with the CBME.
Administration:
Administration will be unified under Paymaster Commander Bradshaw, who is appointed as Assistant Director with the title Assistant Director (A). He will be responsible to the Deputy Directors for meeting their requirements. The JCC is abolished.
Effect is to be given to the above directions with the minimum of delay.
Heads of Sections are to inform members of their sections of this reorganisation.
(Signed) [S.G.M. Menzies, Chief of SIS]
Director125
On 30 January 1942, AGD wrote to Tiltman, Bradshaw, Cooper, Birch, Malcolm Saunders, Denys Page,126 Knox, de Grey, Eric Earnshaw-Smith127 and Henry Maine:128
In consequence of the great expansion of G.C. & C.S., the Director has decided to abolish the posts of Head and Deputy Head and to nominate myself and Cdr. Travis as Deputy Directors and to divide the organisation into two parts:
The Civil side including Diplomatic & Commercial Sections. The Services side including Naval, Military, and Air Sections and ISOS and ISK.
The Civil side will be under my direction with Mr. Earnshaw-Smith as my assistant. Cdr. Travis will control the Services side with Mr. de Grey as assistant.
The Director will circulate his decision as to General Administration in a few days.129
New terms of reference were also approved by the Y Board for the Y Committee on 5 February 1942.130 Birch later noted in his official history that:
What strikes one most about comparing the old with the new is a change of outlook, a better understanding and a broader perspective of Sigint. It is, for instance, no longer ‘the needs of the three Services, and of the GC&CS’ that need co-ordinating, but ‘the activities of the Y services’, and increasing concern is shown with developments overseas, the Y services in the Dominions and the Y Committees now established in Cairo and at Gibraltar, liaison with the Allies and Sigint.
A number of factors led to the restructuring of GC&CS and AGD’s subsequent removal from BP. The Service Ministries had continued to expr
ess concerns about GC&CS and in particular, AGD’s view that all service cryptanalysis work should be centralised at BP. The problems within Hut 3 reflected badly on AGD as he was ultimately responsible for the smooth delivery of BP’s ‘product’ to them. The Americans continued to apply pressure and it was proving difficult to secure agreement on the best division of labour in intelligence work between the two countries across multiple theatres of operation. The ‘Action This Day’ letter certainly would have been noted in Whitehall as further evidence of AGD’s inability to effectively run the BP operation. Interestingly, Gordon Welchman, who had written the letter, believed that it was ill health that had forced him out.131 According to the Official History of MI6, AGD’s removal from BP by Menzies was a demotion.
In the years that followed, former colleagues of AGD began to give their views on his ‘demotion’. Stuart Milner-Barry, who had co-signed the letter to Churchill and then personally delivered it to him at Downing Street, told the historian Ronald Lewin that by the winter of 1940, Denniston was a ‘busted flush’ and incapable of the organisational effort that was necessary if BP was to be put on a war footing. AGD was obsessed by secrecy and had a good relationship with Sinclair, who after all had appointed him in the first place. Sinclair’s successor, Menzies, was a WW1 hero who conducted most of his MI6 business at White’s Club in St James’s and was a different proposition. In his biography of Menzies, Anthony Cave Brown claimed that AGD and Menzies had remained in ‘close and friendly association’ between the two world wars. He also believed that AGD’s removal from BP was one of Menzies’s ‘unhappier decisions’. However, during WW1, Menzies was an intelligence liaison officer between GHQ and the Directorate of Military Intelligence and MI1(c). While he was only a member of the Army in principle, he may well have felt that the Army’s man, Malcolm Hay, should have been appointed as Head of GC&CS over the Admiralty’s nominee, AGD. However, there is no evidence that Menzies knew about AGD being chosen over Hay for the role of Head of GC&CS. In any event, the view that he never fully supported AGD would be strengthened by events that transpired towards the end of WW2.
