CHAPTER 16

Australia’s First National Signals Intelligence Effort

On 3 September 1939, Australia’s armed forces were in no position to fight a modern war against an advanced enemy like Germany. The best they could do was support British forces in Europe, the Middle East and the Mediterranean. In Australia, the small cadre of professional sailors, soldiers and airmen struggled to put the war plans into operation and oversee the mobilisation of the volunteer forces that would serve overseas while bringing the conscripted militia onto a war footing, albeit one restricted to the home front. The administrative overhead associated with this was significant and demanding. To make matters worse, there was no central defence organisation, no supply system, no instructional schools beyond those used for training local militia, no command and control system that could take charge of both the field forces and the national defence, no communications system worth speaking of and no intelligence system, just small intelligence organisations within the forces.

As we have seen, Naval Intelligence Section was the most sophisticated of the service intelligence organisations but it lacked manpower everywhere, with the exception of its honorary coastwatcher and reporting officer systems. The army’s Military Intelligence Branch had lost itself in internal security intelligence matters that it had neither the legal right nor aptitude to manage. Other than this, it was scrambling to fill the positions in the 2nd AIF and its wartime establishment at home. The RAAF intelligence capability was minute, inexperienced and much less developed than the intelligence entities of the other two services.

All this meant that the three service intelligence branches were disconnected from each other and served their respective services, not the broader strategic needs of the nation. There were no mechanisms for coordinating intelligence collection, analysis assessment or distribution at an inter-service level, let alone at a whole-of-government level. There was no vetting and, as we have seen, there was precious little security for HUMINT activity or any other secret activity of government in Australia. In fact, Australia’s intelligence system was closer in character to that of Japan than to that of Britain.

SIGINT was different. First, Australia did not control ‘Y’ Procedure, as SIGINT was then called. ‘Y’ Procedure was controlled by the ‘Y’ Committee in London and by a number of specialist subcommittees. The foundation for the centralisation of cryptanalysis and SIGINT operations under GC&CS was laid in 1922, when the British armed services agreed to second their cryptanalytical staff to GC&CS and put them under the control of the Foreign Office.1 In 1923, the head of SIS was renamed the chief of the secret service and the director of GC&CS. While GC&CS remained separate from SIS, this put the same official in charge of both foreign intelligence organisations. Because of this administrative arrangement and, more importantly, the financial constraints imposed by the economic circumstances of the interwar years, the British armed services were forced into pooling resources and agreeing to a more sensible division of responsibilities under the direction of the ‘Y’ Committee. Thereafter, the influence of the Foreign Office and SIS over British SIGINT became more pervasive, if not more active.

The establishment of the Joint Intelligence Committee (JIC) on 7 July 1936 reduced even further the authority of single services over SIGINT collection.² According to F.H. Hinsley, the British official historian, the JIC was established to provide support to the new Joint Planning Staff by coordinating all of the inter-departmental reports and intelligence appreciations, and to coordinate the tasking of the intelligence services on behalf of a growing number of government departments that were now required to run an economy and government preparing for total war. This whole new structure provided support to the Cabinet, the Committee of Imperial Defence, the Chiefs of Staff Committee and all of the other organs of state involved in planning for or managing defence.

The other factor that influenced the willingness of Britain to allow Australian or any other Commonwealth country to involve itself in high-level SIGINT was the extreme sensitivity of SIGINT to compromise. This sensitivity had always existed, but it was seriously exacerbated by the Hall–Peaslee compromise of August 1925. As we saw earlier, this involved Vice Admiral William Hall, the wartime head of Room 40, handing more than 10,000 highly secret Room 40 decrypts to an American lawyer, Amos Peaslee, acting on behalf of American companies seeking damages for their losses to German saboteurs during World War I. Hall appears to have taken a substantial number of Room 40’s highly secret files home with him after the end of the war, and it was from these that he provided Peaslee with the decrypts. By March 1925, the decrypts were in the hands of the German government, the defendants in Peaslee’s lawsuit, and, according to a National Security Appreciation conducted of the incident sometime after 1955, the Germans responded immediately by introducing one-time pads and purchasing and modifying the stickered ENIGMA machine.3

It is easy to see why Admiral Ragnar M. Colvin, RN, may have been less than keen on an Australian national SIGINT effort, and why the authorities in London were less than enthusiastic in supporting any such action.

The sensitivities in London were not germane to the concerns held by Australian officers and civilians in government about the poor state of Australian intelligence in 1939. One of those most concerned was, as we now know, the DNI, Commander Rupert Long. With the declaration of war by the Menzies government, some impetus was given to dealing with these concerns, even if the government did not back its rhetoric with the money the services needed to develop an effective intelligence system. In late November 1939, as the RAN struggled to find enough sailors to man its routine communications, Long laid the problem of SIGINT before the Naval Staff, particularly Rear Admiral Colvin.⁴ Long, an advocate of the joint approach the British had implemented, had been pushing for combined operational and intelligence structures.

The initiative for a joint operations and planning organisation in Australia came from the RAN, specifically from Long, but it was brought to the Defence Committee on 26 January 1939 by Admiral Colvin.⁵ Colvin then ensured the idea was delayed by agreeing to have it considered by a Joint Planning Committee (JPC), which would report to the chiefs of staff on the issues surrounding the proposal.

On 6 August 1940, the JPC recommended that a Central War Room be established in H Block of Victoria Barracks in Melbourne, and that the COIC be placed next door. The Central War Room was to be the joint HQ from which the three service chiefs and senior ministers could run the strategic direction of the war. It would be continually manned, although the attached COIC would not.

The recommendations made by the JPC were sensible enough, but did not address the needs of civilian organisations or the political leadership. Nowhere was it mentioned that the prime minister and Cabinet ministers might need to be involved in decision-making if the feared attack on Australia took place. There was also a complete lack of consideration as to how the military authorities would manage the impact of operations on the civilian population, and how military and civilian agencies and departments would cooperate.

Despite the Defence Committee endorsement and discussion in Cabinet in January 1941, the recommendations for a Central War Room and a supporting COIC went nowhere. The Central War Room was made unworkable by the very restricted space it was given at Victoria Barracks and because it was given no equipment. It was also made redundant because the chiefs of staff chose to meet elsewhere.6 Even as Malaya was falling to the Japanese, Australia still had no effective coordination between the three services.⁷ As to intelligence, on 27 January 1942 the Chief of the Air Staff (CAS), Air Vice Marshal S.J. Goble, was writing to the Secretary of the Defence Committee complaining that there was no effective COIC, and ‘there still appears to be a disinclination on the part of the intelligence sections of the three services to use this combined organisation’.⁸

SIGINT was a different matter than joint intelligence or operational control centres, as the Australian armed services in the first years of the war regarded SIGINT as a single-service matter. The question that needed to be answered on the approach of war in 1939 was how SIGINT was to be managed in Australia and who was to be responsible for strategic SIGINT collection.

The subject had been brought before the chiefs of staff on 12 December 1939, when Admiral Colvin dutifully tabled his DNI’s minute on the need for an Australian cryptanalytical organisation to deal with traffic from enemy raiders that might operate in Australian waters. Having tabled Long’s minute, Colvin then spoke against his own DNI’s proposal.9

Long’s minute is unremarkable. He was recommending that Australia approach the relevant British authorities for advice on whether Australia should establish its own cryptanalytical organisation. If the British authorities supported this, Long suggested they be asked to supply some of the specialist personnel and equipment that Australia lacked. It may have been this recommendation that put Colvin off. As an RN officer, Colvin would have been well aware that the philosophy at GC&CS was for cryptanalysis to be concentrated in one location, GC&CS, and that any request for cryptanalysts and specialised equipment to be sent to Australia would fall on deaf ears in London.

In his own minute, Colvin carefully undermines Long’s suggestion. He uses Long’s own words to recommend to the chiefs that no action be taken to establish any Australian cryptanalytical organisation without first obtaining the advice and assistance of GC&CS.¹⁰ He even uses Long’s observations that there was a dearth of qualified personnel in Australia, that Australia was too far from the action in the Atlantic and Europe, and that the creation of an Australian cryptanalytical organisation risked duplication.¹¹ Australia’s CNS was arguing for Australia not to develop its own SIGINT capability.

Rear Admiral Colvin makes it abundantly clear that he was not initiating the discussion, but attempting to counter suggestions that Australia should create its own cryptanalytical agency. Colvin expressed a very Eurocentric view:

It would seem that most of the enemy messages intercepted is Atlantic traffic and the reception of such traffic in Australia is certain for only about 12 hours per diem. Only a very few messages are received each week which would not (or may not) have been intercepted by UK authorities and dealt with by the Imperial Cryptographic Organization. Thus it is improbable that there would be obtainable in Australia sufficient material upon which cryptographers could work, without merely duplicating the work being done in the UK, with much less prospect of success owing to obvious limitations.¹²

Colvin was completely wrong: simple physics meant that signals intercepted in Australia could not be intercepted in Britain or the entire western hemisphere. This is one reason Australia eventually became an important part of the worldwide Allied SIGINT system.

As time would tell, the high-power, high-frequency transmissions out of Germany and Italy could not be picked up in Britain, as the distance was too far for the ground wave, and the sky wave bounced off the ionosphere too far above Britain and Europe to be intercepted there. As for Canada and the United States, along with the entire western hemisphere, they were on the other side of the earth, and most of the signals of interest were whizzing by into the vacuum of space.

The sky waves from directional antennae in Germany and Italy directed towards the east could, however, be easily picked up at Australian sites, and in order to capture this traffic, Australian intercept sites would need to be built and operated. This single chip enabled Australia to get into the ULTRA game.

The challenge that remained, the one that Long was raising, was that Australia needed to create its own capability to decrypt intercept, so that signals of importance to the security and interests of Australia were readily available for the Australian government. It was a reasonable argument, although asking Britain to supply the necessary specialists and equipment was not.

The truth may be that Long annoyed Colvin by sending a copy of his 28 November minute directly to the CGS, Lieutenant General Ernest Ker Squires, without first letting Colvin know. This may have been a deliberate attempt by Long to get around Colvin. If it was, he was playing a dangerous game. First, Admiral Colvin was his chief, not Squires. Secondly, Colvin and Squires were British officers brought in to serve as chiefs of staff of their respective services, but this does not mean they were friends and, in fact, they had already butted heads over resources.

By December 1938, the ACNB’s operational budget had risen 16 per cent from AU£2.55 million in 1937 to £2.96 million, while the funding of new naval construction had fallen almost 80 per cent from 1936 to 1938. In the context of such severe cuts to the navy budget, Squires, as the newly appointed CGS, had put a plan to the Lyons government recommending the reorganisation of the Australian Army include a permanent force of 7500 regular soldiers and a program to re-equip and upgrade the militia.13 In response, Colvin had threatened to resign.14 So there was little love lost between Colvin and Squires or their respective services. For Long to approach Squires directly was, to put it mildly, very courageous. It worked, though, because Squires championed Long’s suggestion in his subsequent reply to Colvin:

I consider that we should at least have a nucleus organization in Australia against the contingencies of operations in and about Australia and her territories. The work is clearly of a highly skilled nature and much practice is necessary, and the sooner a commencement can be made the better.¹⁵

This is a significant turnaround in the history of Australian intelligence. Until this discussion, the great champions of intelligence, especially of SIGINT, had been the ACNB and the RAN. In 1939, the ACNB was surrendering the lead in this area to the CGS, General Squires, a British Army officer.

This did not mean that the RAN led by Colvin was not trying to build up its SIGINT capabilities, because it was. The difference lay in the scope of the SIGINT operation. The ACNB, led by Admiral Colvin, appears to have wanted to keep Australian SIGINT focused on tactical intelligence of immediate use to the military and naval forces of the Commonwealth. What Squires and Long were looking for was an Australian national SIGINT effort, something that, as we will later see, was not supported in London. Colvin’s lack of enthusiasm may have been due to the attitude of the British authorities, one that was adamantly opposed to having anyone other than GC&CS deal with diplomatic and other non-tactical SIGINT. As we will see below, the British feared that any Australian attempt to conduct SIGINT operations against Japanese diplomatic traffic would put critically important sources of SIGINT at risk in Europe.

In light of Colvin’s concerns, Squires conceded that Australia did need to approach Britain and ask advice on the matter. He made clear, however, that he was not recommending that Australia seek British permission to establish an Australian cryptanalytical organisation, but that the British authorities be asked for advice on how best to do this and on how to integrate it into the existing British system.

The other major point of difference was that Squires was focused on the Japanese and their operations in China. The Japanese had now been fighting on the Chinese mainland since July 1937, and this was producing a flood of intercept of which only a small portion was being decrypted by FECB and GC&CS. The activity of the IJA in China and the IJN in support of these operations was of great importance for Australia, and General Squires understood this. Squires did not want Australia’s new cryptanalytical organisation to target Germany or her European allies, he wanted it to target Japan:

So far as the Army is concerned the type of material mainly required for practice is that transmitted by the Japanese in the course of their operations in China. Whilst some of this may be intercepted direct (thus giving practice to signal personnel as well) a considerable quantity would have to come from the British organization in the Far East.16

Colvin, however, with the support of the CAS, Air Vice Marshal Goble, countered Squires by the simple expedient of having his fellow chiefs agree to refer the matter to the Defence Committee.¹⁷ This, as Colvin probably intended, delayed further action on Long’s suggestions for more than a year.

The question of whether Australia should establish its own version of GC&CS went before the Defence Committee on 15 February 1940. Unfortunately, by this time Lieutenant General Squires was absent, as he was dying from cancer. Major General John Northcott, the Acting CGS, represented Squires on the committee. Northcott, who did not understand intelligence, fell into line with Colvin’s recommendation that no further action be taken until advice had been obtained from Britain.¹⁸ On 11 April 1940, the prime minister, Robert Menzies, wrote to the Secretary of State for Dominion Affairs for advice.¹⁹

London only replied to Menzies’ letter six months later, on 15 October 1940, a sure sign of serious discussion around Whitehall. Eleven months had passed since Long had first raised the question of Australia creating its own cryptanalysis organisation. Lord Cranborne, the Secretary of State for Dominion Affairs advised that the British authorities felt it was inadvisable, ‘for the present’, to envisage any large-scale cryptanalytical organisation in Australia.²⁰ Cranborne did recommend, however, that Australia should contribute to the combined SIGINT effort, now centred on GC&CS, by forming a small-scale cryptanalytical organisation from the existing SIGINT organisations operating in Australia.²¹ Cranborne advised that this new organisation should continue to work on IJN codes in cooperation with FECB in Singapore and GC&CS in Britain and, importantly, expand its operations to cover fixed commercial wireless station traffic throughout the world.22 These were fair and reasonable suggestions.

Despite the clear recommendation from Cranborne, the Defence Committee continued to prevaricate. At its normal meeting on 5 December 1940, the committee deferred consideration of the matter until it had been discussed with the Admiralty’s Chief of the Naval Intelligence Staff, Captain F.J. Wylie, RN, who was due to visit Australia later that month. In the interim, the CNS would discuss the matter with the CGS and CAS, and raise the question of training army and RAAF personnel in this type of work. Effectively, Australia’s own military advisors had postponed taking action on any aspect of SIGINT for a full fifteen months since the country had entered the war in Europe.²³

While all this was happening, the army had taken the initiative and established a small cryptographic group in Sydney. This group comprised academics from the University of Sydney, including Professor A.P. Treweek (also a major in the Sydney University Regiment), professors A.D. Trendall and T.G. Room, R.J. Lyons and Lieutenant I.H. Longfield Lloyd, Australian Military Forces (AMF) (the son of Colonel E.E. Longfield Lloyd of the Security Service), who worked part-time on Japan’s codes. Administrative responsibility was vested in the GOC, Eastern Command, and Major Reginald Powell of the Military Intelligence Branch oversaw its operations.

The Sydney cryptographic group was not particularly effective, and the claims by Major Powell that they were making considerable progress in breaking Japan’s diplomatic codes can be discounted because the resources being dedicated to the problem were minuscule.²⁴ None of the academics involved, with the exception of Treweek who was a militia officer, was being paid, and they appear to have been working on Japanese codes in their spare time on an honorary basis. There is no documentation for the payment of salaries to them, and the Senate of the University of Sydney had not released them from their responsibilities. The idea that six or seven academics could make advances that large and well-resourced US and British agencies could not is hard to believe.

Through no fault of their own, the group lacked the necessary equipment and supporting services, and the amount of raw intercept they could obtain was small and partial. The number of messages crossing their desks could not have provided the statistical data necessary for decryption. Likewise, they had no secure contact with the intercept operators, and no ability to apply TA to cribs and those cribs to their cryptanalysis. In short, they were operating without any of the infrastructure necessary for making headway against Japan’s codes.

Finally, as far as this analysis is concerned, there is no evidence of any such advances in the archives. There are no files of broken-out messages or reports derived from them. There are no files showing exchanges of intelligence between the army and any other body. All we have is Major Powell’s unsubstantiated claims and a single decryption he forwarded in his report. Australia was now paying the price of its own failure to adequately fund its own defence between 1918 and 1939 and for the destruction of Edmund Piesse’s Pacific Branch.

The one decrypt Powell did forward is telling. It was not a decrypt of any Japanese coded message, but of a love letter between a named British official in Hong Kong and a married Australian woman, who is also named in the files. These two people appear to have shared a shipboard romance en route to their destinations in Asia. Even today, this letter is disconcerting, as the woman was the wife of a senior Australian officer and no attempt has been made to disguise either identity.25

Over at Navy Office, things had changed. The initiative on SIGINT had been lost in the deliberations of the Defence Committee, and the army had its own program. In an effort to meet the needs of the RAN, in September 1940 Navy Office established the Special Intelligence Bureau, a small cryptanalytical section within the NID. But they could not find suitable people any more than the army could.²⁶

The biggest skill gap was the lack of Japanese linguists. The old language program had qualified a few officers, but most of those had been lost to the services and only a few, like Nave, remained available. The Australian authorities also made things even harder for themselves by refusing to consider recruiting Japanese Australians for the work. Harry Freame was the exception, and he was only involved because he was a highly decorated and well known-returned serviceman and had served with E.E. Longfield Lloyd, who was now the Director of the Commonwealth Security Service.

In all of this gloom, a little serendipity now intervened and Eric Nave was back in Australia convalescing from tropical sprue, a serious digestive condition that afflicted many serving in tropical climes. Nave was now a commander in the RN and would normally have been repatriated to Britain, but FECB did not want to risk losing him to GC&CS if he returned to Britain. It was arranged for him to go to Melbourne before the Admiralty could stop it.27 This turned out to be a mistake.

The truth appears to have been that the newly married Nave had had enough of Singapore and did not want to return. As he was an RN officer, the ACNB had no authority to allow him to stay in Australia, but Nave had a friend in Australia, Admiral Colvin, who had been the British Naval Attaché in Tokyo when Nave undertook language training there. Added to this, Colvin was an old China hand who knew of Nave’s good work and who understood the plight of someone suffering from tropical sprue. It is also likely that Colvin was influenced by his DNI, Commander Long, and his Communications Staff Officer, Commander Jack Newman, who were both keen to keep Nave for cryptanalytical and SIGINT work in Australia. Whatever reservations Colvin had about Australian plans for a national SIGINT effort, he was supportive of keeping an experienced Japanese linguist and cryptologist like Nave in Australia. The evidence for this is that in the tussle that now developed between the Admiralty and the ACNB for Nave’s services, the ACNB was able to hold on to the RN’s man.²⁸ Only Colvin, with the help of the RAN’s doctors, could have accomplished that.

As the Admiralty wanted Nave back in Singapore, the ACNB used its medical board to argue that he was not fit enough to be posted back there. The Admiralty then insisted that Nave return to the United Kingdom. The Australian riposte was that Nave was so ill he could not even travel through the tropics to get back to Britain.²⁹ This gambit worked, and the Red Book recorded that in July 1940 Nave was posted to the Melbourne Communications Intelligence Unit (CIU; Special Intelligence Bureau).³⁰

At the Melbourne CIU, Nave’s job was to lead the establishment of the Special Intelligence Bureau. Records suggest that he was the sole commissioned officer until November 1940, when Paymaster Lieutenant Keith Stafford Miller was posted in to join him.³¹ Miller had resigned his commission on 5 November 1927, but was brought back onto the RAN Emergency List on 21 October 1940, most likely at Nave’s insistence.

There is no documentary proof that Miller was a Japanese linguist, but he remained in the Melbourne CIU until September 1945.³² He then stayed with the RAN until January 1949, when he was appointed to the Commonwealth Public Service as an officer of the Joint Intelligence Organisation.33 This shows that by the end of the war Miller was a hardened intelligence professional.

In January 1941, Captain Wylie, RN, the head of FECB in Singapore, visited Australia to provide the RAN with advice on the conduct and management of its SIGINT effort, and to discuss proposals for the division of labour on SIGINT operations against the IJN. The meeting in the Conference Room at Navy Office consisted of Wylie; the Second Naval Member, Commodore J.W. Durnford; Captain J. Burnett; and commanders Long, Newman and R.F. Nichols. The discussion ranged over the RAN’s plans and the potential difficulties they foresaw with the other services, especially with the army’s ‘horror’ of any form of joint operation, and its inability to differentiate between operational and strategic intelligence activity.³⁴ The other matter discussed was the liaison arrangements between the C-in-C, Singapore and the C-in-C of the Dutch forces in the NEI.

Also discussed was the Admiralty’s request that the ACNB provide three additional HFDF stations collocated alongside existing stations. The Admiralty wanted the ACNB to contribute to the DF network in Asia, even though the Australian government had wriggled out of this commitment in the mid-1930s. Raising it now, Wylie admitted, was ‘rather tall’, but it was essential if the British and Allied forces in Asia were to deal with a Japanese attack. Duplication of the HFDF stations had arisen from the experience in the Atlantic, where having just one DF station at a site caused too many errors in fixes and prevented multiple simultaneous fixes being made.³⁵

Captain Burnett, RAN, challenged the effectiveness of this more complex and expensive system, and put the view that the operations staff did not see much value in the intelligence being provided. Burnett was making the usual complaint of all operations officers—that they had no use for messages saying there was no change in the disposition of target vessels.³⁶ What Burnett did not appreciate was that a ‘no change’ message had value, in that it confirmed that the target vessels had remained in position. This intelligence was not a guess, but the product of a lot of work.³⁷ Each ‘no change’ message required 300–400 bearings a day and around 2000 bearings a week, all of which required coordination and plotting in order for the SIGINT system to categorically report that no IJN vessels had moved.³⁸ Wylie explained this to the meeting, for the benefit of Burnett.

Wylie also emphasised the importance of combined intelligence operations, including analysis and assessment processes. While each service element collected, collated and distributed its own intelligence product to its respective service, Wylie said they should also distribute it into a central pool from which the other services and governmental agencies could self-select intelligence product. This important point was not lost on the RAN officers present, as someone noted ‘Very Good. We should remember this’ in the margin of the minutes of the meeting.39

The discussions between Wylie and the RAN also covered the suggested Australian cryptanalytical organisation. It is highly likely that this subject was raised with Wylie by Long and, later, by Nave. Wylie followed the official line already outlined by Lord Cranborne: that Australia should not contemplate a large organisation but look to expanding the existing organisation now being led by Nave.

As for the RAAF, it had no operational SIGINT capability and the RAN had fewer than twenty naval ratings qualified to undertake SIGINT intercept operations. The other factor was the lack of intercept sites. The RAN was only now preparing HMAS Coonawarra at Darwin and HMAS Harman at Canberra to provide cover of the IJN nets, and the army’s intercept site at Park Orchards, in the outer suburbs of eastern Melbourne, was being used to intercept Japanese diplomatic and international commercial traffic. A fourth intercept site, Jandakot in Western Australia, was being commissioned with one receiver to listen for German surface raiders. The Jandakot site would shortly be supplied with a second receiver so that the intercept operators there could work Japanese Morse.⁴⁰ The advice to keep any new organisation small was sensible, and the obvious way forward for Australia was to merge the army’s cryptographic group in Sydney with Nave’s Special Intelligence Bureau in Melbourne.

On 2 May 1941, a conference was held in Melbourne to settle the issue of Australia’s SIGINT effort. This conference was important because it involved everyone: Long from the RAN; Lieutenant Colonel Edwards, Major J.C.W O’Connor and Captain E.H. Fleiter from the army; and Professor Room from the cryptographic group in Sydney.⁴¹ At this meeting, Long ‘intimated’ that the ACNB had established a SIGINT section, before he and Colonel K.A. McKenzie left the meeting to allow the SIGINT specialists to discuss the technical details of the work being done by the cryptographic group in Sydney.⁴²

This conference was also significant because it was not about service SIGINT at all—the focus was on breaking Japan’s diplomatic codes. This was about an Australia strategic SIGINT effort directed towards meeting Australian self-interest. The first order of business was to see if they all agreed that breaking the Japanese diplomatic codes was a feasible objective for the new organisation. They agreed it was. The second order of business was whether Australia should do this, and they argued that as the FECB ‘may not always be available’, it was essential for Australia to have its own independent capability.43 The new organisation’s manpower was to be four officers and three clerks, on top of the existing naval organisation in Melbourne, and it would be a joint organisation along similar lines to FECB.⁴⁴

All of this was passed to FECB in Singapore, and the army agreed to ensure that Eastern Command would continue to provide the unofficial Sydney group with support and that all of their activities would still be covered by the strictest security measures as detailed by the RAN.⁴⁵

On 15 May 1941, Australia’s naval and military authorities tasked the Sydney group with conducting a cryptanalytic attack on Japanese consular and diplomatic messages. The minute from the CNS to the CGS made it clear that the intercepted traffic was to be attacked in Australia by the cryptographic group and that Commander Nave now believed some of the members of the Sydney group would be ‘most useful’ in dealing with Japanese diplomatic traffic.⁴⁶ This is the first time Australian chiefs of staff clearly state a position of Australian independence in SIGINT matters.

Professor Arthur Sadler, who had worked as James Murdoch’s assistant and whom Nave had identified as the most competent Japanese linguist, now fell under the malign eye of the Military Police Intelligence Bureau operating in New South Wales. The problem was that Sadler, a Japanese linguist and expert, had a Japanese wife. This made him unacceptable to the dysfunctional security apparatus and, acting on their advice, the GOC Eastern Command would not recommend Sadler because he was ‘inclined to be indiscreet’.⁴⁷

From May 1941, things began to speed up as the ACNB requested the CGS to consider posting one of the Sydney academics and a clerk to Melbourne as soon as possible, to be followed by the remaining members, plus three clerical staff, immediately afterwards.⁴⁸ The army actioned this request on 27 May.⁴⁹

All of this may appear to a modern reader to be simple, but, as anyone who has served in the military knows, even in wartime the military is a vast and slow-moving bureaucracy. And there are other slow-moving bureaucracies too.

In June 1941, the CGS, General Vernon Sturdee, wrote to the Minister for the Army, Percy Spender, requesting permission to call up the professors in Sydney for military service in the rank of major.50 Sturdee emphasised that the work of these men would be ‘of the greatest importance’ in both collecting intelligence and training military personnel to work on Japanese signals and codes. The best the army could do was to make all four academics temporary majors on special duty, with Treweek, Lyons and Room going to Melbourne and Professor Trendall staying in Sydney.⁵¹

The Senate of the University of Sydney released Room, Lyons and Treweek for service on 4 June, and while the Senate was happy with Treweek and Lyons being appointed majors, it wanted Professor Room to be made a full colonel, to keep him on a pay level equivalent to his university salary.⁵² The university also wanted the army to pay for all the costs of moving the men to Melbourne.

Now the two bureaucracies engaged in mortal combat. Professor Sir Robert Wallace, the Vice-Chancellor of the university, took the matter to Percy Spender, who appears to have been bemused by the fuss. Spender wanted to know from Sturdee what he should tell Professor Wallace, given there was little likelihood of the military establishment agreeing to make Professor Room a full colonel and pay £5000 a year for his services.⁵³ The army suggested employing Professor Room as a civilian under the ACNB, which, luckily, agreed.⁵⁴ Finally, on 25 July, Professor Wallace was informed that Room and Lyons could be employed to undertake the special work in a civilian capacity and at salaries equivalent to those they had been paid at the university.

With the administrative nightmare of employing the four academics overcome, it was the turn of Frank Sinclair, the Secretary of the Department of the Army, to interfere. This took the form of an objection to the interception and reading of Japanese diplomatic traffic because it could contravene international agreements giving ‘diplomatic messages immunity from interference’.⁵⁵ Sinclair was Australia’s version of Henry Stimson, the US Secretary of State who withdrew funding from the first US SIGINT operation led by Herbert Yardley because, as he later wrote in his memoirs, ‘Gentlemen do not read each other’s mail.’⁵⁶

It was left to the Deputy CGS, Colonel John Chapman, to write to Sinclair telling him that the diplomatic telegrams involved were sighted in the cable companies’ offices and copies made, and that everyone was doing it. Chapman also noted that diplomatic telegrams and radio messages were not subject to privilege and thus fair game.57 It was a big boys’ game played with big boys’ rules.

On 14 November 1941, the Defence Committee agenda included an item on Australia’s small special intelligence organisation and, on 28 November, the three chiefs of staff, sitting as the Defence Committee, formally approved the organisation, its establishment and its mission.⁵⁸ They even recommended that the chiefs of staff be given authority to increase the size of the organisation, as and when it became necessary.⁵⁹

By December 1941, Australia’s national SIGINT organisation, the Special Intelligence Bureau, consisted of Professor Room; Eric Nave, RN; Paymaster Lieutenant Commander A.E.N. Merry, RN; a Japanese linguist from FECB and paid for by the Admiralty; Paymaster Lieutenant Commander W.E. McLaughlin, RAN; Paymaster Lieutenant Miller; A.B. Jamison; and three secretaries, Misses Robertson, Eldridge and Shearer. The ACNB was paying for all of these, with the exception of Merry. The army paid for Major Treweek, A.A. Mason, R.J. Lyons and Lieutenant Longfield Lloyd.⁶⁰

The whole process of creating the Special Intelligence Bureau had taken two years, from 28 November 1939 to 28 November 1941, and its creation only preceded General Yamashita’s landing in Malaya by two weeks.

During February and March 1942, reinforcements for the bureau were hastily sent out from the Admiralty and Foreign Office in London. They included the Japanese linguists Lieutenant Commander E. Colegrave, RN, and Hubert Graves, Henry Archer and A.R.V. Cooper of the Consular Service.⁶¹ Archer had served as the British Consul-General in Harbin, China, and was a competent Japanese linguist. Hubert Graves had served as Consul-General at Kobe in Japan and was a very competent linguist.⁶² The posting of these British personnel clearly shows that despite Australia having initiated the development of the Special Intelligence Bureau, GC&CS was heavily involved every step of the way.