The main British stations for intercepting Japanese military codes and ciphers were in India. There had been an intercept capability in India since the First World War, when the Indian Army set one up at its headquarters in Simla. In the immediate wake of the First World War, with Russia the main threat to British interests in India, an intercept site was constructed at Abbottabad on the North-West Frontier. Throughout the 1920s and 1930s the Indian codebreakers concentrated on Soviet codes and ciphers in an extension of the Great Game, the intelligence battle between the Russians and the British that had raged across the region from the late 1900s. Their work was given an early start by Tiltman, who recalled spending most of the 1920s in India.
I was a member of a section of the General Staff at Army Headquarters in Simla, India, consisting of never more than five persons. We were employed almost entirely on one task, to read as currently as possible the Russian diplomatic cipher traffic between Moscow, Kabul in Afghanistan and Tashkent in Turkestan. From about 1925 onwards I found myself very frequently involved in all aspects of the work – directing the interception and encouraging the operators at our intercept stations on the North-West Frontier of India, doing all the rudimentary traffic analysis that was necessary, diagnosing the cipher systems when the frequent changes occurred, stripping the long additive keys, recovering the codebooks, translating the messages and arguing their significance with the Intelligence Branch of the General Staff. A cryptanalyst should take every opportunity of acquiring a general working knowledge of branches of Signal Intelligence other than his own. I realize that I was exceptionally lucky to have this opportunity and that very few others have had the chance of acquiring this kind of general working experience.
The experiment proved a success. The Wireless Experimental Depot moved briefly to Cherat but had returned to Abbottabad by the time the war with Germany broke out. As war with Japan loomed in 1940 and 1941, two more British Army intercept units were set up in Burma, at Rangoon and Maymyo, near Mandalay. Meanwhile, the main Indian Army station, now known as the Wireless Experimental Centre, had moved to a site at Anand Parbat, just outside Delhi, explained W. C. Smith, an Intelligence Corps sergeant who worked there.
The centre occupied the site of Ramjas College, an Arya Samaj, or Protestant Hindu foundation evacuated elsewhere, on the outskirts of old and new Delhi. Its hilly, isolated and self-contained position, accessible to the capital, made it very suitable for a wireless station. We were on the edge of the Thari or Rajputana desert, where the clear air perhaps aided reception, but where we lived with the continual terror of the anhi – the dust storm.
This was a nerve centre of intelligence from military wireless sources against Japan. Well over a thousand are believed to have inhabited it, at its height, made up of Intelligence Corps (the real Wireless Experimental Centre, India Command); RAF, from 166 Signals Wing; Women's Auxiliary Corps (India); Indian airmen, many from the far south of India, perhaps with Malayalam or Tamil mother tongue; West African Signals, from what are now Ghana and Nigeria; occasional British ATS, and more than one WAAF officer; and, of course, a host of Indian Army NCOs and men. There were many civilians too, not least the Sikh postmaster, and Habibullah, the Pathan fruitseller, and last but not least the cycle shop at the gate, our lifeline off duty at a hire of one rupee eight annas a day.
The WEC was subordinate to the British authorities in India and was in theory independent of their counterparts in the UK. It was controlled by the Inter-Service Wireless Intelligence Staff which was in charge of signals intelligence in India. But effectively the WEC was an outpost of Bletchley Park. It was a joint RAF/army unit, ostensibly commanded by an RAF group captain but split into five sections, most of which came under the direct control of Lieutenant-Colonel Peter Marr-Johnson, who had flown to India when the FECB was evacuated from Singapore. He was in charge of A Section, administration; B Section, the intelligence collation and reporting section; C Section, which was the ‘Special Intelligence’ or code-breaking department; and D Section, which co-ordinated traffic analysis and DF results. E Section, interception and communications, was controlled by an RAF wing commander, Smith said.
My own B Section duties were partly concerned with ordering maps to cover the whole of the war in Asia, from Burma to the Pacific, for the adjoining Map Room in which were kept detailed maps of the situation in Burma and the parts of India subject to invasion. The real star of the Map Room was the huge map of the estimated Japanese order of battle. My main task was to make a card index of every place east of Burma under Japanese occupation or in which there were Japanese forces, to help in locating geographical places, and to enter the appropriate units. When I started there was nothing to go on, so I picked up the Statesman (Calcutta and Delhi) and looked at the war news in the East. Even in March 1943 they devoted far more attention to war in Europe than to Asia, except perhaps in the Arakan. However, that day our planes bombed Kavieng, New Ireland. So it was found, and the first card was made with the latitude and longitude and brackets (KABIEN) for the Japanese equivalent. By the end of the war there were many cards and WEWAK, HOLLANDEA, RABAUL (RABAURU) had many entries. Many were the puzzles in identifying BANPON, KANBURI and so forth, and from that began the tracking of the course of the Burma–Siam Railway. Hardest of all to identify were places in China with many variations in spellings – according to various readings of Chinese characters and translations – and the Japanese renamed many places. It was thrilling gradually to build up a picture of the whole area, although individually one knew that was a very small part.
Although communication with other signals intelligence centres was difficult because of distance, the WEC was at the forefront of breaking Japanese army codes. The first success was technically shared with Kilindini. Major John Figgess, one of the centre's code-breakers, was sent as a liaison officer in November 1942 to the Royal Navy unit, where he spent much of his time producing military intelligence from a simple substitution cipher, the WE code, which the Japanese Navy used when sending military traffic.
The system, introduced because the existing naval codes had no facilities to disguise the identities of recipients and senders of army messages, was known as WE because the transposed kana syllables it used were always bracketed by the syllable WE. Despite its lack of security, the WE code was mainly used for preambles of important messages between Tokyo and the main Japanese bases of Palau, Rabaul and Truk; it became a valuable source of information on the Japanese order of battle and provided a number of cribs for more secure code and cipher systems.
The WEC had two main outposts of its own: the Western Wireless Signal Centre (WWSC) at Bangalore in south-western India and the Eastern Wireless Signal Centre (EWSC) at Barrackpore, near Calcutta, in north-east India. The western centre was purely a fixed intercept site covering military operations in Malaya and further afield, but the eastern centre was much more substantial. It had its own intelligence operation, known as Intelligence School ‘C’, which reported to 14th Army at Calcutta. It also provided operational air intelligence to aid the city's air defences, and three mobile special wireless sections, which in April 1943 were deployed close to the front line at Imphal, Chittagong and Maungdaw.
Harry Beckhough, the then officer commanding Intelligence School ‘C’, said operations were hampered early on by inexperience.
Lack of sufficient knowledge of the Japanese language and radio dispatch system meant first tracking Japanese units by strength of radio traffic in different areas. Monitoring units were set up in the jungle and manned by Royal Signals and RAF radio operators, furnishing me with all possible intercepts of Japanese signals for decryption. These were courageous groups, glued to their radios, picking up enemy messages in difficult areas, often cut off from communications except ours.
Abbottabad was independent of the WEC, coming under the direct control of the Inter-Service Wireless Intelligence Staff. It was mainly concerned with diplomatic traffic, having taken over that responsibility from the Royal Navy code
breakers after the evacuation of Singapore. There was a large number of Indian civilian and service wireless operators at Abbottabad, but there were also Royal Signals and RAF intercept specialists, some of them training the Indian staff. Dennis Underwood volunteered to join the Royal Signals at eighteen and was sent to Douglas on the Isle of Man for his special operator training. With manpower desperately needed for the frontline regiments, the army was increasingly turning to members of its female equivalent, the Auxiliary Territorial Service or ATS, as a source for intercept operators. ‘By the time I finished there was only our male squad left,’ said Underwood. ‘All the others were ATS. That might sound like a possible heaven but in fact it was just the opposite. We men were on one duty or another almost every night, guard duty, fire piquet, kitchen fatigues, etc., and when the occasional night off came we were too exhausted to take advantage of the opportunities!’
At the end of his training, he was sent to Abbottabad.
I recall it as being a fairly small mountain town. At that height the Japanese signals came in loud and clear. We had a bearer to whom we each paid one rupee a week. For this he kept the barrack room clean, the fires burning, our equipment clean, woke us in the morning with a mug of cha and then shaved us as we lay in bed. I have no recollection of the food but can remember a canteen which served cha in glasses which were very difficult to carry because of the heat. In addition, a traditional cha-walla toured the camp charging one anna a glass. There were then sixteen annas to the rupee and twelve rupees to the pound.
There was a service club in Abbottabad run by, I think, the ladies of WASBI (Women's Services British India), where we gathered some evenings when I remember eating toast and drinking yet more cha. One lasting memory is of listening to Dvorak's Symphony from the New World via the BBC World Service. We also visited the local cinema on occasions. I can recall sitting on benches reminiscent of church pews but have no memory of any of the films we saw except that sometimes they were Indian. We played football in our spare time, although at that height it was rather exhausting. On one occasion a group of us travelled down to Rawalpindi to watch a match between a team of professional players and a services team. I can't remember who won, but I do recall the hair-raising journey there and back along a road with precipitous drops and no barriers.
Dorothy Ratcliffe worked at the Abbottabad Wireless Experimental Depot, sending the reports back to Bletchley Park via a Type X machine cipher link.
My husband was in the Indian Army and he was sent off to Iraq in 1941, leaving me behind in Secunderabad. A friend who was married to a Gurkha officer had said: ‘If ever you're abandoned come and visit me in Abbottabad.’ So I went to see her and while there stayed in the Abbott Hotel, which was a collection of chalets scattered over a fairly wide area. I was going over to the dining room one day when a man I didn't know stopped me and said: ‘Can you type?’ Well, I'd done a course of typing and book-keeping after I left school but I'd never actually had to work. He said, ‘Why don't you come along to the mess tonight, have a drink and meet Colonel Kenneth Tippett,’ who was the head of the Wireless Experimental Depot. Colonel Tippett asked me a few questions to make sure I was suitable – they didn't want anyone who drank, who might talk, you see – and then told me I should start the next day at a salary of 100 rupees a month. This was only about £7 but I thought it was marvellous. What did I know, I'd never had to work before in my life.
So I was picked up in a car the next day, along with four others, by one of the men who worked there and he took us a couple of miles outside Abbottabad to the Wireless Depot where I was given a typewriter and told to start typing these intelligence reports. After a while I was put on ciphers, sending the messages to Bletchley. First of all it had to be done by hand but then we got one Type X cipher machine and I was put on my own in a little side room where I didn't see anyone. There was a large notice on the door saying, ‘Keep Out, This Means You’, and there were all these young men coming up to the door to ask me for a date who read the notice and went away again. It was such a different life there. We were not normal people. We abandoned wives had a whale of a time. There were always chaps ready to take you out dining and dancing. But after a couple of years my husband was posted back from Iraq and I had to hand my notice in and join him in Lahore.
Central Bureau, the Australian-based military signals intelligence operation, had moved with MacArthur's headquarters to Brisbane. Its director was Brigadier-General Spencer B. Akin, the former military head of the Signal Intelligence Service, the US Army code-breakers, who was now MacArthur's Chief Signal Officer. The day-to-day running of the bureau was initially in the hands of an assistant director, Colonel Joe R. Sherr, the former commander of Station 6, the US military sigint unit at Fort McKinley in the Philippines, and an executive officer, later also assistant director, Lieutenant-Colonel A. W. ‘Mic’ Sandford, a member of the Australian Imperial Force (AIF), the Australian Army, and a veteran of mobile tactical sigint units in the Western Desert.
There were three other key senior officers within the headquarters operation: Major Abe Sinkov, the US Army codebreaker who led the first US delegation to Bletchley Park, was head of 837th Detachment, the US codebreaking component; Wing Commander Roy Booth, the former commander of the RAAF's Darwin intercept operation, was in charge of the air force contingent, eventually becoming a third assistant director; and Norman Webb, now a major, was the most senior British representative.
The bureau was organized in a remarkably similar fashion to Bletchley Park. Its headquarters was at 21 Henry Street, a two-storey mansion close to the Eagle Farm racecourse at Ascot Park. But this soon overflowed and a series of huts was erected on the racecourse itself. Surrounded by wire fences and patrolled by Australian military, it was to become home for the Central Bureau code-breakers, linguists and intelligence analysts working on the Japanese messages. As the bureau grew in size, the racecourse became covered in more and more huts, numbered much like those at Bletchley, each containing a particular section. Central Bureau was divided into four main sections: Traffic Analysis, employing six officers and sixty ORs under Captain Stan Clark, one of the Australian Army officers who had seen service in the Middle East; High-Grade Codebreaking under Major Sinkov, employing thirteen officers and sixty ORs (including a tabulating machine section), of which roughly a third were Australians; Air–Ground Codebreaking under Nave, with the naval air side led by a US Army major, the army air by an Australian major and a meteorological section under Professor Room; and Collateral Intelligence, employing four officers and six ORs, shared between the Australian Army and the RAAF. This last section was a purely Australian operation. It checked the transcripts alongside sighting reports and captured documents, provided intelligence reports to the Australian commands, built up a picture of the order of battle, and liaised with Bletchley Park and the WEC in Delhi. The US Army codebreakers at Central Bureau had no responsibility for circulation of the intelligence they gathered. Their transcripts were ‘edited’ for intelligence value and then passed direct to Akin and General Charles Willoughby, MacArthur's chief intelligence officer, to be disseminated to US customers.
Joe Richard, then a sergeant, was one of the first US Army code-breakers to arrive in Brisbane.
At Central Bureau, during 1942, the Americans contributed very little. They were attacking low-level Japanese Army messages that were sent in three-figure groups. These proved both hard to intercept and mostly unreadable (being mostly one-time pad). The Australian Army [AIF], led by experienced men who had returned from the Middle East, was doing the interception, starting the traffic analysis, and carrying out the field processing of the army air-to-ground messages in three-figure code and the naval air-to-ground messages in 3- kana and 4- kana codes. The Royal Australian Air Force was at Townsville intercepting the Japanese naval air-to-ground 3- kana and 4- kana codes and sending them to Central Bureau. Due to the Europe-first policy, the US did not send many intercept army personnel to Australia until mid-1943. A handfu
l evacuated from the Philippines by air worked with the RAAF at Townsville, teaching them to pick up the air-to-ground frequency. The training of Australian personnel in breaking and reading the Japanese Army and Navy air–ground messages was done by Eric Nave.
14
CENTRAL BUREAU'S BIG BREAK
Despite Fabian's efforts to have Eric Nave sent ‘home’ to Britain as part of the terms of the Anglo-US Agreement, the Australian code-breaker had been transferred instead to Central Bureau. Fabian was unapologetic over his refusal to share FRUMEL's signals intelligence with the Royal Navy codebreakers at Kilindini and the US and Australian Army codebreakers at Brisbane. He later recalled:
Security was a paramount concern for me. I was relieved when Commander Nave, an Australian cryptanalyst, left FRUMEL for Central Bureau. He left because I reprimanded him for his lack of security. I also had to get my admiral to remind MacArthur about the need for security. As the commander of the South-west Pacific theatre, MacArthur informed Admiral Leary, who was set up at Brisbane, that he wanted information produced by FRUMEL. Admiral Leary told me that we had to give MacArthur information but asked my suggestion on the best way to supply such material.
I felt that certain restrictions were necessary to ensure security. Admiral Leary issued the following requirements: (1) Fabian or one of his unit's representatives will report to MacArthur's headquarters each day at 1400 hours. The FRUMEL representative will never be kept waiting in MacArthur's outer office. (2) No-one will be authorized to make copies of any material provided by FRUMEL. (3) During the briefing of FRUMEL material, only MacArthur and his chief of staff, General Sutherland, will be present. Everyone else, including General Willoughby, will be excluded from these briefings.
The Emperor's Codes Page 20