3

Code Club

In January 1940, Professor Thomas Room, head of the mathematics department at the University of Sydney, received an invitation to meet confidentially with the army’s chief of staff at the nearby Victoria Barracks. When he arrived for the meeting, the chief, Lieutenant General Ernest Ker Squires, made a surprising and curious suggestion. Squires asked the professor if he would like to start up a code-breaking club with some of his fellow academics at the university. Squires added that it would be better to meet at the barracks than on campus. It would draw less attention, and if they made much progress the army could offer certain resources.

Squires knew nothing about code-breaking himself, and so could offer little detail about what Room should do next, other than to impress upon him that the army wanted someone local to learn code-breaking. Although Room was a mathematician, cryptanalysis — the discipline of breaking codes and ciphers — was hardly his speciality. His thing was geometry.

Thomas Gerald Room was a friendly but somewhat shy man who always spoke in a formal, correct manner, even in casual settings. He had arrived in Australia from Cambridge, England, five years earlier to take up a professorship at the University of Sydney’s mathematics department, which at the time comprised five staff, including two temporary assistant professors.1

A devout Christian, he joined the Student Christian Movement, and through that campus club met Jessica Bannerman, an Australian arts graduate. They married in 1937 and bought a house in St Ives on the city’s north shore. In 1938, he published a book, The Geometry of Determinantal Loci, a magnum opus that had taken him several years to write.

Professor Room pondered who among his colleagues he should approach to start a weekend code-breaking club. After some consideration, he invited only one other faculty member — his close friend Dicky Lyons. Since moving to Sydney, he and Richard Lyons had formed a friendship around two common interests: mathematics and sailing. They spent their summer weekends together yachting on Sydney harbour.2 It was a somewhat uneven friendship because Lyons, a less accomplished scholar, was in awe of Room’s mathematical talent and achievements.

The two academics went to Victoria Barracks, where they were allocated a private place, and started teaching themselves cryptanalysis. The university library had a copy of The Black Chamber, their most likely starting point.

The Black Chamber was written by Herbert Yardley, a code-breaker for the United States navy until 1929. Yardley’s book described in detail how his United States’ code-breaking unit — dubbed ‘The Black Chamber’ — decrypted Japanese diplomatic messages during the Washington Peace Conference of 1921 and by doing so learned Japan’s negotiating strategy at the conference. The result was that the final deal restricted Japanese naval expansion to the greatest degree possible, tethering them to a ratio of 5:5:3 against the navies of the United States and Britain. Yardley’s book provided numerous insights into the workings of codes and ciphers, as well as its role in modern-day intelligence.3

In 1929, the United States’ president, Herbert Hoover, appointed Henry L. Stimson as secretary of war. When shown the contents of decrypted diplomatic messages of other nations, Stimson was appalled. To his sensibilities, the work of Yardley’s Black Chamber was underhanded, secretive, and immoral, and not the sort of thing that civilised nations did. Stimson ordered that the Black Chamber be closed down immediately because, in his words, ‘Gentlemen do not read each others’ mail.’

Turfed out of his job, Yardley got his revenge with the publication of The Black Chamber, blowing the lid on American cryptanalysis activities against Japan in the process, and providing the ideal introductory textbook to the art of code-breaking for two mathematics professors in Sydney.

After learning the basics, the two professors were ready to practise and hone their newfound skills. The army’s chief of staff was happy to help, and they were provided with sample material to get them started. At first they were given exercises with simple, low-level codes in Japanese. Once they had sharpened their teeth on these easy problems, the army gave them some messages in the so-called LA diplomatic code.

The professors were not told where these practice messages came from, but the LA code messages weren’t just toy puzzles. They were real Japanese consular messages that had been intercepted by an army station at Park Orchards near Melbourne.

The Australian army had recently made the decision to try to build local expertise in making codes (known as cryptography) and in breaking codes (cryptanalysis). They first tried to do it in-house, establishing a small team of army personnel, but the progress of the unit was slow and disappointing. Their contacts in Singapore at the Far East Combined Bureau suggested that they follow the model of the GCCS in Britain and to try to recruit some ‘professor types’.

Room and Lyons couldn’t speak Japanese, but, since they were dealing with coded Japanese messages, they needed to learn. They took courses in Japanese from Margaret Lake, a lecturer at the teachers’ college on the university campus. Lake was an excellent teacher, and they made rapid progress.4

After their initial forays into code-breaking, and with the encouragement of the chief of staff, they sought to expand their circle, and approached the university’s two classics professors, Trendall and Treweek, to join their secret club.

Athanasius Treweek had been a member of the Sydney University Regiment for years, and was a captain in the field artillery. Treweek had long believed that war between Japan and Australia was inevitable, and had been learning Japanese since 1937 in preparation for it.

Arthur Dale Trendall was a New Zealander who had been in Australia for two years. Fluent in Greek and Italian, he was an expert in fourth-century Greek vases, and had a growing international reputation as a classical scholar.

Trendall was renowned for his uncanny intuitive judgement when inspecting a new classical item. Richard Johnson, in one of several obituaries after Trendall’s death in 1995, said of him,

Trendall had a phenomenal visual memory; he could look at a vase he had never seen before and place it to an obscure painter or school which he had last seen years previously, on the basis of the stylistic similarities. When one considers that about 20,000 vases of the period are known, and he knew every one, the magnitude of the memory may be appreciated. Nor, of course, was it only memory; he had an eye for style and for details of execution which enabled him to produce his great classificatory collections: Paestan, Campanian, Apulian, Lucanian, Sicilian.5

This ability to see patterns and connections in ancient vases seemed to translate rather well to the art of code-breaking. While the mathematicians worked on theoretical solutions, Trendall described this ability years later to the historian Des Ball, telling him, ‘You get a feeling for it. Your eye lights up on something, and … bang.’6

In fact, Trendall considered the intuitive approach to code-breaking to be far superior to the mathematical, analytical approach — in part because real messages were incomplete, or contained errors made either by the sender or the interceptor who transcribed it. These messy, error-ridden messages were called garbles in the cryptanalytical trade. Garbles were a diabolical problem for the mathematicians, but Trendall could quickly see beyond the chaos of the garble to the ordered pattern of the message below.7

They met at the barracks on weekends, where the army provided them with a steady stream of cryptanalytical puzzles, all of which were real, intercepted Japanese messages. Quite quickly, they broke the LA code, a Japanese consular code. Trendall boasted that breaking the LA code was ‘child’s play’.

The chief of staff came to them with a different kind of code-breaking task in September. He explained to them that the army censor had come across suspicious correspondence, and wanted them to investigate. A letter from a British officer stationed in China to the wife of a senior air force officer in Melbourne contained curious patterns of dots on a newspaper clipping that came with the letter. The officer and his wife had recently been living in China themselves, having returned to Australia in July, so it was possible that one or both of them had made foreign connections before returning home. And, adding to the suspicion, the letter had been sent first to another woman in Sydney, who had then forwarded it to the woman in Melbourne. The army was worried that the dots were a code being used to engage in espionage. The chief of staff handed them the correspondence in question and left them to it.

The professors studied the letters and the clipping. There were indeed patterns of dots in the margins, and it was clear that the dots were some kind of code. They dubbed it the ‘Dot Code’. Before long, they had broken it.

The secret message began: ‘Ever think of Cathay? Blissful afternoons with precious naked in my arms.’ It then got quite lurid and explicit before ending with the line, ‘Many kisses, much love.’8

One thing was clear. The secrets of the nation were safe from this pair. They weren’t engaged in espionage; they were having an affair.9

Room recommended that the army let the mail through to the lovers, but that it add a series of dots on the outside envelope in a Dot Code message that said, ‘Careful! — The Censor’.

There is no record of whether the army took the professors’ suggestion seriously.10