Fighting to Lose (11 page)

RANTZAU’s message to Eschborn presented an opportunity that had not been thought of: he could be another double agent. He was approached, and, after some arm-twisting, agreed to reply to RANTZAU. MI5 was especially keen to develop the contact because Eschborn was an amateur photographer and the Germans were proposing that he help them develop the technique for making miniature photographs. MI5 saw this as a means of staying on top of this enemy initiative, oblivious to the fact that any commercial photographic studio, either in England or Germany, would be better equipped for the task.33

At about this time, MI5 began applying code names to its double agents: Owens was assigned SNOW, a jumbling of the letters of his surname; Gwyllem Williams became GW; and Charles Eschborn became simply CHARLIE. These code names seem an accurate indication of MI5’s level of sophistication.34

Now things really sped up. Robertson vigorously petitioned his intelligence contacts in the army, navy, air force, and MI6 to come up with convincing material to feed the Germans. Owens was released from all supervision and encouraged to roam around the country looking for items of interest, as he would if he were a real spy. Robertson told him to try to penetrate into restricted areas to give eyewitness credibility to the stories he would tell DR. RANTZAU. He gave him thirty gallons’ worth of petrol coupons and told him to visit as many aerodromes as he could; and when in Newcastle, urged him to see whether he could, by himself, locate the headquarters of 13 Fighter Group. He also suggested that Owens snoop around Harrogate, the spa town that had been taken over by the Air Ministry in anticipation of the bombing of London.35

These were wonderful opportunities for Owens as A-3504. An ordinary spy approaching even one aerodrome in wartime Britain risked execution if caught, and so would go cautiously. Owens could visit as many aerodromes as he could manage with absolutely no fear should he trigger base security. It emboldened Owens to gather information with an aggressiveness no ordinary spy could match. In one case, he talked his way past the entrance gate and drove around an airfield counting the parked aircraft. In the pubs of Harrogate, he cozied up to airmen and asked questions delicate enough to get him shot. Captain Robertson could not have made it easier for Major Ritter’s A-3504.

To better assess the accuracy of the information Owens was collecting, Robertson again enlisted the help of Major Boyle over at the Air Ministry. Between them, they compiled much of the RAF’s order of battle in England, supplementing “the complete list” of squadrons supplied by Boyle with contributions from MI5’s newly created local security control officers (SCOs).36 Neither man seems to have been concerned that by centrally collecting this data they increased the risk that the whole package could be leaked to the enemy.

As for whether Owens was telling the enemy more on his trips to the Continent than he should, Robertson put on record his conviction that Owens was entirely trustworthy: “He is a stupid little man who is given to doing silly things at odd moments, but I am perfectly convinced he is straightforward in the things that he gives me and the answers to my questions.”37

Throughout this period, according to the Hamburg–Berlin teletype messages that Robertson never saw, this “stupid little man” steadily provided the Germans with high-grade intelligence on his trips across the Channel, some of it supplied by Major Boyle and some of it apparently gathered by Owens himself.

4

January–April 1940

By the beginning of 1940, it would have been patently obvious to Ast Hamburg’s Major Ritter that his British opponents had little clue about how to run a wireless spy. He must have shook his head when Owens told him that the MI5 wireless operator who sent A-3504’s reports was a civilian, not a qualified military person, and that the people trying to run him as a double agent had little knowledge of codes and ciphers. It was the perfect circumstance for
Funkspiel
— the epitome of the spymaster’s art.

It was initially not as ideal as Ritter might have imagined. Britain’s War Office wireless listening service — formerly MI8(c) but now renamed the Radio Security Service (RSS) — was not looking for spies, but only for illicit signals that might be used as radio beacons to guide in German bombers. Only suspicious transmissions in Britain were being sought, and they were not being monitored at all for their content.

This changed in December with the arrival at the War Office of a Canadian army signals officer who was looking for advice on how to handle the obviously clandestine wireless traffic being intercepted by a Canadian signals intelligence unit in Ottawa. Within two weeks, E.W.B. Gill, an Oxford professor and former wireless intelligence officer of the First World War, was dispatched to the Radio Security Service, then still based at Wormwood Scrubs (London) along with MI5, to have it abandon the radio-beacon searches and listen instead for what might be German enemy-agent transmissions. As the Canadians had discovered, there was plenty of suspicious stuff out there, but neither MI6 nor the Government Code & Cipher School had shown much interest.1

On learning that an RSS volunteer operator was handling the wireless communications of a double agent code-named SNOW that MI5 was running, Major Gill undertook to have the German side of the traffic studied. It was soon found that Hamburg, SNOW’s control station in Germany, was exchanging messages with another station, whose signal was moving along the coast of Norway — a ship. The cipher used was similar to the
congratulations
code that had been given to Owens, and Major Gill, with the help of Lieutenant Hugh Trevor-Roper, a twenty-six-year-old German-speaking Oxford scholar who had arrived with him, managed to puzzle it out.

The name of the ship was the
Theseus
, and it was reporting on neutral ships headed for British ports. On January 29, Gill forwarded this information to the Government Code & Cipher School, but instead of a thank-you, its chief, Commander Alastair Denniston, fired back that Gill should stick to listening and leave the cipher-breaking to the experts. Undaunted, Gill continued to have the
Theseus
traffic monitored, and extended the listening to other stations across the Channel that appeared to be exchanging transmissions of a similar type.2

The
Theseus
was a “spy ship,” as the term was applied during the Cold War to vessels that roamed the high seas listening in on foreign radio communications using sophisticated wireless receiving equipment. Ships could be positioned precisely where the signals of a distant transmitter could best be heard.

In the case of spies, once their signals were picked up and copied, the ship could be moved to where it could best retransmit them to the mainland station. In the case of the
Theseus
, this would have been Hamburg. It worked the other way, too. The wireless operator for SNOW could be made to hear the
Theseus
, depending on where the vessel was transmitting from.3

The choice of wireless operator for the
Theseus
must have been deliberate. He was twenty-four-year-old Friedrich Kaulen, son of a German merchant. Fair-haired and good-looking, he was R-2220, an agent of Nest Bremen. His Abwehr assignment immediately before the
Theseus
had been two years as a spy in England for Abt I Luft, and he had received kudos for his excellent photographs of British secret air fields, anti-aircraft batteries, and searchlight emplacements.

Ordinarily, the wireless operator on a military vessel would be a naval rating or, in the case of a spy ship, perhaps an army signals specialist. Kaulen, however, was an amateur civilian radio hobbyist, much like the volunteer operators who did much of the listening for the RSS, and like the operator who transmitted for Owens. Kaulen’s only second languages were English and French, an odd choice for a wireless operator who was to remain for months in Norwegian waters. On the other hand, if the Germans wanted to do a wireless deception — a
Funkspiel
— on the operator of Owens’s transmitter, should they have figured out that it was under British control, it would make sense that they would use a person just like Kaulen, and do it from a ship.4

Kaulen’s ability to understand English would be handy for confirming that he had Owens’s signal, since he only spoke English. As for getting the correct frequency, that would not have been a problem: as a Nest Bremen spy, Kaulen’s commanding officer would have been the Abt I Luft chief at Ast Hamburg — Major Ritter.

At Major Gill’s urging, the Radio Security Service in London began avidly collecting the new traffic and, because he and Trevor-Roper shared a flat, they would work on the material in the evenings. By simple anagramming, they found they could break other messages, and these revealed that Hamburg was in contact with spies in Belgium, Holland, and Luxembourg.5

On March 20, at a meeting with Gill and Captain Robertson of MI5, Commander Denniston of the Government Code & Cipher School agreed to assign one of his codebreakers to the traffic.6 This was to be sixty-six-year-old Oliver Strachey, who had briefly held a job with the India railway system before marrying the suffragette Rachel Conn Costelloe and enjoying a reversal of household roles until his First World War stint with MI1(b), the War Office code-breaking agency that Colonel Simpson had headed. He stayed on after the war when MI1(b) merged with the navy’s Room 40 to form the Government Code & Cipher School (GC&CS). He was then with a GC&CS team working on German naval traffic.

Gill and Trevor-Roper were convinced they had found the Radio Security Service’s noble purpose. They saw it as opening a window on the operations of the German secret services. Denniston remained cool. Strachey did not produce a decrypt from the new intercepts until April 14, a week after the German invasion of Norway.7

Denniston’s dismissiveness is indicated by the fact that Strachey was no more than one senior citizen with a pencil and paper, whose output was assigned the lofty and surely sardonic title, Intelligence Service (Oliver Strachey) — ISOS. Denniston had good reason to be skeptical. The ciphers used for the messages Gill and Trevor-Roper were so excited about were of a First World War vintage that, thanks to the indiscretion of a former director of Naval Intelligence, the Germans had to know the British could easily break. Denniston could be sure they would not be using them for messages of any real value.

The former Naval Intelligence head in question was the famous Admiral Reginald Hall, who had presided over the codebreakers of the Admiralty’s Room 40 during the First World War. He is widely credited with covertly releasing the intercepted “Zimmerman Telegram,” which in 1917 helped bring the United States into the war. In 1919, the war over, he left the navy in a swirl of ill-feeling, probably having to do with how difficult he had been to work with while serving in his positions of high responsibility. He tended to be arrogant and autocratic, and there were some in senior government circles who probably were not sad to see him go, shoved a little, perhaps, by not giving him certain honours that he might have felt he deserved. He took with him some ten thousand decrypts of German navy, Foreign Office, and espionage messages, and stashed them at home.

In 1925, an American civilian lawyer approached Hall on the off chance he could help him with a case he was working on aimed at getting Germany to pay for damages for the “Black Tom” explosion that shook New York Harbor in 1916. What he needed was hard evidence linking the destruction of a munitions depot to German agents. Hall led Amos Peaslee to the stacks of decrypts, gave him the run of the house, including servants, for as long as he wanted, and went off to Scotland on a shooting holiday. Three days later, Peaslee had found and copied 264 cables and radiograms pertinent to German First World War covert operations in America.

The decrypts ignited a blaze of publicity when the case was heard in The Hague in 1927. The German government may have had its suspicions, but it had no idea of the extent to which the British had been reading its secret communications, and probably still were. The immediate and lasting consequence was the Government Code & Cipher School was locked out of all German Foreign Office and German army traffic, the Foreign Office switching to unbreakable one-time pads (single-use sheets of random letters whose numerical equivalents are added to those of the letters in a message to encipher it) and the army to high-security plug board ciphering machines (akin to early telephone switchboards, they allow for the creation of thousands of unique electronic circuits).8 It stood to reason that the Abwehr, the German army’s secret intelligence service, would have taken similar measures.

To make the breach in secrecy even worse — and it could hardly have been — the types of German ciphers compromised were subsequently described by the American cryptologist Herbert Yardley in the popular book
The American Black Chamber
(1931) and in detail by Helen Fouché Gaines in her
Elementary Cryptanalysis
(1939). The most vulnerable had been transposition ciphers, which at their weakest were solvable by anagramming.