The illusionist, p.19
The Illusionist, page 19
In civilian life, having roamed across several creative fields, Simpson was regarded as respected in each but not outstanding in any of them. ‘His lively mind refused to be confined,’ The Times would write later. But Maunsell thought he showed great promise. ‘Brilliant but unstable’ was his assessment, noting that Simpson suffered from bouts of depression.
That lively mind had made Simpson the obvious choice to play the role of Nicosoff, the Axis agent working far behind enemy lines. His language skills turned out to be useful after all: Nicosoff communicated with his handlers in French.
In early 1942, Maunsell was handed something that suggested Simpson wouldn’t need to work on the case much longer: decrypted internal Abwehr messages about a spy in Egypt named ‘Roberto’. This, it was apparent, was the German codename for Nicosoff. In October, ‘Roberto’ had been rated ‘credible’ as he passed on information from his friend ‘Piet’ about British troop movements. There had been discussion between Berlin and Athens about how to get money to their man.
But three days after the start of ‘Crusader’, the tone had changed. ‘The reliability of Roberto reports must be subjected to severe doubt,’ one message read. He had reported a division of troops leaving Palestine for the Caucasus when the Abwehr knew it to be in Egypt. Their man’s silence in the run-up to the attack ‘is likewise very striking’. Even when ‘Crusader’ would have been common knowledge in Cairo, he hadn’t mentioned it. By the start of December, they seemed to have reached a conclusion: ‘Interference by the enemy intelligence service in the Roberto network is becoming more and more obvious.’ Around the same time, SIME learned about Levi’s arrest in Italy. The prospects for Cheese seemed bleak.
But Simpson wasn’t willing to give the operation up. The Germans, he argued, might be persuaded that the problem wasn’t Nicosoff but his source, ‘Piet’. Nicosoff might yet be redeemed. He was given time to try.
Simpson approached the problem with a novelist’s eye for character. He constructed a picture of Nicosoff fitted around the details which he knew Levi had given the Germans, and trying to avoid areas where the Italian might have invented new details he didn’t know about. This wasn’t straightforward. Levi had told the Germans that Nicosoff was born in Egypt to a Syrian family. But he had also, without consultation with anyone else, invented the surname Nicosoff. Some in SIME suspected this was a pun on ‘knickers off’, which given Levi’s main interest in life wouldn’t have been a surprise, but whatever the reason, the distinctly Russian-sounding name created a problem. ‘He must be of Slav origin’, a SIME officer decided.
Many of the notes on Nicosoff were in that vein, as though written by agents trying to establish who this man was, based only on his wireless signals. ‘There is no positive evidence at the present time that he has any other profession than that of spy,’ one said. ‘On the other hand he apparently has enough money for his living expenses. He must have, or have had, some financial standing, in order to be able to borrow money at a time when he was – according to the evidence of the messages – absolutely on his uppers.’
Perhaps this was simply how the staff of SIME, who spent so much of their time trying to catch spies based on clues, were used to thinking. Or perhaps they were putting themselves in the place of Nicosoff’s German handlers, trying to work out what picture they would have assembled of their agent.
His job was listed as ‘unknown’. They toyed with the idea that he was a ‘half-commission man’ – someone who introduced clients to brokers, in return for a cut of the proceeds. ‘He would frequent (at the upper end of the scale) such places as the bar of the Metropolitan; and (at the lower end of the scale) would not be out of place in a better-class Arab cafe or an other ranks’ eating-shop. He would not mix socially with British officers, but might with Greek, Polish or Yugoslav.’
As for past jobs, he had to have learned how to operate a wireless transmitter at some point, though SIME couldn’t quite decide how. ‘If we place his age at about 45, he might have been a ship’s wireless operator in his younger days,’ they said, though they generally worked on the basis that he was between thirty and thirty-five.
Simpson was sure that Nicosoff could win the Germans’ trust once again. He painted a picture of a man who had been deceived by a greedy source the previous year, and had since been abandoned by unreliable helpers. ‘I am alone,’ Nicosoff complained to his handlers, ‘and without funds I cannot hire agents.’
Though he transmitted twice a week, the intelligence content of his messages was minimal. He had been abandoned by the rest of his spy ring. ‘All the others, and all the sources of intelligence, have lost faith in the enemies’ promises to send money, and have left him hanging on alone in the faint hope that such money may yet arrive,’ Simpson explained. ‘This is the burden of every message, and it is only varied by occasional and unimportant tit-bits of information such as anyone might pick up in Cairo.’ In February, he reported having seen American troops in Cairo, and that he’d heard an aircraft factory was being built near the city. But when pressed on these, he offered little more detail.
Not that Nicosoff’s handlers seemed to care much. Signals went unacknowledged. Replies, when they came, contained few questions on military subjects. As for the requests for money, there was the occasional reply that it was on its way, but also suggestions that perhaps he could come somewhere – Istanbul, say – to pick it up. SIME was suspicious of a trap. As Nicosoff, Simpson replied that he didn’t have the cash to get to Turkey.
For months, Nicosoff and his German handlers maintained these half-hearted exchanges. Could he locate the 23rd Infantry Division? the Abwehr asked. ‘No money,’ came the reply. ‘No agents to collect information.’ Agent ‘Roberto’ barely appeared in internal Abwehr signals.
To Simpson, the problem with the character of Nicosoff was that this apparent mercenary had been working for nearly a year without pay. This ought to make him suspect in German eyes. With no sign of any money on its way, he played Nicosoff as sulky. He sent a message threatening to sell his wireless to pay off the debts he had run up in his service of the Axis. When still nothing came, he grew sarcastic. Whenever a promise of money arrived, he was immediately enthusiastic.
The aim was to leave the Germans with the impression of a potentially useful agent who was losing faith in his controllers. Simpson’s main worry was that the Germans would lose patience first. They certainly didn’t seem very bothered about their man in Cairo. That lack of interest should have been a clue to SIME that they had a security problem. The reason German intelligence wasn’t bothered about its man Roberto was only partly because he had misled them over Crusader. The other, bigger reason was that they had a far better man in Cairo.
The Cut and Restored Rope June–July 1942
The Magician holds up a rope and a pair of scissors, and proceeds to cut the rope in half. They tie the two halves together and then slide off the knot, revealing the rope has been made whole again.
Chapter 27
Colonel Bonner Fellers, US Army, had been a fixture in Cairo since late 1940. In his mid-forties, still handsome despite his concerns about losing his hair, he was the military attaché at his country’s legation. For more than a year, he had been his country’s main observer in the only place the British army was actually fighting. His reports were read by the president.
With America supplying so much equipment, his good opinion was vital to the British, and as a result they let him go where he wanted and see whom he pleased. It helped that he had a great deal of charm. ‘An original and delightful person,’ Hermione Ranfurly wrote after meeting Fellers in Cairo in early 1941, ‘who seems to say exactly what he thinks to everyone regardless of nationality or rank.’ When her husband, an army officer, went missing during Rommel’s first advance, Fellers had wired to the American delegation of the Red Cross to see what they could find out. A fortnight later he rang to pass on news: ‘Ranfurly captured. Last seen in good health.’
His motives may not have been entirely altruistic. The Countess of Ranfurly, as she was formally known – her imprisoned husband was the sixth earl – was a useful person to cultivate. She worked for the Special Operations Executive, and knew everyone. She was no mere socialite, either. By mid-1941, concerned at the way SOE was operating, she was sneaking documents out of the office in her bra each evening to pass to Wavell. Her subterfuge led to a restructuring of the troubled organisation that summer.
Fellers wasn’t aware of all of that, but he was formidably well-informed, and his reports back to Washington reflected it: troop movements, plans, the opinions of the most senior officers were all in there, along with Fellers’ own often caustic observations about the British position. Which made it all the more unfortunate that in early 1942, Rommel was reading them within hours of their dispatch.
The previous September, Italian intelligence had stolen codebooks from the US embassy in Rome and then copied and replaced them before their absence was noticed. Axis spies were now reading all sorts of US diplomatic traffic, but none was valued as highly as that of Fellers. Here was the man in the room, passing on the details of every discussion. He was the ideal spy, touring battlefields with a soldier’s eye for what mattered. In January of 1942 alone, he had told his masters – and Rommel – of the transfer of 270 aircraft from North Africa to the Far East, and given them complete details of the British tanks in the theatre, including their location and state of repair. He opened February with a report on forthcoming Commando operations.
‘Personnel losses of the British are fairly light but loss in materiel heavy,’ began one long report at the start of June. ‘It is estimated at 70% of British tanks engaged were put out of action and at least 50% permanently destroyed. The air ground liaison was poor and the RAF repeatedly bombed own forces.’ The next day he sent more detail. ‘It is believed that the Southern Brigade of the 50th Division has been completely destroyed,’ he wrote, before going on to give the locations and strengths of specific units. It was all, one of Rommel’s intelligence officers marvelled, ‘stupefying in its openness.’
In Germany, information from what was known as ‘the Good Source’ was rushed through the system, translated, re-encoded and transmitted to Libya with the goal that Rommel, at lunchtime, would know where Allied troops had been the evening before.
There was nothing quite like it. When Allied codebreakers read German messages, they had to go through a process of interpretation, working out who was talking to who and deducing, from the content of their messages, the location, status and intentions of different forces. Rommel had none of these problems. He was reading a report on his enemy’s situation, written with complete frankness by one of the US Army’s best intelligence officers. It was the assistance he needed as he drew up plans to resume his attack.
That April, as he went through the samples of decrypted German communications – known as ‘Ultra’ – that were placed in his daily box, Winston Churchill noticed something odd. The Luftwaffe had received an urgent warning that the British knew the location of its secret desert headquarters. Churchill scrawled a note to Menzies, the head of MI6: ‘Please report on this. How did they know that we had told the Army in Egypt where it was?’
It was a good question. Menzies replied that an investigation was underway, which was surely true when he said it, even if it hadn’t been before then. It was looking for security weaknesses at Grey Pillars. It was hardly surprising that senior British officers quickly settled on the American in their midst: it’s always easier to blame an outsider. On this occasion, though, prejudice delivered the right answer. Bonner Fellers had indeed been the German source.
In April, this was simply a guess. But it led Auchinleck’s staff to begin drafting new rules for what the American could see. It was a tricky area: he was the representative of the ally that was sending them the tanks they badly needed. It wouldn’t do to offend him. No one wanted him to know that he wasn’t being told things. But they wanted to stop telling him things.
They also sat him down for a chat about his security measures. They didn’t tell him why, but Fellers was no fool, and he started to wonder what was wrong. As it happened, he had his own doubts about US codes. In February, he’d asked Washington for assurances that the codes were secure. Of course they were, came the reply.
In May, the staff at Grey Pillars settled on new rules of engagement for Fellers. They asked him to show them anything potentially compromising before he sent it. He promised he would. He didn’t. In early June, Churchill’s box of Ultra decrypts included one with an alarming cover note: ‘Another long report to German Army in Africa from “Good Source” concerning British morale, training, supplies and intentions, evidently based on the Good Source’s visits to British units.’
For the previous two months, MI6 had been hunting the Good Source. At Bletchley Park, they’d collated clues from the messages they’d been able to decrypt. They realised the term wasn’t a descriptive evaluation that could be applied to reliable intelligence, but a reference that applied to a specific channel. But who? Where? One message, complaining that the British were wasting US aircraft and spare parts, sounded like things that had been said by American manufacturers who’d visited the Middle East. Could the Good Source be in the US War Department in Washington?
The message in Churchill’s box nailed it. Menzies wasn’t sending the prime minister another piece of the puzzle to infuriate him, but instead its solution. Included in the signal was the final clue. The Good Source had visited the front in North Africa, and reported back: ‘Training inferior according to American ideas.’
The Good Source was in North Africa, with access to Grey Pillars, able to visit the front line, and his point of comparison was the US Army, which he naturally assumed to be superior. It wasn’t much of a mystery any more. ‘I am satisfied that the American ciphers in Cairo are compromised,’ Menzies wrote in a note to Churchill included in the same box. ‘I am taking action.’
That was even more sensitive than it might have seemed. There is a certain delicacy involved in telling your closest ally that they are the victim of a security leak. The job was given to Bletchley Park, who began an exchange of messages with their counterparts in Washington, the Signal Intelligence Service.
But as well as warning the US that its codes had been compromised, Bletchley had to protect its own work. The British knew that the Germans were reading the American code because the British were reading the German code. If they acted too clumsily, there was a danger that the Germans would start assembling their own puzzle, and realise that someone was doing to them what they were doing to the Americans. The message to Washington contained a plea to say nothing at all to Cairo.
For several days, the two sides consulted, hampered by time differences and delays as the Americans considered pieces of evidence that were being sent piecemeal in transatlantic cables. Despite what Menzies had told Churchill, both ends of the conversation were aware that there were other ways in which Axis intelligence might have read Fellers’ telegrams. An enemy agent might have access to the unenciphered messages, either in Cairo or Washington.
In the US, there was reluctance to believe what they were being told. The implications were too awful. If the code was compromised, US communications from all their diplomatic outposts were being read. Who knew what other secrets had been blown in messages from their embassies?
As for Fellers, Bletchley Park was still compiling evidence about what the Good Source had been giving the enemy, but in Washington, a flick through his cables over the previous six months would reveal that, if he were indeed the leak, he had been giving them everything.
And while they delayed action, he carried on giving. Even as the discussions were going on, he wired Washington about imminent plans for Commando raids on nine Axis airfields. At Bletchley Park, reading Good Source warnings based on that cable, they were apoplectic. ‘Further information from Good Source reveals our future plans,’ read one cable to Washington, ‘and matter becomes of extreme urgency.’
Even as the Bletchley Park analysts read the Commando message, they knew there was little they could do. Too blunt a warning to Cairo that the raids were expected would risk all their work. All they could do was tell Cairo that the Germans were increasing security at airfields. It was, in any case, too late to get a message to the Commando units far out in the desert.
The news from the raids would take days to arrive. The SAS claimed to have destroyed up to forty planes. German reports the next day suggested the number was ten. But it was clear that the raiders had been expected: one group trying to sneak onto a base found themselves surrounded by German soldiers. Only one of those men made it home.
Meanwhile Fellers was still sending his dispatches, unaware they were being read. And now Rommel was closing in on Cairo.
Chapter 28
At the end of May, his strength consolidated, his picture of the British defensive line better than anyone in Cairo realised, Rommel had launched his assault.
He used deception techniques of his own, sending forward trucks mounted with aircraft engines on his northern flank, to blow up a dust cloud that looked like it had been caused by a fleet of tanks, while at the southern end of the line his real force pushed forward, overrunning the surprised defenders.
The Eighth Army – with more men and more tanks than the Afrika Corps – fought back, and at times came close to winning, but while its commanders failed to exploit opportunities, Rommel, short of fuel and ammunition, managed to squeeze his way out of trouble again and again. The men on the ground fought bravely, desperately, but by the middle of June British forces were in full retreat, trying to get to the defensive line on the Egyptian border ahead of their pursuers. ‘The battle has been won,’ Rommel wrote to his wife. ‘The enemy is breaking up.’ He saw the possibility of a victory by the following month so total that he would be able to get home. ‘Perhaps we will see each other in July after all.’
