The illusionist, p.24
The Illusionist, page 24
Next, they set up the stage. When the Eighth Army launched its attack, it would need fuel, ammunition and supplies, vast amounts of them. These needed to be brought up to the front and stored as close to the men who would be doing the fighting as possible. But growing piles of supplies were exactly the sort of clue the enemy would be watching for. These dumps were, at the best of times, a logistics problem. The petrol was a fire hazard, especially given the army’s standard four-gallon can tended to leak. The more mundane supplies were a target for thieves.
Scouting the northern end of the line near Alamein station, five miles behind the front line, Ayrton found dozens of slit trenches that had been dug for defensive reasons the previous year and were now unused. They were lined with brickwork, an army habit that, while it definitely made the positions look neater, drove Barkas spare, because from the air it meant clean lines with jet black shadows that were easy to spot in reconnaissance photographs. But Ayrton saw an opportunity. The Axis photo-interpreters would be used to seeing the trenches by now, and so long as they looked the same, they wouldn’t pay any attention to them. He began experimenting with stacking petrol cans inside the trenches, and discovered that, pressed against each wall of the trench, they made no visible difference to the shadows. By night, two thousand tons of petrol were sneaked into the trenches, ready to refuel the advancing tanks. Even from close-up, it was hard to tell the difference.
That left the question of the rest of the supplies. Sacks of sugar and flour could be hidden here and there in shallow holes in the ground, and covered with camouflage nets, but much of what they were dealing with was in crates that were too large and too many to be hidden. They would have to be disguised.
Over the course of four nights, a work party of 80 Australian infantrymen found themselves stacking boxes of supplies in the shape of army three-ton trucks, and then covering them with camouflage netting. Excess boxes went under small tents, and the result looked from the air like soldiers making camp just back from the line. It took 30 ‘trucks’ to conceal close to 600 tonnes of supplies. The most coveted goods – cigarettes, sugar and milk powder – were in the centre of the stacks, to discourage pilfering.
Fifteen miles further back, 3,000 tonnes of ammunition was hidden under hessian sacking and sand in an existing ammo dump. Larger crates were again disguised as trucks. The stacks were covered with canvas painted to look like vehicles, complete with wheels. Barkas requested reconnaissance flights to check how all this appeared from the air.
While the real dumps in the north were being hidden, in the south another unit was working under the supervision of Brian Robb, a Punch cartoonist whom Barkas had found manning a searchlight in the Sinai desert, bored out of his mind. He was tasked with creating a fake supply dump in an area on the map simply marked ‘Brian’.
The focus of the camouflage efforts was on hiding real equipment, so Robb had to work with what he could get his hands on. Much of it was ‘two-dimensional’, aiming to create something that would fool air reconnaissance. Without the manpower to dig real trenches, matting was laid along the ground the simulate the shadows they would create. Beds turned on their sides and covered with hessian and camouflage netting passed for petrol cans and crates of supplies.
More ingenious – and elaborate – was the water pipeline that snaked south and west towards the front. It too was a fake, built deliberately slowly to give the impression that the British were on a relaxed schedule. ‘Construction’, such as it was, involved digging a trench, with petrol tins beaten into the shape of pipe left alongside, and then filling it in at night and moving the fake pipe forward to the next ‘section’. Every so often there was a ‘pumping station’, with scarecrow figures posed alongside. Real army vehicles going to and from the front were re-routed to travel alongside the ‘pipe’, to stir up dust and add to the impression of activity.
None of this was central to the big trick that Barkas, under Richardson’s supervision, was performing in the desert, but all of it served to reassure the viewer of the false picture he was trying to show: a British force slowly getting ready to launch an attack at the southern end of the line.
On 18 October the grand illusion began, as X Corps began to move into position. A common misapprehension about magic is that sleight-of-hand relies on speed. In fact quick movements draw attention. It is the slowness of the hand that deceives the eye, by reassuring the audience that there’s nothing suspicious going on. In the same way, the advance of X Corps was designed to put the Germans at their ease.
The stage that Barkas had set for Behrendt and his colleagues to watch contained, in mid-October, three visible elements. There were the dummy pipeline and the dummy supply dumps in the south, and, in the north, what seemed to be thousands of trucks, parked across the desert. These vehicles would definitely have been noteworthy to German intelligence. They would be essential in any battle to bring men and supplies forward. But to someone looking for signs of an imminent attack, they were secondary to armour and infantry. And they were just sitting there, as they had been for weeks.
The first change to that came with the move of the 9th Armoured Brigade, in daylight, to a staging area well in the rear, codenamed ‘Murrayfield’. That was exactly the sort of thing Behrendt was looking out for. But what did it mean? That far back from the line, they might just be training. If it was a first step towards the front, they could go either north or south from that position. It was something to watch closely, but not an immediate threat.
That night, things got busy. The 9th Armoured moved north, to the area where the trucks were apparently sitting around. Far from being a random dispersal of vehicles that weren’t needed, this was the part of Barkas’s stage where most of the action was going to happen. It was codenamed ‘Martello’. Days earlier, the tank crews had been taken there and shown exactly where each of them was to go. What was waiting for them was a genuine magic prop designed for them by a real magician.
The previous year, when Jasper Maskelyne had first been put in charge of experimental camouflage, Barkas had shown him a sketch, drawn by Wavell, of a tank with a cover over it that made it look like a truck. Could he make this? It was a task for which he was probably better qualified than anyone else in the army. This was a stage prop that hid something, and Maskelynes had been building those for three generations. Before he’d left the camouflage section he’d come up with a design that met the requirements.
Constructed out of steel tubing and painted hessian cloth, the covers came in two halves, one for each side of the tank. Lifted over the tank and bolted into place, they completely covered it. The result had tracks not wheels and the tank’s gun barrel sticking through the windscreen. But from a distance, it looked like a truck. Aerial reconnaissance was asked to photograph it and give a view. ‘Perfect,’ was the response.
They called them ‘sunshields’ – a name chosen to hide their purpose – and there were hundreds of them now in Martello, each one numbered and allocated to a tank. The camoufleurs had rested them on barrels which, from a distance, looked like their wheels. They weren’t going to fool anyone close-up, but a magician knows where his audience is, and Barkas would be keeping his well back.
The same storm that had grounded the Luftwaffe that week had ravaged Martello, demolishing sunshields and dummy trucks. But the Camouflage section had scrambled to repair and replace then, and when, on the night of 18 October, the 9th Armoured arrived, they found their sunshields waiting for them. Each crew lifted its cover onto its tank and secured it in place. It was plausible, especially if someone wasn’t looking very hard.
Meanwhile the 24th Armoured Brigade had occupied Murrayfield, moving up in the darkness. When the sun came up on 19 October, everything looked as it had the previous evening. There were the trucks in the north, there was the brigade in the centre.
Perhaps if someone had looked very closely at the reconnaissance photos, they might have noticed changes, but there were more interesting things to look at that day. The 10th Armoured Division was on the move, and it was clearly moving towards the southern end of the front. It stopped behind the Brian false dumps, in an area the British codenamed Melting Pot. To someone who already thought an attack in the south was likely – hadn’t that been the way Rommel had wanted to go weeks earlier? – and who expected to be able to see it in advance, this was a clear threatening move.
But nothing happened for two days. The Germans were allowed to relax. The 10th Armoured sat where it was. The pipeline slowly continued. Nothing was imminent.
Then, in stages at night between 18 and 22 October, an awful lot happened at once. All of X Corps, hundreds of tanks and thousands of men, moved to Martello, assembling sunshields over their vehicles in a scramble to ensure that when the sun came up they were hidden. Behind them, teams supervised by Barkas’s camoufleurs hustled to assemble dummy tanks and vehicles at Melting Pot and Murrayfield. The Royal Engineers moved across the desert, erasing the tracks made by the tanks. As the sun came up each dawn, it was as though nothing had changed. But on the morning of 23 October, Montgomery had an entire armoured corps ready to attack, and the enemy knew nothing about it.
The ball had moved from one cup to another, under the eyes of the audience, and the audience hadn’t seen a thing.
Chapter 35
While Barkas was manipulating the shells and balls in sight of the audience, Wild was delivering the magician’s patter, telling the story over the top of the Eighth Army’s manoeuvres: Monty was holding back; the generals were having a conference in Iran; the British were worried about the Caucasus.
The codename for this story-telling was ‘Treatment’, and Wild drew it up and ran it according to Clarke’s instructions. From Tehran invitations went out to Wavell in Delhi and Alexander in Cairo for a conference on 26 October, and staff set to work arranging the logistics. Only the generals knew the meeting would be called off at the last moment.
At the front, senior Eighth Army officers sent messages back to Cairo reserving hotel rooms for the days after 20 October. The staff college at Haifa announced it was extending its current course, and wouldn’t finish it before mid-November. Each of these was a clue that, with luck, would make its way in some form to the Abwehr, a jigsaw piece suggesting that the top brass wasn’t expecting to be doing any fighting at the end of the month.
In Istanbul, Wolfson played his part, spreading rumours as instructed by Clarke, including placing a story in a Swiss newspaper that the Eighth Army was in a defensive posture, with forces being sent to protect the Caucasus. He sent a clipping of that to Clarke. ‘In this particular one I thought it best to stress only the “defensive positions” part of your telegram,’ he wrote, ‘arranging for the “offensive postponed” to reach the enemy simultaneously, but through another channel, as being too obvious otherwise.’
In the desert, a stream of false radio traffic, both coded and in clear voice, was broadcast to give the enemy the impression of an army going about its business. German direction-finders would be able to see that these signals were coming from well behind the front line.
And in Cairo, Cheese was busy reporting sightings of troop transport boats in the harbour of Alexandria – perhaps getting ready to go to Crete – and numerous troops on leave from the front – just as leave was being stopped. At the end of September a question was passed on from Berlin: did he have any news about British plans to attack Rommel? ‘When, how and where?’
Nicosoff didn’t directly respond to that. ‘So many rumours circulate about a forthcoming English attack in the desert that it is difficult to ascertain a definitive date,’ he signalled in mid-October. He said he’d spoken to an American officer friend who was on leave. ‘He is of the opinion that the attack will take place at the beginning of November.’
Anything more direct than that would have been a risk. He wasn’t supposed to be someone with access to that sort of information. He was transmitting most days, messages as short as one sentence and rarely longer than ten: there was less traffic on the desert roads than there had been in August; he needed money; Greeks were always talking about attacking Crete; when would the money be arriving?
Clarke took the view that the best way to smuggle a lie was hidden amid a lot of truth. He told his subordinates that 90 per cent of what they sent to the enemy should be confirmable from other sources – although those sources might be the result of other efforts to mislead them, such as the fake divisional signs of the Cascade deception.
So a lot of what Nicosoff was sending was correct, but trivial – ‘chickenfeed’, the spies called it: the newspapers had reported the execution of five spies in Aleppo; locals complained about German bombing during Ramadan; there were Scottish soldiers in Cairo wearing an ‘HD’ insignia, which his masters would be able to identify as the 51st Highland Division. The impression was of a busy agent, developing sources among soldiers on leave and civilians in sensitive jobs, travelling back and forth between Cairo and Alexandria in an effort to gather information. Sure, he complained about money, but he was coming up with the goods. The 51st duly went into Behrendt’s order of battle.
In the messages he drafted for the Cheese channel, Evan Simpson was telling two stories. The first was the one set out by Clarke, offering puzzle pieces that pointed to a large and expanding British force in Cairo, threatening Crete but not yet ready to move against Rommel. The second was the tale of Nicosoff himself, an enthusiastic spy who was increasingly conscious of his own importance to his masters and required delicate handling. When he had been out of favour at the start of the year, Nicosoff had been humble. Now he was demanding. ‘What happened yesterday?’ began one angry message, after a communications failure. ‘Impossible to work with your incompetent radio operators. My work is not worth the risk if communications are so uncertain. It has to be your best operator every night.’ As for a request for specific information: ‘Impossible to move without money.’
Had he been real, poor radio operation wouldn’t have been Nicosoff’s only complaint. Some of the information requests he received were, to put it mildly, ambitious. ‘Try to find out if the British attack will be combined with a flank action,’ read a request in mid-October. ‘If possible also give the exact direction of this attack and the forces that will take part.’
The place the Abwehr believed this attack might come from was Northern Chad, a thousand miles away on the other side of the desert. It was not clear how Nicosoff’s controllers imagined he would check to see who was on the march from there. Probably they were still yearning for the days when Bonner Fellers sent them detailed reports on British strategy. Like many masters of distant servants, the Abwehr wanted a lot but had only the haziest grasp of what was realistic.
As the date of Operation Lightfoot approached, the team at A Force headquarters considered the part Nicosoff should play. A year earlier, Shearer had been prepared, with Clarke’s agreement, to destroy the Cheese channel as the price of success of Operation Crusader. But A Force was now more ambitious for their imaginary radio man. They wanted to keep him going as a supplier of false information to the enemy. The question was how best to maintain his credibility. Ideally, he would transmit something about the attack in advance, but what – and when? Usually Nicosoff went on the air around 7.30 p.m. – two hours before the operation was due to begin. De Guingand ruled out giving the enemy advance notice. Wild feared it would be implausible for the agent to go on the air two hours late.
Another option would be simply for him to learn of Operation Lightfoot at the same time as the rest of Cairo. That would be credible, but it would also undermine his value to the Germans as a spy.
Instead, A Force began drawing up a little playlet for the evening of 23 October, when the attack was due to begin. It became known as ‘Cheese Cake’, and they went through a couple of draft ‘recipes’ before they settled on the one they used. At 7.40 that night – two hours before the balloon was due to go up – Sergeant Shears began sending a normal Nicosoff message, noting that there seemed to be fewer troops on the streets and that General Alexander had cancelled his plans to travel to Iran.
Then, just before eight o’clock, he switched to Q codes, the unencrypted Morse shorthand used by radio operators across the globe to send immediate information. ‘WAIT, WAIT, WAIT,’ he signalled and went silent. A minute later, he signalled again: ‘WAIT FIVE MINUTES. WAIT FIVE MINUTES.’
Five minutes later he was back: ‘URGENT MESSAGE: WAIT TWENTY MINUTES PLEASE.’ He asked for more time once more before, at half past eight, sending an encrypted message reporting that his girlfriend had called him to say she had urgent information she couldn’t say over the phone. He was now heading to her flat, where she was entertaining an American officer. And then he went silent for an hour and a half.
At 10.17 p.m. he got back in touch. ‘HERE. URGENT MESSAGE. WAIT 15 MINUTES.’ He stalled twice more before, at 10.41 p.m., an hour after the offensive had started, sending his urgent news: ‘My friend learned tonight from an American air force officer at her place that a British offensive is imminent. Will probably start early tomorrow or even this evening. All leave cancelled yesterday.’
Shears was given instruction about how to send his messages as though he were ‘a man at fever heat of excitement with such news’. Nicosoff didn’t usually make coding mistakes, but on this night he would: ‘Errors can be made by switching letters round, e.g. EUKQB can be sent as EUKBQ, a mistake easily made by a man whose eyes are moving faster than his fingers.’
But when Shears sent the second message, ‘revealing’ Montgomery’s imminent assault, there was no response. He stayed on the air for half an hour, but there was nothing. Perhaps the operator at the other end had got bored and decided that whatever it was could wait. Perhaps, with British artillery blazing away along the front, they had been called away to handle other traffic.
