From our archives...

August 1943 : The Germans evacuate Sicily

Determined German resistance slowed the advance of the British forces along the southern coast, aiming towards Catania and then Messina. Montgomery attempted to get round Mount Etna to the north but was frustrated after some very heavy fighting with the Germans, now reinforced by the 1st Parachute Division. Meanwhile Patton’s troops had made a rapid advance to the north, taken Palermo and turned to advance eastwards along the north coast. From 13th July decrypt Intelligence had been plentiful, notably from the Red and Locust keys of the Luftwaffe, from the Albatross army Enigma key, and the Bream (Fish) traffic, and from the newly broken Puffin Enigma key. The German Porpoise naval Enigma key and the Italian C38m naval key added shipping Intelligence. However local Army Y units proved unable to perform as well as they had in the later stages of the North African campaigns because of greatly improved German field communications security. The BP Intelligence went direct to Montgomery and Patton once they were established in Sicily, of course via their attached Signals Liaison Units. From 15th July a number of decrypts made it clear that the Germans were evacuating the Western part of the island. The fall of Mussolini on 26th July and the Allied pressure, not least from the sea and air, causes the overall German commander in Italy, General Kesselring, to recommend evacuation, and finally Hitler agrees on 5th August. Patton uses a series of small amphibious operations along the north coast to speed his progress, though at least one of them lands after the US land force has already passed by. The one move the very considerable Italian naval forces make to impede the operation comes on the nights of 6th and 7th August when two Italian cruisers set out from Taranto to bombard Palermo. However, forewarned by Enigma and C38m decrypts, destroyers and aircraft ensure they turn back on both occasions. During the night of 13th/14th August the Germans disengage everywhere, and complete their withdrawal over the narrow straits to the Italian mainland by the morning of 16th. Late that day a patrol from the US 3rd Division, to the fury of Montgomery, enters Messina ahead of an amphibious party from the Eighth Army who arrive there early the following morning. (General Montgomery gets his revenge by ensuring that bagpipes from the Highland division are there to greet General Patton’s ceremonial entry later that morning). The German fighting withdrawal has been impressive, as always, with the deciding factor in the campaign being Allied air domination.

In a strange echo of the German tanks halting before Dunkirk, the Allied forces, despite their dominance of the air and sea, have done virtually nothing to impede the evacuation, in which the Germans transport some 55,000 men, 9,000 vehicles, and all their heavy vehicles back to the mainland without casualties. The Italians brought away 62,000 troops. There had been an ambiguously worded decrypt of 18th July which suggested an evacuation was being planned, and then on 1st August clear indications that ferrying in the Straits of Messina is being planned. On 6th August BP warns that the Hermann Göring division is intending to evacuate in the near future and on the 8th BP gives further warnings that evacuation is imminent. On the evening of 11th August the Eighth Army’s own Intelligence summary confirms that evacuation is in full swing. Thereafter both decrypts from BP and photo-recce record the progress of the operation. Yet virtually nothing was done to prevent the flow across the, admittedly short and well defended, sea crossing from Messina to the toe of Italy. There seems to be no official explanation for what surely was a considerable missed opportunity?

The Bombing of Peenemünde. On the night of 17th/18th August 1943, 596 RAF bombers attack the German V weapons development site at Peenemünde on the Baltic, dropping 1,800 tons of bombs. 40 RAF bombers fail to return, despite the use of Window over Berlin in a successful attempt to mislead the night fighters. As usual, considerable pressure had to be applied before Bomber Command would divert from their campaign to destroy the German industrial cities. This raid was different in that the attack, from the relatively low level of 7,000 feet, was intended to be not on an area but on a collection of precise targets selected on the basis of good photo-recce information. It proved to be one of the most successful of all RAF raids, the official post-war verdict being that it delayed the V weapons programme by at least two months. 180 German workers were killed including some of the scientists and engineers leading the development programme. Unfortunately in a tragic mistake, the Pathfinders also marked the near-by camp housing foreign forced-labourers, causing many casualties.

The raid led to the dispersal of the V weapon manufacturing into underground factories, and the moving of the test range to Heidelager in eastern Poland. The first British Intelligence about the V-weapons work at Peenemünde had come in August 1940 from a source in contact with Admiral Canaris, the head of the Abwehr, confirming the statement in the ‘Oslo’ report. But it was not until the spring of 1943 that stories from Prisoners of War and agents began to multiply. The British attempt to deduce what was being developed became confused because of information on the several quite different weapons, the jet fighter, the V1 flying bomb, and the V2 rocket. The picture was further confused by reports that put the war-head of the rocket at 10 tons, a weight that thoroughly alarmed Churchill and the Chiefs of Staff and eventually turned out to be about ten times too large. They might have been even further alarmed had they known that on 8th July 1943 Hitler had given the V2 the highest priority, setting a date of September 1943 for the commencement of the bombardment. One of the best of the many reports that now came in was a Purple decrypt from the Japanese Ambassador to Berlin which gave London and the surrounding manufacturing towns as the target for the V weapon campaign. The picture in the UK was further confused because the rocket experts at the RAE, Farnborough, were convinced that the rocket must be solid fuelled, though many reports were now coming in of a new liquid fuel. At BP Prof. Frederick Norman in Hut 3 has special responsibility for looking out for any decrypt material relating to the V weapons. At the end of October 1943 BP began to produce good evidence that the Germans were developing flying bombs.

US Bombes Enter Service. In October 1942 Cdr. Edward Travis, Director of BP and Cdr. Joseph Wenger, the US Navy head of Op-20-G, the US Navy centre devoted to codebreaking, signed an agreement under which BP and OP-20-G were to cooperate on breaking German naval cyphers. As a first step BP was to help the US Navy develop ‘analytical machinery’, the US name for equipment of the bombe type. BP sent them full details and drawings of the British bombe, and a paper by Alan Turing on their use. Alan visited the US to see the prototype at NCR during December 1942 and the US ordered 95 production machines in March 1943. The two prototypes, Adam and Eve, began tests in May 1943 and production machines start to be shipped to OP-20-G in Washington in late August, with 6 in ‘semi-continuous’ operation by 7th September. They are basically similar to the UK model, but employ valves as well as relays. They are high-speed four rotor machines, employing a considerable number of valves (largely thyratrons) for the high speed switching circuits, as relays are considered unreliable at these high speeds. Instead of needing plugs and cables to set up the menu like the British machines, the US version of the Turing bombe has uniselector switches that can be rotated to a desired letter of the alphabet. It has an electronic memory to record a stop, so that the machine can overrun, slow down, and then return to the stop position, instead of having to stop immediately that a stop is detected like the standard British bombes. (The British High Speed bombe, Mammoth, has a device that enables a stop to be investigated without stopping the bombe at all). The first type, of which some 95 were built in total, had only 16 banks of rotors compared with the British 36, and so could only run short cribs. In theory they could run a three-rotor menu in 50 seconds, a four-rotor menu in 20 minutes, but in practice it took considerably longer allowing for set-up time. At this time in August 1943 BP had about 13 British four-rotor bombes (11 Keen or BTM type called Mammoth, 2 Wynne-Williams or Mawdsley called Cobra), so the US bombes were a very welcome addition to the effort to break the U-boats’ 4-rotor Enigma keys. The US bombes were manned by Waves, the US equivalent of Wrens, and used cribs sent to them by secure teleprinter using Type X. Once they were through their unreliable teething stage, the rate of production of US bombes was phenomenal, by British standards. As the number of US bombes rose rapidly, reaching 75 by January 1944, spare capacity became available, at a time when BP had some 100 British bombes in service. But 70 of these British machines were the slower 3-rotor bombes, and the high-speed US bombes could do three times the work of these slower British bombes on standard German 3-rotor Enigma messages. So Hut 6 was able to take advantage of this, once the reluctance of the US Navy to take on work for the Army had been overcome. Such Army and Air Force runs on US bombes were known as ‘Bovrils’; in 1944 these would take over half the US bombe time. As confidence in the Op-20-G team built up, BP was able to off-load ever more naval work to the US team, releasing effort in Hut 8 to concentrate on breaking new cyphers and on the complications that the Germans introduced from mid-1944 onwards. Relations between Hut 8 and their US opposite numbers became excellent.

BP Breaks in August. There was no respite for BP this August. For the first time Hut 8 broke Sunfish, the ‘blockade-runners’ key used by German supply ships and U-boats working to the Far East. It was only broken intermittently, but was valuable for providing information that enabled the interruption of this supply route by which the Germans were sending some of their technological advances to aid the Japanese. The blockade-running season began in earnest in November 1943 and ended in January 1944 when the Axis abandoned blockade-running. In this period, largely due to decrypts of Shark messages to the U-boats and Sunfish decrypts of messages to the surface ships, no German ship succeeded in getting away from the French Atlantic ports without being sunk in the Biscay area, and only one in-bound succeeded in reaching France. Hut 6 broke Puma, an army-air cooperation key used in Sicily; this key became very valuable in November when it became the key of the Luftwaffe liaison officers at the German 10th Army HQ in Italy. The almost daily decryption of the liaison officers reviews of events on their various division fronts made it possible to see the battlefield through German eyes. Hut 6 also broke Peregrine, used by a German SS Panzer Corps in Yugoslavia where they were waging a bitter struggle with the partisans. They also broke Shrike, apparently used by advanced elements of the German XXXIX Panzer Corps arriving in late August in northern Italy. It had become apparent by October 1943 that the Germans were using this key to exaggerate their actual strength in Italy, for the Corps was still on the Russian front. This at first caused a panic at BP because this use of Enigma to deceive suggested the Germans were aware of the success of BP in reading Enigma; however reflection soon established that this deception was to impress their Italian ‘partners’ with their strength in Italy rather than the Allies who the Germans assumed could not read the traffic. This was the first time, but not the last, that BP found the Germans using Enigma signals as part of elaborate deception plans devised to conceal their intentions or strength from their own forces. Examples arose in Greece and Yugoslavia before the German invasion of Hungary in March 1944, in the Ardennes offensive in December 1944, and in the attack in south Hungary in March 1945.

Moves at BP. Accommodation was always in drastically short supply at BP. The very large Block F, at 40,000 sq ft the same size as Block D, had been building since January 1943 and now it was finished, enabling the Fish teams to get together there at last in September. The Testery who were tackling the Fish signals by manual methods had been working in Huts 4, 11 and then in 15A, where their 5 Tunny machines had been transferred in January 1943. They moved into Block F, as did the rest of the large Military Section. The Newmanry, who worked in Hut 11, were trying to get the counting machine, Heath-Robinson, to work for long enough to solve part of the Fish settings, as since June they knew from experience that this could be achieved. The machine problem was due to the unreliability of the electronic valves, and the difficulty of keeping the two tapes in synchronisation. The prototype Colossus is now about to be assembled at Dollis Hill, and will overcome these problems. It is at this time that Max Newman, Tommy Flowers, the brilliant Post Office design engineer, and Alan Turing are seen deep in discussion. The Newmanry is now 25 strong, a number that will grow to about 350 staff in due course of whom 200 were Wrens. At this time the Testery has 40 staff, who are very fully engaged in breaking Fish links with a little help from the Newmanry, in particular Bream from the German High Command near Berlin to Rome, very valuable though out the Italian campaign. They had just broken the Perch link to Army Group Centre, engaged in the new Russian offensive. The bombe complex at Stanmore opened in July and now houses some 500 Wrens and their bombes. Despite this, the number of Wrens working in BP continues to grow so the ‘Wrennery’ at Woburn Abbey opens in August with 6 Wrens; the number there was to grow to 400 by 1945. Stuart Milner-Barry, who leads the cribsters in the Hut 6, is soon to be put in charge of the Hut. He will take the place of Gordon Welchman in September when Gordon is promoted to be Assistant-Director (Mechanisation) with responsibility for matters relating to machine aids in Huts 6, 8, the bombes under Sq. Ld. Jones, the Newmanry, and the Tabulating Section in Block C under Freddie Freeborn. These teams all make use of considerable equipment of the bombe, punched card or electronic machine type.

The Bletchley Park Trust welcomes the preparation of these notes, but the authors are responsible for the statements and the views expressed.

Can you help?

If you have any information relating to the following question please contact us at info@bletchleypark.org.uk

'In August 1943 Bletchley Park broke the German Army key Shrike. It soon became clear that this was the key used by elements of the German XXXIX Corps, apparently being deployed in Northern Italy as the Corps moved there as the Italians surrendered. But BP knew that this Corps was in action on the Russian front, and it soon became clear that the key was being used to deceive the readers of the Enigma messages into believing that the Corps was arriving in Italy. It is known that at first this discovery caused a panic at BP as it suggested that the Germans must know that the British were reading their Enigma messages. Does anyone remember this time and the panic that this caused? How long was it before the strange explanation that the Germans were trying to persuade their erstwhile allies, the Italians, who had been given the key to Shrike, that they were arriving to prevent the surrender was accepted at BP? Does anyone remember the other occasions when this use of Enigma happened? Did it cause a similar panic when these messages were read?

Read other example archive documents from Bletchley Park