Hut 8’s ability to read Shark also confirmed that the B-Dienst had been reading the Naval Cypher No. 3, ensuring that communications security was improved and the Germans were unable to predict the convoys’ routes. Nearly a hundred U-Boats were sunk in the first five months of 1943. As the battle swung toward the Allies, Harry Hinsley and the Hut 4 intelligence reporters detected increasing signs of nervousness among the U-Boat commanders. They began to report torpedo failures, made exaggerated claims, and expressed widespread fear of Allied aircraft. By April, their morale appeared to have gone into terminal decline, with the Shark decrypts containing ‘increasingly frequent references to their fear of air attack and to the efficiency of the Allied surface escorts in following up aircraft sightings’.
By May, the Allied successes against the U-Boats had soared to a level that threatened to wipe them out completely. On May 23, after hearing of the loss of the forty-seventh U-Boat that month, Dönitz ordered the Wolf Packs to be withdrawn from the Atlantic, giving the Allies the respite they needed to get supplies across to Britain in preparation for the invasion of mainland Europe. Bletchley itself was not only involved in the Battle of the Atlantic, it was also heavily involved throughout 1942 in providing intelligence on the fighting in North Africa; playing an important role in operations in the Mediterranean against the Italian navy; providing information that was critical to the Double Cross System and preparing furiously for the invasion of mainland Europe. The 110 staff who arrived at Bletchley Park in August 1939 had swollen to 680 by the end of 1940 and by the end of 1942 to more than 3,500, of whom two-thirds were women.
So many of the people working at Bletchley Park were now women that Edward Travis set up a ‘Women’s Committee’ to advise him on ‘all questions affecting Women at the War Station’ and to ensure ‘the promotion of the well-being of all the women’ at Bletchley. The committee included representatives of all three women’s services plus Foreign Office civil servants, and its chairman, a Miss J. V. Wickham, was available at all times to offer advice and help to ‘any civilian woman who is in difficulty of any kind’.
CHAPTER 9
THE BATTLE FOR NORTH AFRICA
The fighting in North Africa had begun with a sweeping victory over the Italians in early 1941 by General Archibald Wavell, the Commander-in-Chief Middle East. No German troops were in North Africa as yet and the Allied victory was the result of what was virtually an unopposed advance by British and Commonwealth troops against extremely limited resistance. Wavell did receive some assistance from Bletchley from Italian messages decyphered in Josh Cooper’s Air Section, with one message leading to the destruction of twenty-five Italian aircraft in a single attack. Bletchley was determined from early on that it should play as large a role as possible in operations in North Africa, arguing that Ultra would ‘in all probability produce an increasing amount of information of vital operational value’.
The idea of an outpost of GC&CS for the Middle East based in Cairo had first been proposed in July 1938 by the pre-war MI6 Chief and GC&CS Director Hugh Sinclair, but was resisted vehemently by the service intelligence chiefs in Cairo. In early 1940, there was a reluctance at Bletchley to dissipate its resources by sending anyone to Egypt, but by the summer the situation had eased and Bletchley sent a small team of codebreakers out to Cairo in July 1940, reinforcing them a few weeks later with Major Freddie Jacobs, the head of the GC&CS Military Section’s Italian codebreaking team, to form a military cryptographic unit called 5 Intelligence School. A ‘combined inter-service cryptographic bureau’ with the cover title of Combined Bureau Middle East (CBME) was finally set up, in the former Fauna and Flora Museum at Heliopolis and with Jacobs in command, in November 1940.
Denniston arranged for Jacob to be in direct touch with Bletchley via Cuthbert Bowlby, the MI6 regional director in Cairo, for any discussion of cryptographic matters, and asked Jacob to send him a full round-up of what traffic was available there, noting that while the Bureau’s ostensible targets were Italian, Arabic and Russian ‘some day these may be expanded’.
Almost immediately after the Italians’ defeat, an Italian Air Force message referring to Luftwaffe escorts for convoys between Naples and Tripoli was decyphered at Bletchley Park. Hut 4 concluded from this that the convoys were carrying Germans and that Hitler must be sending troops to support his beaten ally. The codebreakers’ views were dismissed in both the Admiralty and the Air Ministry, and the report did not make its way to Cairo. A few days later, British troops had their first contact with the Afrika Korps.
General Erwin Rommel’s arrival in North Africa in mid-February 1941 led quickly to the installation of a direct Special Signals Link to Cairo using Type-X cypher machines and on 13 March, the first direct transmission of an Ultra report from Bletchley to the CBME. Hut 3 could now send their reports direct to Jacobs who could then brief Wavell. The link had not had time to establish itself before Rommel began a rapid offensive that was difficult to predict.
The available Enigma decrypts appeared contradictory. The Red Enigma suggested that Rommel had been told to build up his strength before launching an offensive. But the regional Luftwaffe system used by Fliegerführer Afrika, broken by Hut 6 at the end of February and designated Light Blue, pointed to an immediate advance. Ignoring his orders from Berlin and despite only having a limited force, Rommel attacked immediately and was soon pressing home his advantage against the poorly prepared British troops, taking the Libyan ports of el Agheila and Benghazi before surrounding the Australian garrison at Tobruk. But here he came to a standstill.
The problem for Freddie Jacob, the head of the GC&CS outpost in Cairo, was that only he and George Wallace, the commander of 5 Intelligence School, were allowed to handle the Hut 3 reports coming in from Bletchley, said Henry Dryden, who was originally sent out to Cairo from Bletchley to train codebreakers who had been working on Italian systems to break the German cyphers.
‘The first contact between British and German forward troops occurred at the end of February, and on 24 March el Agheila was reoccupied by the enemy,’ Dryden said. ‘At this point, I was invited by John Tiltman to go out to Heliopolis, “for a month, six weeks at the outside, old boy,” to train the Italian experts in breaking German systems.’ By the time he eventually got there, Jacob and Wallace were completely bogged down in trying to deal with the flood of Ultra messages that was coming in from Hut 3. ‘Because of their sensitivity, the messages were shielded from the eyes of the cypher officers by being encoded in a simple substitution before encypherment and dispatch from the United Kingdom,’ Dryden said. Jacob and Wallace ‘had been taking it in turns, on a twenty-four-hour basis, to decode them before they were sent in a locked box welded to the floor of a special car. As a high proportion of these messages reached Heliopolis during the night, these officers were glad of the 50 per cent reinforcement I provided.’
Given Rommel’s maverick disregard for his original orders, General Franz Halder, the Army Chief of Staff, sent his deputy, General Friedrich Paulus, to Tripoli to agree on a strategy, which was then passed back to Berlin, and to Bletchley Park, via the Red Enigma. The decrypts also disclosed that the failure to secure the port of Tobruk had stretched Rommel’s supply lines and left him desperately short of fuel.
Churchill pushed Wavell to take advantage of this position. But two attempted counter-offensives, forced on Wavell, against his better judgement, by Churchill, failed in the face of the German 88mm anti-aircraft guns, converted by Rommel into an anti-tank role, which had an unrestricted view of their targets across the desert terrain. A single Ultra message which ought to have led British intelligence to question why Rommel needed so many armour-piercing rounds for his anti-aircraft guns passed unnoticed on 18 May 1941. The link from Bletchley Park was at any event of little use during this initial period of the war in North Africa, partly because the Hut 3 reporters were still learning their trade, partly because they were relying on Luftwaffe messages sent using the Red and Light Blue Enigma systems, which w
ere the only relevant ones that Hut 6 could break, and partly because of the time it took to carry out the whole process of interception, decryption and the production of an intelligence report.
During fighting, the Bletchley Park reports rarely arrived in Cairo in what the military called ‘real time’. All too frequently the land battle had moved on before the intelligence could be given to commanders, said Ralph Bennett, one of the Hut 3 duty officers. ‘Very occasionally, the process could be completed in about three hours, but six hours may have been nearer the average.’
A Special Communications Unit (SCU) was sent out to Cairo in April 1941 to assist in the speedy transmission of Ultra intelligence to Auchinleck and his commanders in the field. It was the first of the SCUs to be sent on military operations, with many more to follow over the next four years. Based on the units first introduced during the Battle of France, it received Ultra reports direct from Hut 3, encyphered using the highly secure One-Time Pad system and sent via the RAF communications centre at Leighton Buzzard. It evolved into two separate organisations: the SCU conducting the communications between Hut 3 and the theatre of operations, and a separate Special Liaison Unit (SLU), providing the intelligence to commanders, ensuring that only those who were cleared to see it saw it, and that it was never used without there being an alternative plausible way in which the information might have been obtained.
Momentary tactical advantage is not sufficient ground for taking any risk or compromising the source. No action may be taken against specific sea or land targets revealed by Ultra unless appropriate air reconnaissance or other suitable camouflage measures have also been taken. If from any document which might fall into his hands, from any message he might intercept, from any word revealed by a prisoner of war, or from any ill-considered action taken upon the basis of such intelligence, the enemy were given cause to believe that his communications are not adequately safe-guarded against interception, he would effect changes which would deprive us of knowledge of his operations on all fronts.
The failure of the second counter-offensive, codenamed Battleaxe, led Churchill to transfer Wavell to India, replacing him with Claude Auchinleck, known as ‘the Auk’. This coincided with Ultra’s first great contribution to the campaign. The Luftwaffe provided air escorts for the Italian convoys resupplying Rommel’s forces across the Mediterranean and some detail of the convoys was therefore carried on the Light Blue cypher. But the information derived from these intercepts was rarely good enough to allow the Royal Navy or the RAF to take action against the convoys.
Then in July 1941, Hut 8 managed to break an Italian Navy machine cypher, the C38m, which provided a flood of detailed information about the convoys. The Light Blue cypher gave indications of when a convoy was going to cross the Mediterranean and what it would be carrying while the C38m provided details of vessels involved and the route.
‘Between them, the two were mutually complementary sources of news about the convoys’ routes and their estimated times of departure and arrival at designated ports,’ said Bennett. ‘By the late summer of 1941, so many supply ships had been sunk that the Axis operations were severely curtailed and indeed faced complete strangulation.’
Since collating and analysing this information was a job in itself and the information it produced was required by both the Royal Navy and the RAF, it was carried out initially through collaboration between the Hut 4 ‘Z watch’ and the Hut 3 research section now led by Lucas.
‘From now until May 1942, longer-term research had to yield place to the more exciting duties of handling operational signals on Axis convoys to Africa,’ Lucas said. ‘The essential part of the work lay in the identification of covernames, or covernumbers, for turning-points on the routes.’ The routes taken by the convoys would have on average half-a-dozen legs on the journey to North Africa. The length of each leg could be worked out from the speed and sailing times specified in the message. The difficulty was in locating the turning points.
Pins were stuck into a string at distances equal to the lengths of the legs on the map. The string had its ends pinned to the ports of arrival and departure. Any intermediate points already known were also pinned. The rest of the slack was shifted about by trial and error to give the various alternative possibilities until a general shape of route was obtained that made sense and corresponded with our experience of Italian naval habits, for example a respectful detour to the east or west of Malta. Life in the research section was never dull but nothing again ever quite equalled the excitement of angling for Axis convoys with pins and string.
The convoy reports went to Cairo via the Special Signals Link and also to the navy in Alexandria and the RAF in Malta. The protection of the Ultra secret was paramount. In accordance with the regulations, no offensive action could be taken unless there was a clear secondary source, overwhelmingly created by aerial reconnaissance. Even this could not be directed solely against the convoy lest the Germans noticed the change in routine reconnaissance patterns. But the material supplied by Bletchley Park allowed the Royal Navy and the RAF to wreak havoc among Rommel’s supply convoys.
‘Ultra was very important in cutting Rommel’s supplies,’ said Jim Rose, one of the Hut 3 air advisers.
He was fighting with one hand behind his back because we were getting information about all the convoys from Italy. The RAF were not allowed to attack them unless they sent out reconnaissance and if there was fog of course they couldn’t attack them because it would have jeopardised the security of Ultra, but in fact most of them were attacked.
By the time Auchinleck launched Crusader, a successful counter-offensive which relieved Tobruk in November 1941, the RAF and the Royal Navy were regularly sinking Rommel’s supply ships, causing him major problems. The arrival in Malta in late October of Force K, comprising the cruisers HMS Aurora and Penelope together with two destroyers, heralded a two-month period when supplies to the Afrika Korps were brought to a virtual standstill.
The Luftwaffe keys had revealed Rommel’s own plans for an attack on Tobruk. But the first break into Army Enigma in North Africa confirmed that the failure of supply ships to get through made it unlikely that any attack would take place in the near future. The proliferation of Enigma keys as the Germans sought to improve signals security had led to a change in the nicknames given to them by Bletchley Park. The main keys that had already been broken retained their colour designation. Luftwaffe keys were now named after flowers or insects, Army keys after birds and Naval keys after fish.
Chaffinch, as the new Army key was called, was broken after the capture from the headquarters of 16th Panzer Division on 28 November 1941 of the complete keys for November. It provided details of the Afrika Korps’ shortages of food, fuel and water as well as ammunition for the 88mm ‘anti-tank’ guns. It also gave Auchinleck full details of how many tanks Rommel had at his disposal and useful information on the German dispositions. The captured documents also assisted in the break into a second Afrika Korps Enigma, Phoenix, albeit for only a week. These were the first operational German Army keys to be broken in a British Army area of operations and the Hut 6 successes against them were to be the first step of a major turning point in the way Ultra was viewed by British Army commanders. Up until that point, all the operational intelligence from Enigma had come either from Luftwaffe or German Navy keys, and while the former frequently provided important intelligence for ground forces commanders, it was not regarded as highly by the British Army as it was by the RAF.
Despite the presence of the SLU at the headquarters of the British Eighth Army, the Crusader offensive proved beyond doubt that the best use of Ultra was in providing the details of enemy strength and dispositions, and often future plans, rather than in tactical information during the heat of the battle. The sheer length of time it took for reports to get from Bletchley to Libya meant that Hut 3 could not compete with the mobile Y Special Wireless Section of Royal Signals and Intelligence Corps personnel in armoured cars who were at the front.
&nbs
p; ‘Despite the amazing speed with which we received Ultra, it was of course usually out of date,’ recalled Bill Williams, who served as an intelligence officer in North Africa.
This did not mean that we were not glad of its arrival for at best it showed that we were wrong, usually it enabled us to tidy up loose ends, and at worst we tumbled into bed with a smug confirmation. In a planning period between battles its value was more obvious and one had too the opportunity to study it in relation to context so much better than during a fast-moving battle such as desert warfare produced.
Auchinleck’s defeat of Rommel forced the Afrika Korps back to el Agheila. But within a few months, the Germans had regained much of the ground they had lost and were back in Benghazi. This was at least in part the result of a serious misreading of a decrypt from the Italian C38m cypher which was wrongly seen as suggesting that the Afrika Korps did not expect to reach Benghazi. But it was also a result of the failings of the British tactics. Bennett said,
The troops on the ground on our side were still not used to receiving high-level information. They were also using the wrong tactics. The gunfire was never sufficiently massed to do enough damage to the enemy until Auchinleck managed to change it under pressure in the late summer of 1942.
Although the Red Luftwaffe Enigma gave good warning of a German offensive scheduled for the end of May of 1942 and aimed ultimately at regaining Tobruk, Hut 3 was unable to provide any information about Rommel’s precise plans, either before or during the Battle of Gazala. By mid-June, Auchinleck had decided to withdraw across the Egyptian border to a stronger position at Alam Halfa, leaving the 2nd South African Division to hold Tobruk as a fortress inside the enemy camp.
The Secrets of Station X Page 20