Seventy Years Ago This Month at Bletchley Park

February 1940

The War. 

With no land fighting, except in Finland where the Finns are falling back in the face of the renewed and much better conducted Russian offensive, the war has become fought primarily at sea. There is little air activity, though the RAF mounted a six day mission starting on 25th February, their biggest to date. Berlin and other cities were attacked – with leaflets. This month a specially modified long-range Spitfire obtains the first good aerial photographs of the German naval bases at Emden and Wilhelmshaven. But the RAF had to wait until April before an even longer-range Spitfire was available to cover Kiel, but still not the other German bases further afield in the Baltic.

The War in Finland.

The Russians launch a renewed massive offensive, now led by Marshal S. Timoshenko, against the Finnish Mannerheim line on 12th February.  For three days the Finns hold out but the line is finally breached on the 15th. The Finns fall back to their second line but on 28th February this too is assaulted by overwhelming numbers.  On 24th February the Russians had launched an outflanking assault across the frozen Baltic close to Helsinki. The Finns now know they are defeated, and ask for an armistice, effective on the 13th March.  The British still believe that the Russians have shown themselves incapable of winning outright; the much improved quality of their renewed offensive, like their brilliant 1939 campaign in Manchuria (led by an unknown commander, G. K. Zhukov), went hardly noticed in the West.

Plans to Invade Scandinavia. 

One impact of the fighting in Finland is that both the Allies and the Axis planners focus attention on Scandinavia. The Allies know that Hitler wishes to acquire bases in Norway to protect his iron-ore supplies coming by coastal waters from the north of Sweden.  The British and French Prime Ministers agree on 5th February to mount an expedition to help the Finns.  This force would land on the coast of Norway on 20th March “to be sure of forestalling the Germans” according to Neville Chamberlain. In passing, it would take control of the iron-ore mines in Northern Sweden. From their penetration of the British naval codes the Germans become aware of the Allied plans, and so Hitler orders that urgent preparations be made for a counter-move, an invasion of Norway and Denmark.  The Allied plans to go to the help of the Finns are abandoned on 4th March, though alternative British plans for landing on the Norwegian coast are being developed. Both sides plan to ignore the stated neutrality of Norway, Sweden, and Denmark.

The Altmark Rescue. 

On 16th February, the Royal Navy violates Norwegian neutrality to board a German ship in coastal waters on route to the Baltic.  The Altmark had been one of the ill-fated Graf Spee’s supply ships.  299 British sailors who had been taken prisoner in the South Atlantic are released to great acclaim in the UK.  (Information on the passage of the Altmark had been obtained from the French network of agents along the Norwegian coast.  These observers were a double-edged sword, for the Germans allowed some of their ships to be seen when they wanted to divert Royal Navy effort from more important matters).  The Altmark incident reinforced Hitler’s decision to seize the Norwegian bases, before his attack on France.  The British Air Intelligence branch was still advising that Hitler’s spring offensive would be limited to seizing Holland – to provide a base for an all-out attack on Britain.

The U-33 Sinking. 

On the early morning of 11th February the minesweeper HMS Gleaner, detects the submarine U-33 as she is about to lay mines in the entrance to the Firth of Clyde. Depth charges bring her to the surface, and the Gleaner opens fire and sinks her.  Three of the submarine’s Enigma code-wheels are captured, perhaps from the pockets of a German seaman. Bletchley Park finds that two of these are hitherto unknown wheels, for the German navy is using a choice of 8 code-wheels for the three slots in the Enigma machine, instead of just the 5 employed at that time by the other German services. (That there is a third new code-wheel was not detected until it was found in a naval capture of August 1940).  The capture of U-33 certainly helped BP towards the breaking of the German Enigma naval keys, but it would not be until August 1941 that BP could break a German naval key regularly, though before then they were to read the odd one on occasions by various captures of key sheets. The fact that the German navy are using a choice of 8 wheels rather than 5 adds very considerably to the number of possible key settings, but another problem is that the naval Enigma operators cypher discipline is far better than that of their Luftwaffe’s comrades.  So they tend to make fewer of those careless mistakes that helped BP to break into the other Services keys.  Of course, the more messages that are read on any key, the easier it becomes to break, as the habits of those operators and the nature of the messages the key handles accumulates.

The Early Breaks into Enigma. 

Using the Netz sheets prepared at Bletchley Park by John Jeffreys, based on the Polish Zygalski method, the Poles have broken the first wartime Enigma message on 17th January 1940.  BP followed shortly thereafter using the same sheet method, for on about 22nd January the Research Team made BP’s first break into German Enigma, being the Green (Army administration) of the 25th October 1939.  Two others breaks were made within a couple of days, probably the Red (Air Force general purpose operational) for 17th January 1940, and then the Red for 6th January.  The Germans had not modified the Enigma machine or the procedures, so BP is in serious business at last!  Then on 29 January BP broke the Blue, which turned out to be the Luftwaffe training key.  By the end of March some 50 key settings were solved, though generally with considerable delay, using the Netz sheet method aided by the detection of ‘cillies’. According to the Poles, BP and the Poles broke some 126 keys in total from January to mid-June, 83 % of these were being solved at BP.  The two teams were exchanging material over a line between Bletchley and the Château de Vignolles at Gretz-Armainvillers 30 kms to the east of Paris, until the Poles were forced to evacuate on 9th June. (The Polish accounts say that Enigma machines were used on this line to preserve security!)  Most of the early breaks were of the not-very-useful Blue, though the Red had been made the top priority. 

The Cillies. 

Dilly Know and his team in Cottage 3 had long been studying the properties of the Enigma message indicators from an early set of intercepted encrypted messages. It became clear that the German cypher clerks were not choosing their message indicator at random as they should have done. They tended to have ‘bad habits’ that could be used to assist in code-breaking, in particular by reducing the number of wheel-orders that had to be tried by the Netz sheet-stacking method (or other methods such as, in due course, the Turing bombe). It was soon clear that it was unlikely for a message setting indicator to contain a repeated letter. The German signallers tended to use a set of three letters that were pronounceable, and so the middle letter was most likely to be a vowel. Once decryptions became available, it was apparent that the German cypher clerks had other bad habits; one important one was that, to avoid resetting their machine code-wheels, they sometimes used the position of the machine wheels at the end of one message as the ground-state indicator for the next message. This was common for messages that had more than one part; as each message was limited to 250 characters for security reasons, it was quite common for messages to have more than one part. It now became apparent that it was easy to spot the various parts of a multi-part message, and that they tended to use the last setting of a part for the ground state of the next part. (Was this an authorised practice? Possibly, in the interest of speeding up the transmission of long messages). Of course if the setting of the last wheel-position was known, it was easy to count back the characters in the message to determine the start position. Then there were the easy-to-type keyboard sets, like RFV, CGU, etc. And then there were first three letters of a favourite pronounceable name, such as MARthe, WALter , HITler, and CILlie – did the lady ever learn that her name had given rise to Hut 6’s collective term for these helpful habits: the ‘Cillies? It was not unusual for the first three letters of a word to be used for the ground state, the next three for the message start. These ‘gifts’ did not solve the message decryption alone; the codebreakers had to determine which code-wheels were in  each of the three slots, what the wheel ‘ringstelum’ setting was; and what the ‘stecker’ connections were. But the poor procedures of the German operators did much reduced the work it took, and this went on until mid-1944

The Value of the Early Breaks. 

Establishing a key setting enabled all the messages intercepted in that setting to be decrypted just as a German cypher clerk would do, typically for that day until the key setting changed at midnight. So BP is now reading hundreds of German messages.  Despite the great news, for security reasons an immediate blackout was imposed so that few people in BP or Whitehall knew what had happened. Because the war was static at the time of the first breaks, the Intelligence information obtained was not of great value, but the potential for when the war went mobile is – or should have been - now clear! And indeed, though the Germans achieved total surprise in their invasion of Norway on 8th April, two days later BP broke the German campaign key, the Yellow, and thereafter read it continuously and almost currently until it lapsed in July, after the end of the Norwegian campaign. When Germany changed the setting up procedures for other keys on 1st May, the Yellow retained the old procedures, presumably because it was too difficult to change in the middle of a campaign. So the Netz sheet method could continue to be used for this key alone, and the large increase of radio traffic, due to the military campaign, proved a great help to the cryptographers.  Unfortunately after the first breaks in January only the Navy had made special arrangements for feeding the highly-secure Intelligence to their forces.  But both the Red and Yellow had little information of direct value for the Fleet and the first breaks into German naval Enigma keys were not achieved until March 1941.  (The Germans were reading our navy and merchant navy codes at this time). So until ways were devised for handling this valuable information, it has to be recorded that this success of BP had little, if any, direct impact on the conduct of the fighting in Norway, and indeed for many months to come.

The Staffing of the Enigma Handling Huts. 

It had been decided to establish a separate team from the Research Section for the exploitation of the anticipated break into Enigma. This led to the building up during January 1940 of an “exploitation” team for Enigma Army & Air Force cryptography work in Hut 6. Probably responsibility for the team in Hut 6 was nominally still vested in Dilly Knox. John Jeffreys looked after the Sheet-stacking Room where the key settings were found, and the Machine Room where the plain text of the intercepted messages was recovered by stripping out the cypher once the key settings for the day were found. Gordon Welchman took responsibility for the rest. In total some 30 people were then employed in Hut 6.  The build-up of the team in Hut 3 soon followed, dealing with the interpretation and Intelligence extraction from the decrypted messages, and the preparation of the material to be sent to Whitehall. Hut 3 was led by Lieut-Commander Malcolm Saunders, from the Naval Section; with Sq. Ldr Shaw from the S.I.S.; Capt. Edgar from the Military Section; and F. L. Lucas, from Kings, Cambridge. The first Hut 3 was a small building on what is now the Mansion car park, close to the tennis courts. The Military Section was pushed out to Hut 5, close to where Hut 4 still stands. In the autumn, as the team grew to tackle the fast growing number of decrypts, Hut 3 moved to its current location alongside Hut 6. Both the Hut 3 and 6 teams moved in Block D in February 1943.

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