by Alastair Denniston- Code-breaking From Room 40 to Berkeley Street
It is unlikely that when AGD accepted Ewing’s offer of some work without pay during his vacation, he could have imagined that it would set him on his career path for the next thirty years. However, events in continental Europe would soon put him on that path.
Chapter 2
British Signit in World War One
On 28 June 1914, the Archduke Franz Ferdinand, heir to the throne of Austria-Hungary was visiting Sarajevo on the final day of his trip to Bosnia. Early in the morning his motorcade was attacked and a Bosnian Serb named Gavrilo Princip shot and killed the Archduke and his wife. Immediately following the assassination, Britain and Germany remained united until Germany backed Austro-Hungarian action against Serbia. On 26 July, Britain proposed a peace conference to try to avert war in the Balkans. However, by 3 August, Germany had declared war on Russia and France, and when Belgium refused to allow German soldiers to cross its borders, the die was cast. The following day, Britain declared war on Germany.
When war broke out, the Admiralty wireless intercept (also known as Y) stations intercepted a number of German messages and passed them to the Admiralty. They were programme messages sent out by the German high power W/T station, Norddeich, and consisted mainly of weather and intelligence reports. These were the messages that Oliver had shown to Ewing on 4 August. Early in September, Russell Clarke, a former student of Ewing, told him that he and his friend Colonel Hippiseley, a wealthy landowner, had been intercepting German messages on their own receiving sets in London and Wales. Clarke was a prominent amateur W/T expert and Hippiseley was an amateur mechanic with an interest in W/T. It is remarkable that the police had not confiscated their equipment, but Clarke was a barrister and well connected. They told Ewing that there was a considerable amount of German naval W/T signalling going on at a lower wavelength than that being used for the Norddeich traffic. They offered to provide the Admiralty with enough German signals for it to study the movements and plans of the German High Seas Fleet. To do this, they only required access to the appropriate facilities, which Ewing provided, and permission was granted for them to install their equipment at Hunstanton Coast Guard Station. This was a sensible choice as it was already equipped with W/T and was ideal for interception in Flanders and Northern France. They involved another amateur in the work, Leslie Lambert, and the new station was able to keep a continuous watch on German field stations. As they grew more successful, additional radio masts were added. In all the intercept stations, each aerial had an assigned range of wave length to monitor and it was arranged to take down every message in two separate locations. Each was connected to the Admiralty by telephone line to a telegraph office in the basement. From there, messages went by pneumatic tube to Ewing’s section, where they fell into a waiting basket.
By the end of the war, Clarke presided over four stations at Hunstanton, and one at Slough, Ashford, Lizard and Ballybunion which were manned by General Post Office (GPO) operators as well as Lambert. Traffic was worked on in the Admiralty and War Office during the day and by the night watch in the War Office overnight. Additional intercept facilities were provided by the Marconi Company using its own operators. In all, over 200 wireless stations were in use in England during WW1.1
Within the first four months of the war, the Admiralty was in possession of all three principal code books of the German Navy. The first was captured from the German-Australian steamship Hobart, which was seized off Port Philips Heads, Melbourne. The haul included the Handelsverkehrsbuch (HVB), the code used by the Admiralstab (German Admiralty) as well as the High Seas Fleet. It reached the Admiralty by the end of October 1914 and remained in use until March 1916. On 26 August, the German light cruiser Magdeburg ran aground in thick fog on the Island of Odensholm off the coast of Russian Estonia in the Gulf of Finland. Before confidential papers could be removed and the ship blown up, two Russian cruisers attacked her as the fog lifted. The Russians seized the ship and captured two copies of the Signalbuch der Kaiserlichen Marine (SKM), the current key, a copy of the German naval gridded chart for the Baltic, the bridge log and the ship’s War Diary. According to Winston Churchill, the First Lord of the Admiralty at the time, the SKM code book was found in the arms of a German signalman whose body washed ashore a few days after the Magdeburg ran aground.2 The Russian Naval Attaché met with Churchill on 6 September. The Russian Admiralty had successfully used the cipher and signal books but felt that the Royal Navy, as the leading naval power, ought to have them, along with the charts.3 Churchill ordered a ship to be sent to Alexandrov to transport two officers of the Imperial Russian Navy to Britain. He duly received them in person on 13 October at Scapa Flow. The Naval Attaché, Captain Wolcoff, delivered the code book to the Admiralty on 17 October. The book only worked on unimportant messages because the rest used a key along with the code book. It continued to be used by the Germans until May 1917. Its code consisted of groups of letters, usually three to a group. The usual process of keying consisted of substituting each letter for another letter based on a prescribed plan.
At the end of 1914, the ‘tool set’4 of Ewing’s section was enhanced when an English trawler, fishing in the North Sea, hauled up a collection of German books and documents. They were forwarded to DID in the Admiralty and included a naval signal book, not previously available, as well as confidential papers and charts. They came from a German destroyer which had been sunk in the North Sea off the Dutch coast on 17 October 1914. Among this treasure trove was a most secret code book, the Verkehrsbuch (VB). On 3 December 1914, Ewing was given a parcel of books which had arrived from Lowestoft. The Germans applied keys to the VB so these still had to be found. This was achieved later the same day while Ewing dried the code book by his fire! Some days earlier, two versions of a message had been sent out independently from separate German stations. One was encrypted completely using the VB book and easily read. The second was identical, except for a few words which had been encrypted using another code. Comparing the two messages allowed Ewing’s section to discover the key to the new book. This would be known at GC&CS years later as a ‘kiss’. According to AGD: ‘The VB was found to be of the greatest immediate value in dealing with the German cruiser fleet, while the fact that it was solely used for the correspondence with the Naval Attachés abroad, especially in Madrid, escaped notice for some months.’5
By the middle of October, a naval officer, unknown to Ewing’s team, had taken over the office of Ewing’s secretary, Mountstephen. He was Fleet Paymaster Rotter at the time, the head and principle German expert of the Intelligence Division. Rotter had spent four or five years prior to the war trying to persuade the Admiralty to take cryptanalysis seriously but had failed. He had been placed at Ewing’s disposal and was investigating the assertion of the Russian naval attaché that the SKM book was the one now in force in the German Navy and that any naval intercepts in the Admiralty’s possession could be decrypted with it. Within a week, Rotter had identified the German procedure of numbering messages and the re-encrypting of the numbers, and this provided him with a good starting point. The weather reports came from the book and were then subjected to a simple substitution cipher. The messages were a numbered series sent out by Norddeich (K.A.V.) to all ships (A.S.). The Germans had re-encrypted the numbers of the messages, which provided a simple and certain way into their reencryption tables. Clarke claimed that he could intercept hundreds of such messages on short wave, which would provide the daily movements of the German fleet. Rotter eventually produced the key to the captured book and with Clarke proving true to his word, the Hunstanton traffic flowed into the Admiralty and on to Ewing’s team.
It was now clear that the Admiralty cryptographic section under Ewing had found its niche and expansion was necessary. The section was formally known as NID 25, being part of Naval Intelligence Division (NID). AGD, Herschell and Norton returned full-time to Ewing’s team and Rotter was formally transferred to the section. In early November, additional office space in the Admiralty was obtained and the section moved in
to a 24ft by 17ft room on the first floor of the Old Building of the Admiralty. The number on the door of their new accommodation was 40 and gradually the section became known as Room 40 OB or just Room 40 and ultimately would be considered by many historians to be the birthplace of modern British Sigint.6 By 6 November, Rotter, Henderson, Curtis and Parish were working as day men and AGD, Herschell & Norton as watchkeepers. Room 40 was expanded to around fifty staff, and new recruits included Claud Serocold, a stockbroker; Lord Monk Bretton, a former private secretary to the Liberal politician Joseph Chamberlain; H.A. Morrah, former President of the Oxford Union; W.H. Antsie, a Dartmouth schoolmaster; H.W. Lawrence, an expert on furniture and art; and Commander Fremantle, a member of a famous naval family.7 Of two others, Hooper and Bond, little is known.
Ewing and his counterpart in the War Office had continued to cooperate in attacking the military ciphers but once Room 40 started to solve naval ciphers, Ewing could no longer help the War Office team. Ewing also took on responsibility for diplomatic traffic. In the early days of the war he visited the Eastern Telegraph Company to discuss monitoring German wireless stations in Africa which were receiving signals from the Nauen transmitter. They provided old material sent on this link. The main concern of the Admiralty was to solve the enemy’s naval cipher used to communicate information and orders by wireless from ship to ship or from shore stations, often conveying the orders of the Commander-in-Chief of the German Admiralty. The number of intercepts increased thanks to the work of the Post Office, the Marconi stations, the single Admiralty ‘police’ station at Stockton and Clarke’s Hunstanton operation. Hunstanton and Stockton became the core of Britain’s Y Service and along with the Post Office and Marconi intercept stations, eventually recorded almost every German naval, diplomatic, commercial and consular wireless message transmitted.
Secrecy was paramount and apart from Ewing, his staff and a few responsible officials, there was no knowledge of Room 40’s work, even amongst the admirals at sea. To quote Ewing:
Within the small circle of initiated persons the mysterious agency thus set up was called ‘Room 40’. The origin of the same was that when the staff of deciphers first overflowed the tiny limits of my private room, a large room numbered 40 in the old building of the Admiralty was assigned to us. Later more spacious quarters and room after room was added, but the original title stuck – a title that suggested nothing and stirred nobody’s curiosity. Lord Balfour, after he took office as First Lord, sometimes used another piece of camouflage by calling us ‘the Japanese’ – perhaps in tribute to the impenetrability of the oriental mind. Mr Churchill has told how, while he was having a fateful interview with the Prime Minster in the political crisis of 1915, he was summoned by telephone to come back to the Admiralty at once because ‘very important news of the kind that never fails’ had just come in. It was an agreeable periphrasis, ‘news of the kind that never fails’.8
The Admiralty made a number of changes to its hierarchy in the latter part of 1914. On 30 October, Admiral John (‘Jacky’) Fisher was appointed as its professional head and First Sea Lord. One other key change was the appointment of the 50-year-old Rear-Admiral H.F. Oliver as Chief of the War Staff (COS).9 According to Churchill:
The decision to recall Lord Fisher to the Admiralty was very important. He was, as has been here contended, the most distinguished British Naval officer since Nelson. The originality of his mind and the spontaneity of his nature freed him from conventionalities of all kinds. His genius was deep and true. Above all, he was in harmony with the vast size of events. Like them, he was built upon a titanic scale. But he was seventy-four years of age. As in a great castle which has long contended with time, the mighty central mass of the donjon towered up intact and seemingly everlasting. But the outworks and the battlements had fallen away, and its imperious ruler dwelt only in the special apartments and corridors with which he had a lifelong familiarity. Had he and his comrade, Sir Arthur Wilson,10 been born ten years later, the British naval direction at the outbreak of the Great War would have reached its highest state of perfection, both at the Admiralty and afloat.11
Oliver had been Director of the Intelligence Division (DID) and his replacement was Captain Reginald Hall12, who had had to give up his previous command as captain of the new battlecruiser Queen Mary due to ill health. Hall’s nickname was ‘Blinker’, apparently because of a chronic facial twitch. It was only when he took up his position that he became aware of the small section working for Sir Alfred Ewing13 on intercepting German naval wireless signals. They were successful enough to be supplying the Operations Division with information about German fleet movements.
Herbert Hope, a commander at the Admiralty, was given charge of the activities in Room 40. He was to keep the Operations and Intelligence Divisions informed of the activities of the German fleet from the Section’s cryptographic work. They had expanded and nine rooms were now assigned to them. Hope was responsible for receiving decrypted messages and attaching remarks to them before and taking them to Hall, who in turn took them to the First Lord (Churchill) and then to the COS (Admiral Oliver). Ewing’s section became part of Hall’s Naval Intelligence Division but Hope was kept away from the cryptanalysts in Room 40 and this isolated position was a problem. On 16 November 1914, he became Hall’s representative in charge of the staff of cryptanalysts in Room 40.
Hall was not prepared to simply sit back and oversee this successful operation. It was his job, he believed, to find out everything he could about possible future theatres of naval operations and to be ready for any future conflict involving Germany. He strengthened his staff and Claud Serocold became his PA. Others new recruits included Lord Herschell, Lord Abinger, Mr Vaughan-Williams, Ralph Nevill, 2nd Lieutenant Leeson of the Middlesex Regiment – who had been invalided out of the Army in France – James Randall, a City wine merchant, L.G. Wickham-Legg, a Fellow of New College, Oxford, George Prothero, editor of the Quarterly Review, Algernon Cecil, a distinguished historian, Harold Russell, a barrister, Gerald Fitzmaurice, Dragoman at the Constantinople Embassy, Thomas Inskip, KC and Sir Philip Baker Wilbraham, Fellow of All Souls, Oxford and Chancellor of the Diocese of Chester. All had specific tasks. Wickham-Legg and Leeson, for example, prepared the summaries of intelligence that were issued to the Fleet twice a week. Hall also took the unusual step of recruiting women, something quite new in a department undertaking secret work. The criteria for selection was that candidates needed to have naval connections, know at least two foreign languages and be able to type. They became known as ‘Blinkers’ Beauty Chorus’ and in this way he was able to expand the establishment of Room 40. AGD continued to be on loan to Room 40 from Osborne until September 1915, when he makes his last appearance in the Osborne Lists as ‘Serving at Admiralty’. Interestingly, his last appearance on the Navy List as a member of an educational establishment is April 1915.
Soon after taking up his post, Hall turned his attention to the organisation which was in place for the censorship of telegrams, cables and wireless traffic and for the censorship of all incoming, outgoing and transit mails. He initiated discussions with Colonel Cockerill (later Sir George Cockerill) who soon after became Director of Special Intelligence at the War Office. They agreed to set up a private censorship organisation which would open all foreign mail coming into the main post office at Mount Pleasant. Hall took on the task of finding the manpower and money to do so. Hall won over Churchill to his proposal and within a few days the new department was hard at work. Eventually the Foreign Secretary, McKenna, and Prime Minister Asquith supported Hall’s proposal and a permanent department, a War Trade Intelligence Department, was set up in the War Office to carry out the work.
***
In 1914, like every army in Europe, the British Army was badly prepared to acquire and exploit intelligence from its enemy’s communications. Before 1914, European armies assumed that communications would be based on telegraph, telephone and despatch riders and no army developed effective cryptographic systems
which exploited wireless radio with an acceptable level of security. However, intelligence was available through other channels such as reconnaissance, the capture of prisoners and documents as well as networks of agents. As the use of wireless radio increased between 1896 and 1914, enemy communications could be intercepted and there were more opportunities to attack new cipher systems as they were introduced by the enemy. After the Boer War ended, the next significant British Sigint effort was established in 1906 as part of the Intelligence Branch of the Indian Army’s headquarters at Simla, today the capital of the northern Indian state of Himachal Pradesh. A single cryptanalyst, George Church and one linguist, Captain Gerald Palmer, were employed and, by 1912, they were producing decrypts from Russian, Persian, Chinese and Tibetan communications.14