by Alastair Denniston- Code-breaking From Room 40 to Berkeley Street
Until June 1917 the British Army’s overseas Sigint service, both in terms of operations and intelligence, existed as a separate though parallel branch of the Army Signal Service. Afterwards, the intelligence sections were transferred to the newly formed special branch of the Intelligence Staff at GHQ, Section I9e at Le Touquet, headed by Captain O.T. Hitchings. Its success was measured for the first time as a force multiplier, in this case, the equivalent of four divisions. The Army also supplied intelligence for the Royal Flying Corps, enabling early warning of enemy attacks. While GHQ had a Wireless Observation Group (WOG) monitoring the traffic of German radio stations behind the lines, interception of tactical communications was covered by other WOGs, each with around fifty personnel. WOGs were also established in other operational theatres. The first, No. 2, was formed in Egypt in 1916 to cover operations in Palestine. Later that year, No. 3 was started in Salonica and No. 4 in Mesopotamia, covering Southern Russia and Anatolia. They worked closely together, exchanged intercepts, traffic analysis reports, cryptographic data and translations. They shared a cipher with MI1(b) to which they supplied data and material. While detailed information about the operation of these groups is unknown, their legacy would be seen in WW2.
***
The existence of Hall’s cryptographic section came to the attention of Churchill, in his capacity as First Lord, but his interest was greeted with some scepticism by AGD. Writing in 1919 he said:
The First Lord, Winston Churchill, now took official note of the existence of the Section and issued its charter. As is seen, he laid down certain instructions for the distribution of the translations. One is bound to confess that the First Lord’s view of the possibilities of cryptography appear now distinctly limited. To have carried out his instructions literally would, no doubt, have safeguarded the secret but must also have nullified the value of the messages.38
Churchill dictated his ‘charter’ to his personal secretary Edward March on 8 November 1914 and initialled it ‘WSC’ in red ink. It was addressed to Ewing and Fisher, who initialled it ‘F’ in green ink. The charter document39 said:
An officer of the War Staff, preferably from the I.D., should be selected to study all the decoded intercepts, not only current but past, and to compare them continually with what actually took place in order to penetrate the German mind and movements, and make reports. All these intercepts are to be written in a locked book with their decodes, and all other copies are to be collected and burnt. All new messages are to be entered in the book, and the book is only to be handled under direction from COS.
The officer selected is for the present to do no other work.
I shall be grateful if Sir Alfred Ewing will associate himself continuously with this work.
Fisher passed it on to Hall, who replied the next day:
I have consulted with Sir Alfred Ewing and propose that Fleet Paymaster Rotter be detailed exclusively for this work (he discovered the code). The system at present in force is as follows. All intercepts are decoded immediately they are received. The original intercept is then filed and kept under lock and key. The translation is entered in a book which is kept under lock and key. Two copies are made out of the translation – one sent by hand and given personally to COS, the other given to DID. This system ensures that the information is given at once to the responsible people, the COS to act as necessary, the DID to compare with information from other sources. DID’s copy is kept under lock and key and is seen by no one but DID. In future, the envelopes will be marked ‘To be opened only by – - -.’ I would point out that to carry the book round will entail much delay and will not save copies being taken as so many messages are being received. I would therefore propose that the work be continued under the direction of Sir Alfred Ewing on the lines indicated above.
AGD’s assessment of Churchill’s charter document proved to be correct, as his insistence on security proved unhelpful. It only served to cut off Room 40 from the rest of Naval Intelligence and hence limited the impact that its work might have had. Churchill’s ‘charter’ seemed to make Ewing and Room 40 responsible to Oliver as COS. However, given Oliver’s workload, it in effect meant that Ewing was not responsible to anyone, a situation which was unlikely to sit well with Hall. Oliver would later record that ‘only the First Lord, the First Sea Lord, the Second Sea Lord, the Secretary [of the Admiralty], Sir Arthur Wilson, me and Hall, the DOD [Director Operations Division], the Assistant Director and my three Duty Captains were supposed to be aware [of Rooms 40’s work] in the Admiralty. Churchill may have told the Prime Minister but I never had any evidence that the rest of the War Cabinet knew.’40
The Operations Division was also sceptical due to two unfortunate incidents which occurred even before Ewing’s staff moved to Room 40. It stemmed from the fact that any signal that could be read was circulated without proper analysis or comment. According to AGD:
Owing to poor interception and lack of knowledge on the part of the staff, a signal was circulated alleging that the Ariadne was proceeding to the Jade. The Operations Division knew that the Ariadne was sunk in the Heligoland Bight41 action. Worse than that, a message was circulated on two or three successive evenings purporting to order destroyers to patrol the Inner Gabbard. The COS took counter-action and, at some considerable trouble and expense, English destroyers also patrolled that spot and never found the enemy. Subsequently it was found that the German destroyers had merely been ordered to proceed to Heligoland, an island which could only be distinguished from the Inner Gabbard by the bar over the letter ‘A’ which had escaped the notice of the inexperienced and geographically ignorant watchkeeper.
Further, any signal which could be read was circulated without comment and for reasons best known to W/T experts, many of those emanating from Bulk were among the best intercepted and hence most easily read. The poor watchkeepers had the haziest of notions as to the whereabouts of Bulk but the Operations Division cannot be blamed for the lack of enthusiasm for the times at which the Kiel barrier was opened. The watchkeepers knew nothing of the German Fleet, very little of the geography of the German coastline, while their ignorance of English and German naval phraseology was profound. Hope did his best for them while Lord Fisher pointed out that warships did not ‘run in’ and begged the staff to adopt the word ‘proceed’.42
British Sigint was truly born in November 1914 and AGD was at the heart of the action. Room 40 consisted of Ewing in charge, Commander Hope and Fleet Paymaster Rotter dealing respectively with the intelligence and cryptographic sides of the work, Herschell, Denniston and Norton as watchkeepers, and Russell Clarke and Hippiseley at the Hunstanton intercept station. This was supplemented by the educational staff, Naval instructors Parish, Curtis and Professor Henderson. Their ‘tools’ consisted of a copy of the German Navy Signal Book (additional copies were made) and the HVB which had been captured by the Australians. It was used by the whole German High Sea Fleet, submarines and airships, albeit in re-encrypted form, until March 1915. It proved invaluable, particularly in gaining advance warning of air raids. They also had the VB book which DID had handed over to Ewing when it arrived in the Admiralty. Herschell was moved to DID to head up a team to deal with the translation of the mass of secret papers recovered along with the VB book. The watches were brought up to two-man strength with the addition of Monk Bretton, Hopkinson, Fremantle, Lawrence and Morrah in December. According to AGD:
They knew ordinary literary German fluently and they could be relied on. But of cryptography, of naval German, of the habits of war vessels of any nationality they knew not a jot. Their training was of the shortest before they were sent off onto watches of two men each and given the responsibility of looking after the German Fleet. Worse than that, they had to learn the intricacies of the office routine. They probably had more than their fair share of log writing, they had to sort and circulate. They had to turn the German squared chart into latitude and longitude of which they had not heard since the geography class of their school days. There
was no traditional routine to be followed. New methods had to be evolved to meet new problems.43
Gradually during November, Ewing’s team had succeeded in translating intelligible portions of various German naval messages. Most were routine, such as: ‘One of our torpedo boats will be running out into square 7T at 8 p.m.’ According to Churchill: ‘Admiral Oliver has the foresight to begin setting up directional stations in August 1914. The Admiralty ‘thus carried to an unrivalled and indeed unapproached degree of perfection our means of fixing the position and, by successive positions, the course of any enemy ship that used its wireless installation.’44
By the beginning of 1915, Russell Clarke’s new Y stations were intercepting most German signals and Room 40, having mastered the SKM and HVB codes, was reading all of them. Room 40’s decrypts let them keep a watch on German ship movements as well as important inferences to be drawn regarding such things as the composition of the enemy squadrons, the presence of new ships, the position of minefields, etc. Deducing such information from these messages required specific expertise, which was why Rear-Admiral Hope was lent to Room 40 from the Intelligence Division to help with this work.
AGD recorded Room 40’s daily routine in November 1914 as follows:
Hope and Rotter were present daily from 9am till 7pm, the former dealing with the translated messages, the latter working on the many fragments and examining the unknown. The man on watch had to sort, decode and translate the new.
Hunstanton, Stockton Leafield and Hall Street had direct lines to the Admiralty. There was a never-ending stream of postmen delivering bundles of intercepts. In a few months these men were replaced by an automatic tube which discharged the goods into a basket with a rush which shook the nerve of any unwitting visitor and very much disturbed the slumbers of a night watchman taking his time off.
In the very early days every message which appeared to give sense to the man on duty was ‘logged’ and ‘sent’. That is, the translation was written in the current log book and three copies were made for circulation, one for the COS and one for DID and one for Hope. With luck, there were three or four copies of every message from the various stations. These had to be pinned together and stacked in the file of logged messages. But still there were a vast number of fragments, of messages which failed to satisfy the fastidious German taste of the watchkeeper, or messages in unknown codes and languages. All these were bundled into a tin on which was printed ‘N.S.L.’. It was a very important tin, nearly always very full in those days, but to explain it to the many newcomers was one of the most complex points in a very complicated system. Truly N.S.L. only meant ‘neither sent or logged’. When the war was finished there was still a box called N.S.L. when there had been no log for the last two years. N.S.L. was a living thing with a specific meaning, and it recounted how a night watchman woke trembling in a sweat – he had dreamt he had been sent to the N.S.L. and got lost.
The log became an object of hatred before long. The First Lord had called into being that particular form of filing the current work and it was over two years, when its originator was elsewhere, before a more labour-saving and less soul-destroying method was allowed to replace it. In the days when a watchkeeper averaged 12 messages it could be written up, though even then it was the fashion to let the messages accumulate and allow the new watch to write up the log, and thus appreciate the situation! But it was beyond a joke when naval actions were pending or zepps fluttering and the watchkeeper had 12 to 20 pages of the book to write up.
For two months at least the night man had a lonely time, though he was probably too busy to note it. It was no good bringing pyjamas in those days or hoping the Admiralty would provide a bath. All that was needed was plenty of sandwiches. Tastes in drinks varied and only one man is alleged to have worked throughout the night with a revolver at his elbow.45
By 1915, all German naval signals which could be intercepted were read and circulated since almost all German communications were encrypted using one of the three books in the possession of Room 40. The disposition (strength as well as general location) of its High Seas Fleet was known, along with information about its submarines and airships. In January 1915, Hope began submitting a daily return to Churchill, Fisher, Wilson and Oliver which set out the last known position of every U-boat mentioned in decodes. He had built up an accurate idea of the total operational strength, of the location and state of readiness of every individual boat if in port, and when and if it put to sea. However, at this stage Room 40 was not an intelligence centre. It was not allowed to keep a plot of British warships or merchant ships, so could not know whether a decrypted U-boat position posed a risk.
According to AGD:
No attempt was made to develop any intelligence side of the work, beyond Hope’s duty of instructing the authorities on the real meaning of certain signals. The request that 40 OB should be allowed to keep a flagged chart of the German coastline was vetoed as an unnecessary duplication of the work in the Operations Division. (In May 1917 this request received sanction). But all naval signals were read even without intelligence. True it is that in certain cupboards there were increasing piles of ‘stuff’ which was not read but it was not naval German. The art of reading other peoples’ telegrams was still in extreme infancy; no one then imagined that all those piles contained telegrams possibly of the greatest interest which could be read and, in 1915 it may be said, read without extreme difficulty.46
However, intelligence was not always accepted from NID without questions, as shown by this note from Churchill to NID on 7 November 1914:
With reference to your report of yesterday, apparently attaching credence to a statement that from 100 to 200 small submarines have been manufactured secretly in Germany, have you considered how many trained officers and personnel this important flotilla would require? What evidence is there at your disposal to show that the Germans have trained this number of submarine captains and officers? I have always understood that their flotilla of submarines before the war did not exceed 27. There are no personnel that require more careful training than the submarine personnel. All the experience of our officers shows that a submarine depends for its effectiveness mainly upon its captain. The function of the Naval Intelligence Division is not merely to collect and pass on the Munchausen tales of spies and untrustworthy agents, but carefully to sift and scrutinize the intelligence they receive, and in putting it forward to indicate the degree of probability which attaches to it. It appears to me impossible that any large addition to the German submarine force can be made for many months to come. Even if the difficulties of material were overcome those of personnel would impose an absolute limit. It is very likely that a few small portable submarines have been prepared for coast work.
In early 1915, Room 40 was clearly understaffed and Ewing and Hall both began looking for suitable new recruits. Two naval schoolmasters, Edmund Green and G.L.N. Hope, were brought in first, followed by B.F. Talbot, P.A. Somers-Cocks and George Young from the Diplomatic Service. On the academic side, Dilwyn ‘Dilly’ Knox may have been the first recruit. He was a 31-year-old Greek scholar and Fellow of King’s College, Cambridge.47 Knox started off ‘learning the ropes’ as a watchkeeper but quickly demonstrated a flair for cryptanalysis. He was given Room 53, a cubbyhole of a room which housed a bath for the use of men on the night watch. This was no problem for Knox and legend has it that he did his best work in a hot bath.48 Knox shared a house in Chelsea with another King’s fellow, Frank Birch,49 seven years his junior, who arrived at the end of 1915 or early 1916. He specialised in the analysis and assessment of the intelligence, rather than in cryptanalysis. As a serving Royal Naval Volunteer Reserve (RNVR) officer, Birch was one of the few men working in Room 40 who wore naval uniform. He was active in the theatre and apparently played the Widow Twankie in professional pantomimes at Christmas. A third King’s man joined them, Frank E. Adcock, lay dean of the college.50 Adcock did not share the eccentricities of his King’s colleagues and stayed true to his chosen fi
eld of Ancient History. In contrast to the King’s men, the Rev. William Montgomery51 came from Westminster Presbyterian College, Cambridge, and contributed linguistic skills. Also arriving in 1915 was Edward Bullough from Caius College, Cambridge and another linguist, along with Professor L.A. Willoughby, a German specialist from London University.
Hall personally recruited Nigel de Grey52, who had joined the RNVR and become an observer in the Balloon Section of the Royal Naval Air Service in Belgium. In mid-1915, Hall formed a separate Diplomatic Section with George Young as its head and early members including de Grey, Flaudel-Phillips and Montgomery. It occupied Room 45 but retained the name Room 40. They struggled to break German diplomatic codes, which were far more complex than those used by the Imperial Navy until a copy of their codes was captured. W.F. Clarke,53 a barrister, arrived in early 1916, having secured a commission as an Assistant Paymaster RNVR at the beginning of 1915. Further recruits were drawn from wounded Army officers such as Lionel Fraser, a friend of Serocold. Benjamin Flaudel-Phillips was another Serocold recruit and he succeeded Young as head of the Diplomatic Section.
Room 40 assumed that eventually a change in cipher key would occur and, in early January, the evening watch was unable to read a batch of signals delivered to them. Through a concerted effort by all available staff, the new key was obtained by the next morning. Churchill called to offer his personal congratulations. Ironically, subsequent work during the day revealed that the key had not been changed but just modified slightly, something that should have been detected quite quickly. Room 40 kept the information to itself to avoid embarrassment, and when the key did actually change a few days later, it was solved in a few hours without fanfare. By 1917, the key was changing every night at midnight. The number of intercepting aerials controlled by Clarke and the Marconi Company increased during the spring of 1915 and this led to a very large increase in intercepted traffic. The operators were starting to discover different frequencies being used by different German communication networks. AGD summoned up their increasing expertise: