‘Blinker’ Hall was convinced that the failures during the Battle of Jutland would not have happened had Room 40 been allowed to submit full intelligence reports, based on their experience of German signals, instead of just passing on transcripts whose significance was not always grasped by the Operations Division. He was determined this would not happen again and began by setting up a war diary, but it would take his next move to achieve effective intelligence reorganisation. He decided the time had come to get in on strategic as well as tactical intelligence through setting up a new diplomatic unit under his own control. This began in the summer of 1915 when a diplomatic codebook was obtained from the Persian city of Bushehr on the Gulf. The German consul had left it behind when his consulate was hurriedly abandoned. Ben Faudel-Phillips, the Rev. Montgomery and Nigel de Grey would later join this diplomatic sub-section, headed by Sir George Young, himself an experienced diplomat. In order to conceal that they were engaged in reading diplomatic communications between neutral countries, they were referred to as the ‘research party’, rather than the diplomatic sub-section. In early 1917, they were dealt a diplomatic trump card which would greatly increase the reputation of Room 40. Dilly was also involved, and greatly admired the way Hall played his hand, particularly how he covered his tracks and protected the secret source of the intelligence provided.
On 16 January 1917, Arthur Zimmermann, the German Foreign Secretary, dispatched a long coded telegram to the German ambassador in Mexico, Heinrich von Eckhardt, which proved to be political dynamite. It had to be sent first by transatlantic cable to the German ambassador in Washington, Johann von Bernstorff, before retransmission. President Woodrow Wilson had granted German diplomats the privilege of sending their material under cover of US diplomatic traffic, hoping that this would enable Germany to keep in touch with the United States and further Wilson’s aim of mediating an end to the war. The Zimmermann telegram to Washington was intercepted, as the cables touched on British soil at Porthcurno near Land’s End, and was received in Room 40 soon after it was sent. The telegram was sent in the high-grade diplomatic code 7500, which Room 40 had only just begun to solve. Unlike the telegraphese of operational naval messages, diplomatic telegrams were formal and verbose, necessitating large codebooks.
When the encoded Zimmermann telegram first arrived down the pneumatic tube on the night watch, it seems it was taken first to Dilly. He could make out that it began with the alarming sentence ‘We propose to begin on 1 February unrestrained submarine warfare’, which of course would be of immediate concern to Admiralty Operations, but the gist of what followed sent Dilly scuttling over to de Grey. It turned out to be nothing less than a daring proposal that the Mexican government should form a military alliance with Germany against the United States in exchange for the return of Texas, New Mexico and Arizona, lost by Mexico to the US in the 1846–8 war between the two countries. The precise detail of the telegram was not immediately clear because so few of the 7500 code’s groups had been recovered. De Grey was not only an expert German linguist but also a professional editor and as such was, like Dilly, very used to emending texts or in this case filling in gaps. Writing in 1945, de Grey recalled the moment as though it had occurred the previous day:
The telegram was first sorted to Knox, whose business it was to fill in any known groups. His knowledge of German was at that time too slender for him to tackle any difficult passages in telegrams. So that the procedure was that if the telegrams appeared from what could be read to have any interest he brought them to me for further study. We could at once read enough groups for Knox to see that the telegram was important. Together he and I worked solidly all the morning on it. With our crude methods and lack of staff, no elaborate indexing of groups had been developed, only constantly recurring groups were noted in the working copies of the code as our fancy dictated. Work therefore was slow and laborious, but by about midday we had a skeleton version, sweating with excitement as we went on, because neither of us doubted the importance of what we had in our hands.
It was then de Grey’s responsibility to report to Hall all matters emanating from the diplomatic ‘research party’ in which Dilly had no official role.
As soon as I felt sufficiently secure in our version even with all its gaps I took it down to Admiral Hall … although Ewing was nominally our head, ‘Blinker’ Hall had made a compact with a few of us of the ‘research party’, that if we ever dug up anything of real importance we were to take it direct to him without showing it to Ewing, whom he distrusted as a chatter-box. Blinker was always accessible to the lads of Room 40, at least he was always to me at that time because I was getting him all the news from diplomatic Germany. I was young and excited and I ran all the way to his room, found Serocold alone and Blinker free. I burst out breathlessly: ‘Do you want to bring America in the war, Sir?’
‘Yes, why?’ said Blinker Hall.
‘I’ve got a telegram that will bring them in if you give it to them.’
Hall discussed it animatedly with Serocold and de Grey, but however important the message was, even in its incomplete form, he knew it would have to be handled very carefully diplomatically. He asked for the original telegraph in code, which he locked in his safe, and said that on no account was anything to be said about it to anyone and nothing was to be put on the files. First of all, he had to try and calculate the motives behind this extraordinary action by the Germans and the likely reaction of President Wilson. Secondly, as there was every chance that the Americans would declare war if they received the information, he had to weigh that against the great disadvantage of the Germans then finding out that we were reading their communications, just as Room 40 was providing good intelligence from Dilly about the movement of German ships and U-boats through breaking their naval codes. It was essential not to let the Americans know we had been intercepting neutral messages as well as preventing the Germans from discovering that we were breaking their codes. Hall’s solution was nothing less than brilliant.
Hall rightly guessed that von Bernstorff would have to transmit the same message to Eckhardt in Mexico in a lower-grade code than the 7500 code. He also knew that the British embassy in Mexico had a source with access to the telegraph offices in Mexico City, whom he dubbed ‘Mr T’. This was in fact Edward Thurstan, the British chargé d’affaires, who ran the embassy’s source inside the telegraph office and was therefore able to obtain a complete copy of the encoded Zimmermann telegram as sent on by von Bernstorff to Mexico. It had been sent in diplomatic code 13040, an old code very similar to the one obtained at Bushehr, which Room 40 had been working on since 1915, with Montgomery in charge of trying to reconstruct its 11,000 groups. Although there were still a lot of unsolved groups in Room 40’s reconstructed codebook, Montgomery and de Grey managed to use it to fill in the gaps in the Zimmermann telegram and were able to present Hall with the complete text. This was a triumph as Germany would think that the text of the telegram had been achieved by cloak-and-dagger work, which although regarded as dirty tricks was universally carried out, as indeed in part it had in this case, and would not reveal the interception and breaking of the secret diplomatic 7500 code, which had alerted Room 40 in the first place. However, when the moment came to reveal the content of the message, the doubtful US ambassador in London, Walter Page, wanted proof that it was not a fake and that they could in fact decode 13040 messages. De Grey obligingly went along to the US embassy with the codebook they had compiled, and to his great satisfaction the embassy secretary, Edward Bell, was able to decode a dozen groups of the Zimmermann telegram himself. But de Grey did the rest, and recalled having to bluff with some, as they weren’t all written up in his version of the book:
Being in a hurry, I grabbed my own version of 13040 without thinking and went off to [Hall’s] room. There Edward Bell produced a copy of the telegram and invited me to decipher it in his presence and to explain the system as I went along. I gaily proceeded and all went well with the first few groups but then on coming to the
next I found my book blank and realised with horror that I hadn’t done my homework. I had not written up my book and this was by way of being a demonstration to the Americans of the absolute cast iron certainty of our story, good enough to carry firm conviction to their hesitating hearts. If I stopped and fetched another book he would suspect at once that we had faked it up for his benefit. If I let him see that I was writing it down out of my head he would not believe me. If he did not believe me we should fail and have lost the greatest opportunity ever presented to us. Several seconds of bloody sweat. Then I bluffed. I showed him all the groups when they had been written in my book and passed quickly over those that were not, writing the words into the copy of the telegram by heart. Edward Bell, most charming man, was thoroughly convinced – the more easily I think in that he wanted to be convinced anyhow and regarded the whole thing as black magic.
Once persuaded by Bell that it was genuine, Page transmitted the content to President Wilson on February 23 and when it was released to the press, the ensuing outrage easily persuaded Congress to declare war on Germany.
Hall hushed everything up and destroyed the evidence, and it has been difficult to verify later stories from Room 40 insiders who were not actually there at the time. Dilly’s family knew that he was frustrated by the secrecy imposed within the department. Penelope Knox, his niece, revealed in The Knox Brothers that the way in for the Zimmermann triumph had been through the second telegram obtained from the telegraph offices and that for its solution Dilly had felt ‘professional admiration but also professional jealousy’. The latter was that he was not in at the kill as de Grey and Montgomery had been with Hall. The ‘professional admiration’ was for Blinker’s getting 13040 from the telegraph offices, which gave rise to Dilly’s future plaintive murmurs of ‘Can’t we get something from the post office?’ even in inappropriate situations.
After the great success of the Zimmermann telegram, Room 40 expanded into a set of rooms nearby to become known officially on 17 May 1917 as ID25, Section 25 of the Naval Intelligence Division, with small units dedicated to specialised tasks. In spite of its upgraded status, members preferred to continue to call themselves Room 40, as would happen at Bletchley Park when the staff of the iconic small Hut 6 and Hut 8 were moved to large blocks. Commander William James was now appointed as head to see that in future signals intelligence would be closely co-ordinated with the Operations Division. Hall sent de Grey out to Rome to liaise with Italian-based codebreakers and co-ordinate Mediterranean intelligence and Faudel-Phillips was then put in charge of the diplomatic section, now in Room 45. As the Germans had threatened in the Zimmermann telegram, unrestricted submarine warfare was now carried out and 835,000 tons of allied shipping were lost in the month after America declared war. A convoy system was set up and ID25’s most important work would be concentrated on intelligence relating to the position of the U-boats. A submarine tracking room operated in the Admiralty under Fleet Paymaster Ernest Thring and was directly linked by pneumatic tube to the new direction-finding section Hall set up in ID25 under Frank Tiarks, whom he recruited from the City. The direction-finding network locked onto the signal from enemy wireless transmitters and took a bearing to their location. By using several direction-finding stations to track down the transmitters, a series of bearings could be traced on a map and where they met was the location of the enemy transmitter. The creation of the new sections to work on the codebreakers’ material meant there was no longer any delay in it reaching the Operations Division, who could make immediate use of it, and there was soon a noticeable increase in the number of U-boats sunk.
There was great alarm when the SKM codebook, which had served Room 40 so well, was completely replaced by the new FFB book, at a time when it was essential to be able to read the flag code in order to track the submarines. This meant not just a key system that had to be deciphered, but a new codebook to be compiled from scratch without the benefit of a capture as hitherto. By the time the convoy system was set up in June, Dilly had made good headway with the new code and it became possible to divert convoys from dangerous areas. Renewed efforts were made to recover papers from any submarine sunk in water shallow enough for divers to operate. Dilly had acquired the small Room 53 in ID25 to work on the flag code. The room had its own bath, which was useful for Dilly since this was where he would do a lot of his thinking, whether on the mysteries of Greek papyri or German codes. He was also given a secretary, Miss Olive Roddam, the daughter of a Northumberland landowner, to help him. There was one problem, however, as he kept mostly night watches to be ready for the key change, but his secretary worked normal office hours and he had to rush to get out of his bath and into his uniform before she arrived. In common with all the other civilians in Room 40, Dilly was commissioned into the Royal Naval Volunteer Reserve in July 1917. Miss Roddam was the first woman outside his own family with whom Dilly had ever spent any time in close proximity. She meanwhile had come to London in the wake of the loss of her fiancé in France. It was perhaps inevitable that they would become close and once the war was over, and sights could be set on better times, Dilly would marry Miss Roddam.
There had been great celebrations when Dilly Knox found a way in to the new high-level SKM flag code and its key system. He had found a telegram which gave the notification that it was a practice message for the new code in the preamble. A practice message would come straight out of the operator’s head and would almost certainly have a lot of words outside normal naval terminology so the operator would have to use code-groups representing syllables. Dilly noticed a pattern of a repeated syllable, which he assumed to be ‘en’, which – since in German many plurals and all the basic forms of verbs end in ‘en’ – is the most common German syllable; but there seemed to be rather a lot of them for such a short message and he soon detected that it was a poem with rhyming dactyls. Dilly believed that if the German operator had sent out a romantic poem, it was likely to include roses, which is rendered in German as Rosen and therefore also ends with the ‘en’ syllable, and in any case there would not be much opportunity to use naval telegraphese in a poem, whether or not it included roses. Normally, in the old SKM codebook, unusual words would have to be made up from relevant coded syllables and the inclusion of a large number of plurals and infinitives would account for the frequent syllable substitution of ‘en’ in the practice message. Even so, there were rather a lot of them for a short message. Dilly soon detected a pattern and tried out various syllabic metres, writing down examples of possible spaced syllables, with the repeated syllables he considered to be ‘en’ in what would have looked like this:
— — — — en
— — en — — en
— — — — en
— — — — — en
He then took his scribble along the corridor to the German scholars in the hope that they could piece it together through metre as he had done with his Herodas fragments. Dr Leonard Willoughby, the Oxford Taylorian lecturer, was an expert on the German Romantic movement and it did not take him long to identify a Schiller epigram, indeed with roses, heavenly ones at that, from Dilly’s spaced out ‘en’ syllables:
Ehret die Frauen; sie flechten und weben
Himmlische Rosen ins irdische Leben
[Give reverence to women; they plait and weave
Heavenly roses in this earthly life]
This enabled Dilly to find the substitution ‘key’ and painstakingly work back to compiling the FFB code. Commander James, who was in charge of analysing the intelligence derived from the breaking of the U-boat code, remembered the little room where Dilly was ‘labouring’ over the apparently insoluble as one of the most ‘astonishing sights’ of his department.
Professor Walter Bruford, then the last surviving member of Room 40, himself a German scholar, was able to give this information about the break to Penelope Fitzgerald for her book. Hall felt personally involved and every time a submarine was sunk he said: ‘Willoughby, go and fetch the rum.’ Willoughby was in cha
rge of writing up the translation of each message in the logbook; but doubtless he and Dilly and the other ‘professor types’ would have much preferred to celebrate with the Madeira or brandy they were accustomed to in their common rooms.
The Germans introduced many more different key systems which needed constant vigilance but, astonishingly, this small group of codebreakers in Room 40 had decoded and deciphered about 20,000 messages by the end of the war. The last of these messages was the thrilling climax, giving the news that Admiral Reinhard Scheer, the German chief of naval staff, hinted that the High Seas Fleet was to be scuttled. Alastair Denniston went out from ID25 to act as interpreter for Admiral Sir David Beatty, commander-in-chief of the Grand Fleet, at the surrender of the German fleet in Scapa Flow. Dilly Knox and Frank Birch thought that a suitable end-of-war celebration would be a pantomime parody of codebreaking to be performed at the Chelsea house they shared with a friend.
My interpretation of Alice in ID25, which follows, is from Dilly’s own copy.
THREE
Alice in ID25
‘I’m afraid it’s all Greek to me’
Alice in ID25, written by Frank Birch with poems by Dilly Knox, both devoted Carrollians, is a brilliantly comic, but very authentic, skit of Room 40, following closely the pattern of Lewis Carroll’s own Alice’s Adventures in Wonderland and full of the Carrollian ‘chopped logic’ which Dilly loved so much. It was performed by the codebreakers as a pantomime for their own entertainment during a concert they held at the end of the First World War. Many of the characters would reappear at Bletchley Park. One of the songs performed at the concert makes reference to Dilly and Olive:
Dilly Page 5