FOUR
Between the wars
Demobilisation in 1919 meant a far-reaching decision for Dilly Knox. Like so many of his Room 40 colleagues, he had come to see the excitement of wartime codebreaking as a way of life; as he said at the end of Alice in ID25:
Oh, if a time should ever come when we’re demobilised,
How we shall miss the interests which once our life comprised!
Dilly had every opportunity to return to King’s and had even been appointed librarian fellow in his absence; there he could have accomplished the task of completing the Herodas text he was determined to do, in an academic environment. After careful consideration, however, he decided not to abandon cryptography, but to become part of the new Government Code and Cypher School (GC&CS), which the Cabinet Secret Service Committee decided to establish in 1919. Its public function was ‘to advise as to the security of codes and ciphers used by all Government departments and to assist in their provision’; its secret directive was ‘to study the methods of cipher communications used by foreign powers’. Dilly felt that a peacetime cryptographic unit, with normal working hours, would allow him to pursue his papyrus research with the British Museum to hand, and finish Herodas in the home he planned to set up with Olive Roddam.
Meanwhile, he and Frank Birch would keep on the shared house in Edith Grove, Chelsea. Birch, having achieved a double first in the Historical Tripos, was an accomplished historian and would return to King’s as a history lecturer, but not until 1921. On leaving the Admiralty’s service in 1919, he was awarded an OBE and was officially engaged to put Room 40’s records in order and to write a comprehensive internal history, A Contribution to the History of German Naval Warfare 1914–1918, in collaboration with William ‘Nobby’ Clarke. After Clarke had finished his part of the historical research in 1921, he decided to abandon his legal career to join Dilly in the new GC&CS. At a salary of £500 a year he regarded himself a public benefactor, but for Dilly, like the rest of the Bishop’s family, money was never a major consideration.
Although Dilly was not officially concerned with the history of Room 40, he clearly was involved with the discussions at Edith Grove on the implications of their wartime work. He is on record as saying that ‘at the end of the last war, Birch, Clarke and I issued at great expense a report in which we were unanimous in denouncing the system of “working in Blinkers”’. Although Dilly whole-heartedly supported ‘Blinker’ Hall’s achievements and his determination to keep their work out of the public domain, he felt secrecy about results within Room 40 had been counter-productive. He clearly still resented the Zimmermann telegram episode. Doubtless the paper would have included Dilly’s perpetual grouse that a scholar could only operate if he could see his work through from start to finish.
Hall had left his post, however, and his dynamic energy was no longer officially available to the new organisation. His replacement, Admiral Hugh Sinclair, who held the new title of director of naval intelligence (DNI), was a bon viveur of some repute, known to his friends as ‘Quex’ from the title character in Arthur Pinero’s popular play The Gay Lord Quex, who was supposedly ‘the wickedest man in London’. Sinclair took on the task of forming the new GC&CS from the remnants of Room 40 and MI1b, the War Office cryptographic department, while still keeping it under Admiralty control. The navy’s codebreakers had undoubtedly been a success but the Army’s MI1b had also had considerable success against German military codes and ciphers. There had, however, been very little co-operation between the two, so that they were entering uncharted waters when they amalgamated. Sinclair moved the codebreakers out of the Admiralty and War Office and into new premises at Watergate House, on the Strand, a short distance from the Savoy Grill, his favourite eating place.
One person determined to see that the Admiralty would have a decisive input was Winston Churchill, who, although now Secretary of State for War, always associated himself with Room 40’s activities and the importance of the intelligence derived from them. Before the meeting on the future GC&CS he wrote to Lord Drogheda at the Foreign Office setting out the unique skills they had acquired, which he had personally observed and which he almost seemed to be adopting as his own:
Our work has been done in the face of the enemy and always against time. The messages we have had to decipher were from ships at sea, engaged in actual operations, or from airships also operating. We have had to master a new key every morning before we could begin to read the messages, and sometimes we have had to grapple with two or three keys in one day! This has of necessity developed a kind of aptitude for the work, which depends on its success more on the study of the psychology of the persons sending out the messages and a sort of instinctive ‘flair’ for the kind of things they are saying, than upon careful study and analysis for which there is no time.
Churchill made it quite clear that the Admiralty should consent to pool Room 40 staff with those in the War Office only on condition that Commander Alastair Denniston was in charge of the new department. This was agreed and the ‘Little Man’ was duly appointed. Oliver Strachey, brother of Lytton Strachey, had been chief cryptographer in MI1b and Dilly would work closely with him in the years ahead. It was, however, a very small organisation in 1919, for the Treasury had throughout the negotiations insisted on cutting expenses. Only one wireless intercept station was retained, on the grounds that only cable traffic would be dealt with in future. ‘The inevitable had happened,’ Clarke recalled later. ‘There seemed no longer any need to study the communications of a naval or military nature. The navy and army of Germany had disappeared, never – it was supposed – to rise again, and the idea of watching those of Allied or friendly powers did not seem worthwhile.’ On the other hand, the volume of diplomatic traffic grew steadily, as there were conferences still at work for some years settling the boundaries of Europe. Under a section added to the 1920 Official Secrets Act, the international cable companies were obliged to hand over copies of cables passing through the United Kingdom. Dilly and Oliver Strachey spent most of their time working on American traffic and a new US code in which messages were first encoded and the encoded message was then super-enciphered. The cipher tables changed quarterly. It took a year to compile the codebook as the code-groups were not in alphabetical order but ‘hatted’, that is to say presented at random or out of a hat, but the task was finished in time to provide intelligence on American policy during the Washington Naval Conference in 1922.
Dilly’s domestic arrangements had changed. Birch married Vera Gage, daughter of Henry Charles, fifth Viscount Gage, in the autumn of 1919 and soon afterwards Dilly married Olive Roddam at her Northumberland home in 1920. Bishop Knox could not attend as he was in the thick of the Lambeth Conference, battling for the Book of Common Prayer. Olive was the daughter of Lieutenant-Colonel R. J. Roddam of the distinguished land-owning family the Roddams of Roddam Hall, and after a honeymoon in Scotland, the newly wed couple returned to Edith Grove, where it was thought they could set up home with the Birches. John Maynard Keynes had organised a generous gift from King’s for both couples. Dilly enthusiastically bought a cookery book and opened an account with the Kensington Unique Laundry, but the housekeeping arrangement was not going to work the way Frank and Dilly had planned. Olive did not like living in Chelsea as her heart was in the country.
Fortunately, Dilly had recently found himself in a position to enter the housing market, with a legacy from the Bishop’s first wife, Ellen French, the daughter of the unworldly missionary bishop, who had never spent a penny of his own family’s accumulated wealth. A condition of the outstanding French family inheritance had been that the youngest Knox child should have turned thirty and at the end of the war Ronnie had done so. Most of the legacy was in Great Western Railway shares and, in 1921, with his portion, Dilly went house-hunting with Olive in the country, a place he had hitherto only thought of as suitable for holidaying in or enjoying in classical imagery or with ‘the Shropshire lad’. He could not afford the upkeep of the kind of c
ountry house his wife was used to, but it would at least be a house in the country – albeit in the commutable Home Counties. At the asking price of £1,900 they settled on Courns Wood in 40 acres of Chiltern woodland a few miles from High Wycombe. Dilly caused some consternation among the estate agents by sending a classics lecturer and an electrical outfitter to value the property.
There was no way that a Knox would become a country gentleman. Dilly took no part in country sports, nor did he feel at home in the trendy High Wycombe stockbroker belt, although he could occasionally be persuaded to join tennis parties. There were also a number of retired servicemen in the neighbourhood, whose wives Dilly seemed particularly allergic to. He kept a notebook of verse describing his neighbours and wondered at the prevalence of adultery when all the wives looked so much alike.
Sir John has bought an aeroplane
Nor in his place would I refrain
From scouring earth and sky and sea
To get away from Lady D.
He also identified his nouveau riche neighbours with characters in his favourite ‘Just William’ stories. Olive did do her best to provide him with the sort of peaceful background to continue his life’s ambition to finish Walter Headlam’s uncompleted Herodas. He used to travel up from High Wycombe to GC&CS by train spreading out photographs of the fragments of the mimes on his knees.
Dilly explained some of the problems he had to deal with in piecing together the Herodas in the introduction to the Headlam/Knox Herodas. The papyrus, written in the first century or the first half of the second, was not only fragmented with many pieces missing, it was a copy of Herodas’s and the work of a basic copyist, not of a highly trained scribe. The copyist had miscopied and sometimes simply misread the material he was copying, Dilly said:
Not only was the writer constantly puzzled by the form of the letters which he was copying. Not only was he prone to all the common errors of copyists, but worst of all he suffered from a schoolboy-knowledge of Greek, and, where he followed the sense roughly, made, unconsciously, stupid alterations.
The copyist’s errors were ‘those of a man following the sense of the passage, often at a considerable distance’, Dilly said.
As well as using the metre, or the occasional lack of it, in the copyist’s writing, to work out where things had gone wrong, Dilly used the better preserved portions of papyrus as a ‘crib’ which allowed him to become used both to the Herodas style and the copyist’s handwriting, so he could better identify partial letters and work out where the copyist might have miscopied or misread the original text. But the problems were not simply caused by the copyist’s errors or the deterioration of the papyrus. Herodas himself moved easily and deliberately between two different Greek dialects, Ionic and Attic.
Just as Herodas allows himself the frequent use of different word-forms, Attic as well as Ionic, so in grammar, vocabulary and style he varies between Attic and Ionic. Sometimes his piquancy comes from giving an Ionic cast to an Attic word; sometimes it is an Ionic word in an Attic sentence where one translates into Attic to arrive at the sense. The grammar, normally Attic, assumes an Ionic cast occasionally. The rule that style is Attic and forms are Ionic is true in a broad sense, not absolutely. The pleasing incongruity, at which Herodas aims, binds him to no hard and fast rule.
Despite all these problems, Dilly succeeded in completing the Headlam/Knox Herodas, which finally appeared in 1922 and was lavishly praised by Professor W. G. Arnott as the restoration of an old master, ‘a glowing achievement not merely for the text of Herodas but for the Greek language and literature in general’. The following year Dilly made another contribution to Greek scholarship when he published an elegant dissertation on Cercidas, the third-century BC poet and cynic. Dilly was fascinated by the way Greek dramatists and poets used metre to stress their themes. For some time he had been studying the Cercidas papyrus fragments from the excavations at Oxyrhynchus. The longest of these fragments contained a discourse on how the nature of gods and beliefs did not seem to match up to the facts of life. Dilly now identified in a new papyrus an introduction to a collection of poems collected and introduced by Cercidas and published it as The First Greek Anthologist. Dilly was offered the professorship of Greek at Leeds University as a result. However tempting a professorship might be, it is doubtful whether Dilly gave it much consideration and in any case Leeds was out of the question for Olive, who was then expecting their second child. Their first son, Christopher Maynard, had been named after his godfather, John Maynard Keynes, with whom Dilly kept up a firm friendship. Now that he had no need to work on papyri on the train, Dilly bought a motor bike to get himself to London.
In 1922, the Foreign Office, having realised how useful the deciphering of diplomatic traffic under the Admiralty was to them, assumed control of GC&CS, which then had a largely civil role. In 1925, GC&CS moved to 54 Broadway Buildings, across the road from St James’s Park Underground station. The DNI, Admiral Hugh Sinclair, had now become ‘C’, the head of the Secret Intelligence Service (SIS), in charge of foreign espionage, but had retained control of the codebreakers, whose ‘special intelligence’ would become known simply as a ‘most secret source’. GC&CS was on the third floor of Broadway Buildings and SIS on the fifth; between them was a missionary society, which must have brought back memories to Dilly.
The main focus of the codebreakers soon became Bolshevik communications. GC&CS had enjoyed a great scoop for the team by acquiring Ernst Fetterlein, formerly the Tsar’s leading codebreaker, and Dilly must have enjoyed working with such a character. Another of Dilly’s colleagues, who also worked closely with Fetterlein, recalled that the Russian was a brilliant codebreaker.
Fetty, as we addressed him, would arrive precisely at 9.30 and read his Times until 10 when he would adjust a pair of thick-lensed glasses and look to us expecting work to be given to him. On book cipher and anything where insight was vital, he was quite the best. He was a fine linguist and he would usually get an answer no matter the language. When he deposited his first cheque at a London bank he was asked for his references, to which he replied: ‘Pardon me? It’s my money. Where are your references?’
When Winston Churchill became Chancellor of the Exchequer in 1924, he wanted to catch up with the backlog of Soviet intercepts. He told the new Foreign Secretary, Sir Austen Chamberlain, that ‘I have studied this information over a longer period and more attentively than probably any other minister has done’. Speaking as one who had an intimate knowledge of Room 40’s codebreaking and the intelligence derived from the messages, he added: ‘I attach more importance to them as a means of forming a true judgement of public policy in these spheres than to any other source of knowledge at the disposal of the state.’ Signals intelligence, or Sigint, as the produce of the codebreakers’ work would become known, was a lifelong passion of Churchill’s, which would serve Bletchley Park well during the Second World War.
At no time did the Knox family, apart of course from Olive, have any idea what Dilly did in the office. Dilly’s niece Penelope Fitzgerald said the family thought his work might have something to do with the threat from Moscow, with anti-Bolshevik fever gripping the nation, especially after the publication in 1924 of the Zinoviev letter. The letter, supposedly written to the Communist Party of Great Britain by Grigory Zinoviev, the president of the Comintern, called for the mobilisation of sympathetic forces in the Labour Party and subversion of the British armed forces. It confirmed the widely held belief at the time that the Labour Party was soft on communism and in the middle of the 1924 election campaign was leaked to the Daily Mail by members and former members of the intelligence services, with both Sinclair and ‘Blinker’ Hall believed to have been involved. The newspaper put it on the front page with banner headlines, ‘Civil War Plot By Soviet Masters, Moscow Orders To Our Reds’, and Labour lost the election, although this was already a foregone conclusion. The General Strike in 1926 had also been seen by many to be part of the great Soviet plot to take over Britain.
&nb
sp; Dilly’s family were right to see a connection. In 1926, GC&CS received a new source of telegrams through the Peking post office, which must have pleased Dilly; a whole network of Russian intrigue was involved and these messages were also broken after discovering super-encipherments of the codebook that proved to be political dynamite. The Knox family knew that something had happened, as Dilly bought a new Burberry overcoat and took them all out to dinner at the Spread Eagle in Thame in January 1927; everybody knew that a celebration meant success in an enterprise of some kind, even though Dilly could not tell them what it was about. GC&CS had obtained the cables sent to the Soviet trade delegation based at the All-Russian Co-operative Society (ARCOS), which had been set up in London as a cover for their subversive activities. Dilly and his colleagues were able to read them, leading in May 1927 to a raid on the ARCOS premises and the decision to break off relations with Moscow. However, as it turned out, there was no cause for a celebration as, in order to justify the break with Moscow, the Prime Minister, Stanley Baldwin, and Chamberlain read out the text of some of the diplomatic telegrams in Parliament and the Russians lost no time in abandoning the cipher they had been using and substituting an unbreakable one-time pad system. Needless to say GC&CS was furious.
Dilly now concentrated on some lesser traffic from Austria, Hungary and the Balkans. His colleague Joshua Cooper, who had joined the section in 1925, felt that Dilly himself thought the intelligence derived from their work at that time was of little consequence. Cooper recalled:
Dilly Page 7