The Bletchley Park Codebreakers

by Michael Smith

  Dilly was ready with a poem to celebrate the occasion. Each verse began ‘When Cunningham won at Matapan by the grace of God and …’, mentioning all his girls with a rhyming tribute; the rhyme for Mavis conveniently being the flattering ‘rara avis’. All very heady stuff for a nineteen-year-old. Dilly was considered by some to see cryptography only as a theoretical problem unrelated to real events; but that was a mistake. Undoubtedly aerial reconnaissance played an important part in allowing Cunningham to draw up his battle plan, but it was Dilly who rang the Admiralty immediately to make sure that when the battle was reported in the press its success should be accredited entirely to reconnaissance to cover the real source. He chuckled in his poem when he referred to the reconnaissance ‘cottage aeroplane’. Clare as usual had organized shifts and kept our spirits up and so was given a special tribute.

  When Cunningham won at Matapan by the grace of God and Clare

  For she pilots well the aeroplane that spotted their fleet from the air.

  Tribute was also paid to Margaret whose aim ‘straddled the target’ and, amongst others, to Hilda who ‘sank the Vittorio Veneto – or at least they can’t rebuild her’.

  In 1974, The Ultra Secret by Frederick Winterbotham appeared with a picture of Bletchley Park on the front. All those who had kept the secret for thirty years were amazed. Could we now tell the family why we were so good at anagrams, Scrabble and crossword puzzles? But what was this ‘Ultra’ the jacket cover told us went on at Bletchley? We could never remember using the word. Dilly’s ‘girls’, some of them now grandmothers, eagerly looked in the index for some mention, but not a squeak about the Cottage, and Matapan was credited to a Luftwaffe break in Hut 6, as Winterbotham had only known about air force intelligence. What about the rods and the drink with Cunningham? Maybe Dilly’s mentor, Lewis Carroll, had been right after all and life was but a dream.

  Fortunately, it was the Italians who came to our rescue and proved it was real after all. The Germans had always accused the Italians of having traitors in their midst, which was made worse when, in 1966, H. Montgomery Hyde published the story of the beautiful spy, Cynthia, who had seduced Admiral Lais, the naval attaché in the Italian Embassy in Washington, and obtained the codebook from him which resulted in the Italian defeat at Matapan. As one reviewer observed, ‘treason in bed and death at sea made a libretto which sold well’, and the Admiral’s family felt obliged to take out a libel action, such a course being permitted in Italy on behalf of the dead. Montgomery Hyde was found guilty but the real evidence was not then available.

  When The Ultra Secret was published, however, the Italians were delighted as it paid off an old score by proving that it was the Germans and not the Italians who were the culprits. They wanted to have Winterbotham’s book translated into Italian at once. However, Dr Giulio Divita, who was asked to edit the book, was determined to investigate the matter more fully and, when records were released in 1978, he found ample evidence that it was in fact decoded Italian and not German messages that had given the game away. The BBC ran its Spy! series in 1980 with an accompanying book, rehashing the Cynthia story of seduction and treason, now that it was official that breaking Italian ciphers had been responsible for Matapan; it was she after all who sent the codebooks to the Admiralty, which allowed the signals to be read. Dr Divita wrote to The Times about this misrepresentation of the facts, having by then decided to end the matter by tracking down the Bletchley Cynthia, or at least one of them. This wasn’t difficult as Ronald Lewin had already mentioned names in his book Ultra Goes to War, the first book about Bletchley since the release of official records.

  I was able to scupper the idea that we had been given codebooks captured by Cynthia or anybody else; if we had had such books we shouldn’t have needed codebreakers as it would have been child’s play, given that we had a simulated machine. The Italian messages began with a five-figure number to which the operator referred in his list of keys and then inserted the wheels in the given order, adjusted the clips on the side of them and put up the setting in the four windows as directed. Not having such a list, we had to break every message separately by Dilly’s rodding method as has been described. At last the Italians had got what they needed to exonerate poor Admiral Lais. They asked me if they brought the actual Matapan battle messages from Rome whether I would show them how they had been broken individually without a codebook. I warned them that as this all happened forty years ago there wasn’t much chance of my remembering the actual break.

  Nevertheless, losing no time, Dr Divita brought the Admiral in charge of naval history to see me and when I held the message headed SUPERMARINA in my hand it seemed as if time had stood still and I was nineteen again and wearing a green jumper. I spotted the word incrociatore, the Italian for ‘cruiser’, for which we had a chart and remembered that it was a new recruit, Phillida Cross, who had found the clinching click with a lovely starfish and earned herself an honourable mention in Dilly’s poem. The Admiral would have been amazed that a starfish could have sunk a cruiser if I had been able to tell him so. Margaret Rock and I had been able to rod out enough of the rest of the message to ensure that we had the right wheel position. It had been raining all day and was still pelting when I rushed it over to the machine room. Dilly was rung up at home and came in to take charge of any further messages and the ‘girls’ on duty went thankfully home to bed. Cynthia was finally put to bed too; no seduction and no codebooks but just hard cryptographic slogging and a lucky break, or as Dilly put it in his epitaph on Matapan to Mussolini:

  These have knelled your fall and ruin, but your ears were far away

  English lassies rustling papers through the sodden Bletchley day.

  My new Admiral friend drank the health of ‘Dilly and his girls’, just as Admiral Cunningham had done in the Cottage forty years before. On his return to Supermarina, the Admiral wrote a charming letter of thanks, but, with the lesson of Cynthia still in mind, he ended on a cautious note: ‘Hoping to be given the opportunity of meeting you again, on work matters.’

  There has been a tendency of late to relegate Dillwyn Knox to the status of an old-time professor who did brilliant work on ciphers in Room 40 but was defeated by Enigma, the credit for the breaking of which must go entirely to Turing and Welchman. It is to be hoped that this book will go some way to redressing the balance and that when Bletchley Park’s restoration plan goes ahead, the Cottage will be dedicated to his memory. ‘Nobby Clarke’, an old friend of Dilly’s since Room 40 days, who was in overall command of the naval section at Bletchley Park at the time of Matapan, summed up the situation when he added a verse to Dilly’s Matapan poem.

  When Cunningham won at Matapan

  By the grace of God and Dilly,

  He was the brains behind them all

  And should ne’er be forgotten. Will he?

  7

  A BIOGRAPHICAL FRAGMENT: 1942–5

  JOHN CHADWICK

  Introduction

  In Chapter 7, the late John Chadwick relates his experiences in breaking several Italian naval codes, while serving in the Royal Navy in the Mediterranean. His account is all the more remarkable, since he had no formal training whatsoever in cryptanalysis, and knew no Italian - only Latin. However, his later work at Bletchley as a Japanese translator, which he also describes, was probably even more difficult than codebreaking.

  Chadwick attended one of the six-month courses introduced in 1942 by Brigadier John Tiltman in a desperate attempt to increase the number of Japanese interpreters, who were in very short supply throughout the war. The London School of Oriental and African Studies had advised that no course could produce useful results in less than two years, but Tiltman’s unorthodox views were completely vindicated by the success of the courses, which were mainly taken by bright young Classics students. Tiltman even achieved considerable popularity at Oxford and Cambridge because he was then the only person in England who wanted their young scholars purely on account of their skills in Greek and Latin.


  Many of the signals translated by John Chadwick’s section at Bletchley Park had been sent by Japanese naval missions abroad using the Coral cipher machine (JNA-20), since no codebook could cope with the technical subjects covered by their reports. The signals sent from Berlin, in particular, were highly technical, since they dealt with all aspects of German wartime scientific research, covering radar sets, infra-red search devices (in which Germany was more advanced than the Allies), jet aircraft (including ‘sweep-back’ on wings), advanced adhesives, and much more. The translations of the technical decrypts from the Japanese Navy’s technical mission in Berlin provide a fascinating overview of German technology and production. They are by no means easy to understand without a good knowledge of the subject being dealt with, yet the translators appear to have lacked any technical training, and the many ambiguities of Japanese words were greatly increased by the phonetic form required for signals being transmitted by radio. The decrypts therefore presented the translators with a most formidable challenge, but they met it fully: the translated decrypts are superb examples of the translators’ art. John Chadwick, who later acquired an international reputation as a philologist, has rightly paid tribute to their vital work, which until now has gone virtually unrecognized.

  John Chadwick later achieved world renown for his most famous feat of codebreaking, which had nothing to do with. Sigint. He collaborated with Michael Ventris in the 1950s to decipher the ancient Minoan script called Linear B. They proved it to be a form of early Greek, contrary to the views of established scholars who had been attempting to solve its mysteries for decades.

  RE

  In May 1942, I was an Able Seaman serving on board HMS Coventry, an anti-aircraft cruiser attached to the Eastern Mediterranean Fleet based in Alexandria. I had been trying in various ways to improve my position, and one of the things I had done was to acquire a reasonable degree of competence in speaking modern Greek. This was a much rarer accomplishment than it is today. I was therefore not unduly surprised when on putting into Alexandria after a long absence I received orders to go ashore and report to an office in the naval base called COIS. No one told me what those letters stood for, but since it bore some resemblance to the Arabic word for ‘good’, I took it for a favourable omen. When I found the place, I discovered that it stood for ‘Chief of the Intelligence Staff’. To my astonishment I was interviewed by the COIS himself (Captain Bousfield, RN), who had with him a middle-aged English civilian. I assumed he was there to test my knowledge of Greek.

  But I was puzzled by the questions I was asked. They seemed quite uninterested in my linguistic ability and Greek was never mentioned; they concentrated on my schooling and my first year as an undergraduate at Cambridge. I had abandoned my course in Classics to join up after the fall of France in 1940. I was even more perplexed when I was asked to demonstrate my ability to write small numerals clearly. However, after a short delay I was given to understand I had passed the test, whatever it was, and I was sent back to my ship with orders for my immediate transfer to HMS Nile, the naval base.

  Two days later I reported for work (at the absurdly late hour of 0900) and was directed to another office, where I met the head of the section I was to join. He had apparently been absent when I was interviewed. His name was Commander ‘Jock’ Murray, RN. As I learned later, he had had as his confidential clerk an English lady who was the wife of the Fleet Engineer Officer. Murray was well known for his violent temper, and he had been sufficiently provoked to call this lady a liar, whereupon she had walked out. It occurred to the COIS that if he had a rating as his clerk, he could call him any name he chose without creating another incident.

  He gave me a long lecture on security, and explained that I must keep what I was doing secret from my messmates, not an easy thing on the lower deck. When he was satisfied I had grasped its importance, he calmly told me that his section was trying to decipher Italian naval signals. ‘You say you know Latin,’ he remarked, ‘so you shouldn’t have much difficulty with Italian.’ He then produced a large pile of old intercepts, an Italian code-book (the Cifrario Mengarini) and some recipherment tables. He demonstrated how to use them and left me to decode the remainder. After making a few mistakes I soon learned how to do this, although of course I could not understand all the telegraphic Italian that emerged. There was no sort of Italian dictionary in the office, not even the unilingual one we later acquired.

  After a few days I was initiated into the art of indexing (hence the need for small numerals), building a ‘depth’ and eventually reconstructing the recipherment key. This was of course much more interesting and I soon picked up enough Italian to follow what was coming out. But before I had gone very far, the news from the Western Desert became alarming; the British forces had been driven out of Cyrenaica and Rommel was invading Egypt. The Army announced its intention of holding him at a place no one had ever heard of, called El Alamein. The Navy promptly ordered the evacuation of Alexandria, and overnight not only did the fleet sail, but all the auxiliary and depot ships too; we awoke next morning to find the harbour empty except for a single minesweeper.

  There were two civilians working for the Commander, the one I had met at my interview, W. Stanley Backhouse, and a lady, Pauline Smith, who was the wife of an army officer. At this stage I had no idea what their duties were. When the panic started we were ordered to pack up our work in sacks labelled with innocuous tags like ‘Baccy’ meaning Backhouse, and to stand by to move at a moment’s notice. Next day the order came to destroy all confidential documents, so the sacks were unpacked again and I helped incinerate thousands of intercepts and indexes made from them. I quite expected to be abandoned to take part in the defence of the naval base while my superiors got away, but in fact I was ordered to accompany them on a journey by train to Cairo. We were taken to Heliopolis Museum, the main outpost of Bletchley Park in the Middle East. My arrival created administrative chaos, since I was a lone naval rating attached to an Army Intelligence Corps unit, itself attached to an RAF station. There was no work to do, but we had not long to wait before we were moved again. We had the impression that the German Air Force were aware of our location; one night they dropped a stick of bombs which straddled the building, fortunately with little damage.

  The ‘I’ Corps unit to which I had been attached received orders to leave and took me with them; we travelled by train to Ismailia on the Suez Canal, were ferried across, and then began a most uncomfortable journey in a train of covered goods wagons across the Eastern Desert. When we reached Gaza, in what was then called Palestine, we were taken to a tented camp. Conditions there were quite tolerable since it was summer, though we had nothing to do; and in little more than a fortnight we repeated the journey in reverse, but this time in slightly more comfort. From Cairo I was sent back to Alexandria by train.

  The unit now reassembled and had to start afresh with only the latest intercepts to work on, together with the information stored in the memories of the workers. We were also disrupted by the move of the naval base to new quarters in a suburb of Alexandria, about five miles to the east. A luxurious villa (the Villa Laurens) was taken over for the main HQ, and various adjacent buildings were included in a secure perimeter for offices. We were allocated a small bungalow. We had not been there long before Commander Murray announced that he was leaving, for what reason I never learned. Mr Backhouse was left in charge and proved a much easier person to work for. He had, before the war, been a shipping agent in Naples, so he spoke the language fluently and I learned a lot from him. Our duties were now reorganized, and I became his assistant, working mainly on a low-grade code we all called ‘Ouzo’, although the name was a risky allusion to its Italian name, Cifrario per Uso di Mare. I was to spend a long time on this, so it may be worth describing in some detail.

  The signals sent in this code were easily recognizable, since each group consisted of a letter and three figures. The letter normally stood at the beginning of the group, but occasionally occupied other positions.
Mr Backhouse had worked on it for some time before the evacuation and established that it was un-reciphered, but cleverly designed to make penetration difficult without a very large amount of traffic. It was divided into twelve sections or ‘pages’, each of which was indicated by the letter in the group; there were two letters for each page, to be varied arbitrarily, and from time to time these were shuffled. This change caused us little trouble, once we had grasped the procedure.

  On each ‘page’, each word in the book was supplied with ten three-figure code groups, one beginning with each digit from 9, 8, 7, and so on to 0. Thus in theory the encoder had a choice of twenty code groups for each word, but since it was easy to pair the letters, in practice we had ten codes to break simultaneously. We were able to reconstruct the arrangement of the columns of figures on the page, since, human nature being what it is, the 900 column was the one most frequently used, and the frequency of each column fell off regularly as the columns stood further and further away from the list of words. Only the 000 column showed a slight rise in frequency.