Gordon Welchman

by Joel Greenberg

  In his 1987 book Official Secrets, David Hooper quoted from the letter Welchman received from Sir Peter Marychurch, the Director of GCHQ, dated 12 July 1985. Stuart Milner-Barry waited until 1993 before finally writing publicly about his work at BP and something of Welchman’s contribution. He had previously authored a powerful obituary following Welchman’s death and written to the press about his outrageous treatment by the intelligence services in the UK and US.

  When we first came to GC&CS, Bletchley Park was a tiny organization, probably not more than thirty strong. It consisted of a few old-time professionals who had worked in Room 40 at the Admiralty in the First World War, such for example, as Dillwyn Knox, a Fellow of King’s who died during the war, and A. G. Denniston, and new recruits such as Welchman and Alan Turing. Knox, had, so I understood, been defeated by the Enigma and subsequently exploiting its success, should (subject to the Poles) probably go to the other two.

  Turing was a strange and ultimately a tragic figure. But as an admirable biography of him has been written, I shall say no more here. Welchman, on the other hand has, I think, never received his just deserts, quite apart from being ridiculously persecuted on security grounds for revealing, some forty years after the event, how the job of breaking the Enigma had been done. Welchman was a visionary, and a very practical visionary at that. In spite of Knox’s failure, he always believed that Enigma could be broken. He also realized the enormous importance of the success, and took it for granted that, when the phoney war ended, the Germans would rely principally on the Enigma for their military communications. He foresaw much of what would be involved in the way of expansion of staff, machinery (the bombes), and all the other necessary substructure. And he had the fire in his belly that enabled him to cajole higher authority into supplying our wants. If Gordon Welchman had not been there, I doubt if Ultra would have played the part that it undoubtedly did in shortening the war.4

  So what lay behind Welchman’s harsh treatment by GCHQ and the NSA? In 1981, Chapman Pincher had published a book in which he publicized, for the first time, suspicions that a former Director-General of MI 5, Roger Hollis, had been a spy for the Soviet Union. He had also described MI 5’s and MI 6’s internal inquiries into the matter. Its publication even prompted the British Prime Minister, Margaret Thatcher, to give a statement about it in the House of Commons.5 She announced that she had asked the Security Commission ‘To review the security procedures and practices currently followed in the public service and to consider what, if any, changes are required.’ She went on to say that ‘My concern is with the present and with the future.’ By 1982, Nigel West had already published two books about MI 5. In the USA, the NSA had been trying to stop James Bamford from publishing his book about their organization. The material in The Hut Six Story was hardly the threat to security that an exposé of the NSA and MI 5 might be. Yet the FBI felt compelled to interview Welchman several times, tap his phone and place agents in an unmarked car outside his house in Newburyport.

  A recently declassified government document has shed some light on the matter. In a note to Prime Minister Thatcher, dated 30 April 1982, the Secretary of the Cabinet, Sir Robert Armstrong, informed her of the publication of Welchman’s book in the USA and its imminent publication in the UK. He also summarized the security and legal issues raised by its release. With regard to security, Armstrong stated that while there was no direct damage to UK-USA sigint co-operation, there was indirect damage. His carefully worded rationale for this was that:

  so long as Sigint and cryptanalysis are kept in the public eye by books of this sort, foreign COMSEC organizations will receive more funds and will be stimulated to greater efforts.6

  With regard to the legal position, Armstrong stated that the US Justice Department did not think it was feasible to prosecute Welchman and that any legal action to stop publication in the UK offered no guarantee of success.

  It seems that Welchman was simply collateral damage at a time when both GCHQ and the NSA were trying to restrict the information being published about themselves and related government agencies. He was a 76-year-old veteran of BP largely unknown to all but former wartime colleagues. Pressuring him might, at the very least, dissuade former colleagues from publishing their own accounts of BP’s wartime work. However, it didn’t prevent Peter Wright from publishing his insider’s account of MI 5 in 1987. Pincher and West both went on to publish further books about MI 5 and MI 6. Sir Peter Marychurch may well have been urged by higher authorities to write his distasteful letter to Welchman in 1985. It is interesting to note that Marychurch was only the second Director of GCHQ not to have worked at Bletchley Park during the war.

  By the new millennium, things had changed quite dramatically. As noted above, Christopher Andrew had arranged to publish Welchman’s last piece of work, ‘From Polish Bomba to British Bombe’. Welchman had also entrusted him with telling the story of his treatment by the intelligence services. Andrew had been outraged by the story but had managed to get D-Notice Committee to approve the paper before its publication. In February 2003, Andrew accepted the post of official historian for the Security Service, MI 5. He was asked to write an official history of the service due for its centenary in 2009. At the end of April 2002, James Bamford was writing his second book about the NSA, Body of Secrets. According to Bamford, NSA Director Michael Hayden gave him unprecedented access to the NSA campus in Fort Meade, MD, senior NSA officials and thousands of NSA documents.

  Throughout his life, Welchman had been strongly influenced by religion and music. He cut a dashing a figure as a young man, excelling at sport, and he certainly had an eye for the ladies. However, he greatly valued the teachings of the Church of England and greatly admired his father. He was also very patriotic and his son Nick remembered him being moved to tears when told of the death of King George VI in 1952. He was also very loyal to friends and former colleagues and tended not to talk about people unless he had something good to say about them. A friend remembers him being very upset and quite emotional after he learned about Turing’s suicide. Apart from recognizing his genius, he remembered Turing as ‘a completely gentle soul’.

  Music was a constant theme throughout his life and he used to sing madrigals together with his first wife Katharine while they lived in England. They continued to do so when they moved to the USA and on one occasion even included their son Nick, who was a choir boy at the time, in a production of Three Kings. A friend recalled him having ‘a gorgeous tenor voice’. His stepson Michael remembers him sitting in his music room with headphones on, listening to classical music and conducting along with it. His musical tastes were quite wide and he loved the ragtime music of Scott Joplin. He was always open to new types of music and, on one occasion, introduced Michael to records by Simon and Garfunkel. Michael in turn played him a record by his favourite band. Welchman liked the first album that Michael played him, although not the second one. The fact that he liked the first one was surprising enough as the band was none other than the famous West Coast hippie rock group, Grateful Dead.

  Ros Welchman and her family spent the summer of 1985 in Newburyport, renting a house not far from that of her father. Welchman had experienced an extraordinary number of health problems throughout his life and, by the summer of 1985, it was becoming clear to both his family and himself that even he was unlikely to recover from pancreatic cancer. Ros was able to help Welchman complete ‘From Polish Bomba to British Bombe’, being a mathematician in her own right. He felt very good about finally finishing the paper but still had concerns about sending it to Britain by post. This time, he was determined to get official approval before publication. His close friend Diana Lucy was paying a visit to England and he asked her to deliver the manuscript personally to the publisher. Diana recalls it being all very cloak and dagger and that she travelled in casual clothing with the document stuffed up her jumper. She had a number to call when she got to England to arrange its delivery by hand.

  By now Welchman had to spend most of
his time in bed but he was surrounded by family and friends. He turned to an elderly homeopath for treatment as it was clear that his doctors could do little for him. At last his prodigious research and writing had come to an end and his working life was over. Now he could devote all of his attention to his family, particularly his grandchildren.

  Gordon Welchman did not live to see some of his story told by authors such as West and Hooper or by former colleagues such as Milner-Barry. He passed away on 8 October 1985 at the age of seventy-nine at Anna Jaques Hospital in Newburyport, Massachusetts. His funeral was held at Linwood Crematory in Haverhill, Massachusetts, on 12 October. Before his death, Welchman and his wife had made enquiries about holding his funeral in Bristol Cathedral but the cost was exorbitant. As the son of a former archdeacon of Bristol, Welchman was quite annoyed by this and would not let the family ‘waste’ money on a funeral there. The British Government was represented by its consul based in Boston. When he heard about the final years of Welchman’s life, he asked, much to the bemusement of the family, why they had not come to him when the problems with the intelligence agencies began. Teeny chose all German music for the funeral initially but then changed her mind. On reflection, she chose the English music that Welchman loved such as Vaughan Williams’s ‘The Lark Ascending’.

  On 12 November 1985, McGraw-Hill signed over rights to The Hut Six Story to Teeny Welchman while retaining the right to sell or remainder existing inventory. In January 1986, his final paper was published in Intelligence and National Security. When The Hut Six Story was republished in 1997, the publisher, with the agreement of the Welchman family, replaced the last section with that paper, ‘From Polish Bomba to British Bombe’.

  At his request and somewhat ironically, the ashes of a man who, along with Alan Turing and a few others, had made one of the most significant individual contributions to the Allied victory in the Second World War, were interred in his wife’s family plot in Aschau cemetery in the province of Chiemgau, Germany.

  Appendices

  The Enigma machine.

  (Diagram by Dustin Barrett and Jack Copeland)

  Appendix 1

  A Beginner’s Guide to Enigma and the Bombe

  When he arrived at BP on 4 September 1939, Gordon Welchman was initially assigned to Dilly Knox’s team in The Cottage. He wrote years later that he didn’t learn much there about the Enigma machine but it didn’t matter because any reasonably intelligent person could learn all they needed to know about it in an hour or two!

  Enigma

  The Enigma machine was used to encrypt and decrypt a message. Sending and receiving the message was a separate process. It could run off a battery as well as a mains supply which made it ideal for mobile communications. The machine had a keyboard with 26 keys, one for each letter of the alphabet. laid out in three rows; there were no number or punctuation keys. Above the keyboard was the lampboard which consisted of 26 small circles of glass, each embossed with a letter of the alphabet and laid out in the same pattern as the keyboard. Underneath each circle of glass was a small bulb, much like one in a modern torch. Each time a key was pressed, one of the bulbs switched on and the letter that was illuminated above it was the encryption of the letter on the key that had been pressed. So if the O key was pressed and the bulb under Q switched on, then Q was the encryption of O (see opposite).

  Inside the standard German military machine were three wheels with 26 contact points on each side wired together internally with strands of wire. At the front of the machine was a plugboard consisting of 26 double sockets also laid out in the same pattern as the keyboard. Pairs of letters were connected with cables having plugs at each end. Around each wheel was a ring with the letters A-Z (Navy Enigma) or the numbers 1–26 (Army/Air Force Enigma) which could be set to any one of its 26 positions. When the wheels were put in the machine and the lid shut, their serrated edges protruded through slots in the lid. Adjacent to each slot was a small glass viewing window and the topmost letter or number on the ring around each wheel could be seen. Each wheel could be turned freely by hand by rotating it to any one of its 26 positions.

  From late 1938, German Army and Air Force Enigma operators were issued with a set of five wheels from which they would choose three. Each wheel was wired differently and they could be put in the machine in any order. So when the wheels were clamped together in the machine, you had in effect, a 26-way circuit running through it. When a key was pressed on the keyboard, a current passed from contact point to contact point through the machine as shown in Figure 1. Each contact point represented a different letter so, as the current passed through the machine, the letter on the key pressed could change up to nine times before it reached the lamp-board. To the left of the three wheels was a fixed wheel called the reflector which had 26 contact points connected in 13 pairs. The other ingenious design feature of the machine was that when a key was pressed at least one wheel would move forward one position. The right-hand ‘fast’ wheel moved at every key stroke while the movement of the middle and left-hand wheels was determined by the ring settings. The result of this was that the circuit through the machine changed every time a key was pressed.

  The starting configuration of the machine, consisting of the wheel choice and order, the ring setting and the pairs of letters plugged together, was known as the key. As the Germans had a number of communication networks, each with its own daily key, Welchman devised a naming convention, based initially on the coloured pencils used to mark them up on lists. The main German Air Force key was known as Red. Army and Air Force keys usually changed every twenty-four hours and operators were issued with monthly key sheets.

  The Enigma machine was designed to have reciprocal characteristics so that the processes of encryption and decryption were essentially the same. This meant that if O was encrypted as Q, then at the same starting position, Q would be encrypted as O. The machine achieved this by always partitioning the 26 letters of the alphabet into thirteen reciprocal pairs of letters so that for any of these pairs, one letter would always be encrypted as its partner. When the wheels moved to another set of positions (which happened every time a key was pressed), the circuit configuration in the machine changed along with the identity of the thirteen letter pairs, but the reciprocal relationship between the letters in each pair remained.

  It was the reflector which delivered the reciprocal characteristic of the machine and thus also guaranteed that a letter could not be encrypted as itself. This was a weakness that the Germans were apparently prepared to accept as it provided the machine with the extremely useful property that the encrypting and decrypting processes were the same and one machine could do both tasks. This would have greatly simplified the training of their Enigma operators.

  Once the machine had been set up using the daily key, a final layer of security was implemented. The procedures for this changed several times but, from early May 1940, the Air Force and Army operators were told to choose 6 letters of the alphabet (known as the indicator) at random. The machine would be set to the daily key and then the wheels turned until the first three letters (known as the indicator setting) appeared in the windows adjacent to the edge of each wheel. The next three letters (known as the message setting) would then be encrypted. Finally, the wheels would be turned to the message setting and the plain text of the message encrypted. Both the indicator setting and encrypted message setting would be sent along with other information before the encrypted message. The settings of the machine provided by the key sheet combined with the message setting, defined the starting position of the machine.

  The receiving operators would set the wheels of their machines to the indicator setting and then recover the message setting by entering the encrypted message setting letters. They would then turn the wheels to that setting and recover the original plain text by simply entering the encrypted message. The whole encryption and decryption process assumed that both the sender’s and the recipient’s machines had been previously configured to the daily key.
<
br />   The Bombe and the Diagonal Board

  A more detailed description of how the bombe was used at BP is given in Appendix 2. It has been written by Frank Carter, a leading expert in the field, and includes details which have not previous been published in a mainstream book of this type.

  To start, it is worth restating the problem faced by Turing and Welchman as they designed the bombe. In order to decrypt a message sent on a particular German communication network on a particular day, they needed to know the daily key and the message setting for each message. The daily key provided the following settings:

  • The choice and location of the three wheels in the machine. There were 10 ways of selecting three wheels from five (in the case of the Army and Air Force) and six ways of arranging the three wheels in the machine. The total possible number of wheel orders was therefore 10 × 6 = 60. (three wheels from eight in the case of the Navy yielded 56 × 6 = 336)

  • The identity of the set of 10 ‘plug pairs’ typically used on the plugboard; the total possible number of combinations was 150.7 million.

  • The wheel turnover positions which were determined by the ring settings. The ring around each wheel had 1–26 (Army/Air Force) or A–Z (Navy) embossed on it and could be set to any one of the 26 positions. Thus there were 26 × 26 × 26 = 17,576 possible ring settings. More details of the effect of multiple wheel turnovers are given in Appendix 2.

  The number of daily keys was therefore 60 × 150.7 million, million × 17,576 = 158.9 million, million, million. As the number of possible message settings was also 26 × 26 × 26 = 17,576, the number of possible ways of electrically configuring the Enigma machine before encrypting a message was the same extraordinarily large number.