by Alastair Denniston- Code-breaking From Room 40 to Berkeley Street
18502 hinweisen point to
18500 comma(,) comma(,)
15857 dass that
2188 rücksichtslos ruthless
5376 Anwendung employment
7381 unserer our
98092 U-boote U-boats
16127 jetzt now
13486 Aussicht prospect
9350 bietet offers
9220 comma(,) comma(,)
76036 England England
14219 in in
5144 wenigen few
2831 Monat- month-
17920 en s
11347 zum to
17142 Frieden peace
11264 zu be
7667 zwingen compelled
7762 stop(.) stop(.)
15099 Empfang Receipt
9110 bestahigen acknowledge
10482 stop(.) stop(.)
97556 Zimmermann Zimmermann
3569 stop(.) stop(.)
3670 Schluss der Depesche BERNSTORFF End of despatch
Appendix 4
Examples of Room 40 Decrypts with AGD’s Initials
Crown ©. Reproduced by kind permission of Director GCHQ
Appendix 5
How News was Brought from Warsaw at the end of July 1939
This is AGD’s personal account (and the only British first-hand account) of the famous July 1939 meeting between British, French and Polish intelligence personnel. It was written on 11 May 1948 from memory and using his pocket diary to check dates (TNA HW 25/12):
Before attempting from memory a description of the visit of July 24– 27 it is necessary to outline previous events which led up to the visit. The hero or mystery man or deus ex machine was undoubtedly the French officer Bertrand alias Godefroi. I never really grasped how G.C. & C.S. came to be involved in this liaison.
From 1937 onward it was obviously desirable that our naval, military and air intelligence should get in very close touch with their French colleagues for political and military reasons. The Admiral had always wished and worked for a close liaison between SIS and G.C. & C.S. but I have always thought that Dunderdale, then in Paris, was the man who brought Bertrand into the English organisations. Menzies it is true, had a close liaison with Rivet under whom Bertrand worked but I think it was Dunderdale who, entirely ignorant of the method of cryptographers, urged the liaison on the technical level. Bertrand was his man in the French military branch. The French had no interservice organisation. In fact all four (including Quai d’Orsay) thoroughly disliked and distrusted each other. Bertrand was no cryptographer and never pretended that he was. Any technical staff under him must have been very second rate but he had a genius for making use of others. By July 1939 he certainly was obtaining cryptographic results from us and from the Poles and he also salvaged five Spanish republicans who had worked at Barcelona & installed them in Paris as his party working on Spanish and Italian material.
Bertrand was, no doubt inter alia, a peddler & purchaser of foreign government codes and I think that, as he never had an opportunity in his own organisation of obtaining results from crypt he had decided to concentrate on careful purchase as the surest method. I would say he did not have access to very much cash and frequently asked us to go 50-50 in a possible deal. We were averse to this as we already read most telegrams except the Russian and German neither of whom was among Bertrand’s customers. But the Admiral thought that occasionally it was well spent money to keep in with Bertrand. He must have been useful to SIS in other ways as I well remember an official lunch in ’38 or ’39 at which Menzies presented Bertrand with a gold cigarette case engraved with the royal arms.
It should be noted as of some ultimate importance that in the course of some of his visits to us in Broadway we arranged for Bertrand to have conferences with certain sections such as the Spanish and Italian. He heard explanations of our methods and saw results and with the Admiral’s permission received a regular supply of certain results.
I think he was impressed by our success and by the ability of certain officers and it might well be that this had an important bearing on his subsequent actions.
The Spanish war had rekindled our interest in the German Enigma as, at last, we were intercepting a considerable volume of German Naval material and a little Military or Air traffic. Knox was working on the naval material with intermittent assistance from Foss. Tiltman was in the background but was really fully occupied with the military section and the solution of the Japanese military cyphers. In addition to this we had found that the Italian Navy were using the commercial Enigma. This was cracked by Knox and a small section was formed with Bodsworth in virtual charge to handle the current traffic, leaving Knox clear for the German machine.
So far as I can remember this was the position in January 1939 when Bertrand asked me to come over to Paris as he wanted us to meet certain Polish experts. We assumed that such a meeting could only concern German & Russian work on both of which we were so weak that we felt we might well profit by such a meeting. So I went over with (I think) Tiltman, Knox and Foss. We met in an atmosphere of secrecy and mystery. One of the Poles, a Major Ceski [sic], gave us the alleged results of their research into the Enigma machine as used by the German services. He had no English and his French and German were only fair but it was clear that he was describing the results of his work on the ‘message setting’ which at that period appeared as the first group in all messages in all services. (It was only in May ‘39 that the German Navy adopted the bigrammatic indicator which Knox quickly diagnosed & found himself to drop the Naval & concentrate on the Military/Air). The Poles gave us a long and pedantic description of what Tiltman had spotted and learnt in an hour namely that the message setting was soluble because no letter could by cyphered by itself and the German operators were simple souls with childish habits. The Poles finished by saying that they were obtaining a lot of material and pursuing their research and it would keep us informed of any results. At that time we had to regard the meeting as a flop and that the Poles were no great find. Since then I have thought that it was our party who were being vetted by the Poles who were at the time reading the traffic and only wanted to know what progress we had made. Bertrand himself let it be understood on this occasion and always when the question was raised that he himself was ignorant of the ability of the Poles, or of what they could read. He was aware, he said, that they obtained some results. Later he admitted that he had to visit Warsaw once a month at least to exchange raw material and results and further he considered them exceptionally well-placed to do ‘business’, as he understood it, with treacherous Germans & Russians.
For the next six months we heard nothing more from the Poles probably because they felt that they did not require any help from us. In the middle of July however we received an invitation, through Bertrand, to pay them a visit in Warsaw. The international situation was now tense. England was pledged to go to the assistance of Poland if attacked by Hitler. Each party knew that the other had an active cryptographic organisation. So it appeared quite natural that the Poles should invite us. But it must not be forgotten that the German Navy had changed in May. The Polish invitation specifically included Knox who was known both by the French and Poles to be working on this subject. I do not think though now I cannot be sure that they were aware of our success on the commercial Enigma as used by the Italian Navy.
However, the Admiral instructed me to take Knox with me. I could well wish now that I had added others of our party but everyone was fully occupied and could ill spare the time to travel to another flop as in January. D.N.I., in close touch with the Admiral, suggested that Sandwith should go also, why I don’t know as Sandwith was not a cryptographer and the Poles had no navy. Bertrand was of course of the party, Knox and I went by train as we wished to see Germany probably for the last time, the other two separately by air. We left on the 24th & were met by the Poles & Bertrand and lodged at The Bristol. We were there for work on the 26th & 27th & leaving on the 28th. I was back in London by Sunday the 30th. Knox, whose passport had be
en wrongly stamped for his return transit through Germany, had to go back from the frontier to Posen and get the visa from the British Consul in that town. The 26th (Wednesday) was THE day. The Poles called for us at 7 am and we were driven out to a clearing in a forest about 20 kilometres from Warsaw. Here we found a new, strongly built and strongly guarded office with some underground accommodation and here we met the Polish cryptographers some of whom luckily spoke some French or German.
A prolonged full dress conference with the Polish senior officer in the chair (I have forgotten his name). Colonel Langer was in charge of the organisation in the Forest. Major Ceski with the stecker Enigma now described in full detail the steps they had taken to break down the cypher and obtain the wheel order and to read the messages; we followed him each to the best of our ability. Knox, as our expert, was alongside Ceski and in the best position to follow his explanation. He, however, reacted very badly to the explanation which took about three hours with a break for a cup of tea. I confess I was unable to understand completely the lines of reasoning but when, as [the] second part of the conference, we were taken down to an underground room full of electronic equipment & introduced to the ‘bombs’ I did then grasp the results of their reasoning and their method of solving the daily key. Knox accompanied us throughout but maintained a stony silence and was obviously extremely angry about something.
It was only when we got back into a car to drive away that he suddenly let himself go and assuming that no one understood any English, raged and raved that they were lying to us now as in Paris. The whole thing was a fraud he kept on repeating. They never worked it out. They pinched it years ago and have followed developments as anyone could but they must have bought it or pinched it.
Our position became increasingly difficult as even Bertrand, who knew no English, was aware that Knox had a grudge against the Poles, who, so far as Bertrand knew, had only been successful where Knox had failed.
I assume that he did not believe that they had constructed the order on the wheels and he may have been correct in his surmise. Ceski’s explanation was very lengthy and involved and in a language which he knew only fairly well and I was not competent to judge of the possibility of complete success. So far as I was concerned they were reading messages up to the previous May and were now quite sanguine as to the reading of current messages again by means of their ‘bombs’. I have since thought that they were finding considerable difficulty especially with the naval and knowing Knox’s reputation and ability, felt it was a good moment to come clean and gain his assistance especially in view of the political situation.
The rest of the day remains a nightmare to me. Knox remained aloof and alone. Bertrand, Sandwith and I discussed the situation at length and decided to get away as soon as possible. The next day had been allotted to personal contacts with their technical and intelligence officers. Knox had cooled down considerably and spent a long morning with their technical staff examining the machine and their methods. Language was of course still a difficulty but he now seemed to understand their reasoning and in his conversations with me never referred to his outburst after the disclosures.
I was shown the mass of telegrams they had read, the naval traffic being practically only between Berlin and their fleet in Spanish waters, which formed the bulk of their material. They undertook to send us copies and a model of the machine through Bertrand as we, of course, had no line of communication with them. The machine did arrive during August but I doubt if any messages arrived. As the situation was now clearly war and service sections had already gone to Bletchley we were not interested in back material but were naturally concentrating on current air/army traffic.
I should add that on the second day Knox was really his old bright self and won the hearts and admiration of the young men with whom he was in touch.
If only that first day of formal disclosure could have been avoided and pompous declarations by senior officers had been omitted, Knox’s mind and personality in touch with men who really knew their job would have made that visit a very real success. They were all simple and straightforward.
To me Bertrand’s attitude remains a mystery. I still believe that he knew all about their work and feigned the surprise which he manifested at the Polish success.
For a full set of notes to AGD’s document, see Erskine’s paper.
Appendix 6
Approximate Strength of GC&CS on Move to War Station, August 1939
PLUS: Secretariat, Communications Section (Teleprinters, Telephones, etc.), Typing Staff, etc
APPROXIMATE TOTAL: 200
Appendix 7
Naval Sigint in the UK, December 1940
Appendix 8
Military Sigint in the UK December 1940
Appendix 9
Air Force Sigint in the UK December 1940
Appendix 10
GC&CS Diplomatic and Commercial Sections (Civil) Structure in 1944
Appendix 11
The McCormack Report
Colonel Alfred McCormack was deputy head of the US Army’s special branch which supervised signal intelligence in the US War Department. He visited GC&CS in 1943 with William Friedman and Lieutenant Telford Taylor, who rose to the rank of colonel in the Office of Strategic Services, and spent most of his time at BP, where he served as a liaison between American and British intelligence. After the war he served as an American war crimes prosecutor at Nuremberg.
The report consists of a series of signals (messages) sent by McCormack and occasionally others, to the US Army Arlington Hall Station Message Centre in Washington via the American Embassy in London. The messages were for the attention of General George Strong, the US Army’s Military Intelligence chief, and Colonel Carter Clarke, Strong’s deputy. They were sent between 21 May and 1 July 1943. The standard signal specific information has been removed for ease of reading and each is a separate section. The actual text of each message has not been edited.
***
Denniston’s show, commonly called Berkeley Street, has none of the hectic atmosphere of Park but rather gives [the] impression of [a] well established operation that goes along through wars and peace. General impression is typified by the two ladies who receive and sort incoming traffic. (They started as telegraph clerks in Post Office in reign of Queen Victoria and were fully familiar with general field when they joined present organisation in 1919.) These little birdlike old ladies receive and register incoming material and they have acquired such great familiarity with it that [they] can do everything except actually decipher it. Whole organization is very simple and they seem to accomplish a great deal with quite limited personnel. Whole outfit consists of two hundred.
38 bodies divided, subject to some doubling up, as follows:
Deputy Director and admin staff – 5
Distribution and reference section – 2 Sorting section – 2
Typing – 14
Teleprinter clerks – 2 Geographical sections – 155 Commercial sections – 50 Research section – 5
Geographical sections are divided into enemy countries, neutral countries and allies.
Personnel in enemy country sections are:
German – 26; Italian and Vatican – 19; Japanese – 36; French (which now overlaps enemy and allied groups) – 14; Siamese – 1 (who in off moment doubles up on Irish); Bulgarian – 2; Roumanian and Hungarian – 1; Finnish is now being worked on in Research Section.
Neutral countries are divided up as follows:
Spanish – 6 bodies; Portuguese – 10; Swiss – 4; Irish – (see above); Latin America, Spanish language – 5; Near East (which covers Turkey, Saudi Arabia, Iraq, Iran, Afghanistan, Egypt, Palestine and Syria) – 17.
All Scandinavia is handled in off moments by Rees who is Chief Administrator Officer.
Allied countries account for full time personnel of 11, of whom eight work on Chinese, two on Belgian and one on Greek. Dutch is done by people who work in Near Eastern section, Brazilian by Portuguese section, Polish, Jugoslav and Czeck
[sic] by the man who handles, with a clerk assistant, Bulgarian material, and Mexican of course is done by Latin American section.
Taylor will report later on composition of commercial subdivisions. Principal sources of material, other than by exchange with United States, are censorships at London, Ottawa, Mauritius, Barbados, Gibraltar and Middle East. (Where there are joint censorships at Cairo, Baghdad, Teheran and possible elsewhere). Foreign Office intercept stations in United Kingdom (Denmark Hill, Sandridge, Saint Albans, Whitchurch, Cupar, and Brora) and, through BP, other intercept stations at Ottawa, New Delhi, West Africa and Mauritius. You will observe what extensive coverage this adds up to. Other source of material is secret intelligence service in neutral capitals. General outline of workings of outfit follows: sorting clerks route material to proper geographical section, where it is registered in appropriate book in which notations are made under following headings:
First serial number,
Second cipher,
Third signature,
Fourth reference (that is, to cipher keys),
Fifth subject (briefly indicated),
Sixth remarks and distribution number (this is inserted if material is distributed, so that by looking at any page you can tell what has been distributed and what not),
Seventh source (that is, intercept station or other point which it was picked up).
Traffic then goes to crypt people and if successfully processed results are passed to Head of Section, who sets aside what he thinks not important enough to give to anybody and the rest to distribution and reference section, which is heart of organization and which I will report on in more detail in another message. This section decides whether material is worth circulating and if so gives each item to be circulated its proper serial number and sends uncirculated items back to geographical sections for filing or destruction. Up to this point everything is done in longhand. Only what is going outside is typed up. Note that commercial material does not go into this section but processes along another routing which Taylor will describe.