Book Read Free

Alastair Denniston

Page 35

by Alastair Denniston- Code-breaking From Room 40 to Berkeley Street


  ***

  Iranian principal diplomatic book has been photographed and was sent today to Baker Peter for transmittal to Washington. Have been discussing smaller enemies and will have further information next week. Reference FLORADORA agreement Berkeley Street has sent off two batches for you and are sending specially about eight Dublin items at my request; they are interested to know how Arlington is coming on its share of back traffic, which is to be interchanged on fortnightly basis enciphered with forty figure subtracter. Washington does not figure subtracter, Washington does not hold ISMET which is perhaps the more secret of the two except at London which treats CANKARA as more secret, Tokyo has neither and uses one of three older books of same type enciphered with five to forty figure subtracter and used except at Tokyo only for administrative traffic. Subtracter changed at frequent intervals. In addition Turks have French book for treaties, obsolete consular book, and military attaché book held at all capitals one part thirty thousand groups enciphered with short subtractor or by switching order of figures within the groups.

  Saudi Arabia intake fifteen to twenty messages per day, circulate two or three per day. Capital is at Riyadh but King nomads around and traffic may center at Taif summer residence, Mecca during pilgrimages, or Jedaah where foreign diplomats are kept. Circuits are to Ankara, London, Baghdad, Cairo, Damascus, Jerusalem and Vichy (now defunct). Principal code is two years old three letter two thousand random groups used unenciphered, also two new books used at Ankara two letter seven hundred fifty random groups. For administrative financial and less secret political use simple two figure cipher.

  Iraq traffic intake four or five per day circulate perhaps one a day. Circuits are between Baghdad and Aleppo, Alexandria, Beyrouth, Bombay, Cairo, Damascus, Jeddah, Jerusalem, London, Kabul, Teheran, Washington and Sanna (defunct). So far this year 21 messages Baghdad to Washington of which British have four, and 34 messages Washington to Baghdad of which British have all but one. Iraq messages are really in clear, they have arbitrary Roman letter equivalents for Arabic alphabet and only element of cryptography is that letter equivalents vary from post to post; also use cover letter combinations for names of people, places, dated etc.

  Iran traffic intake about 1,000 messages per month, circulate about one hundred per month. Circuits covered are between Teheran and Ankara, Baghdad, Baku, Basra, Berne, Beyrouth, Bombay, Bucharest, Cairo, Hamburg, Herat, Istanbul, Jerusalem, Kabul, Karachi, Khanegin, Kuibyshev, London, Mosul, Quetta, Rome, Simla, Smyrna, Soleynanieh, Stockholm, Tokyo, Trebizond, Vichy and Washington. Washington traffic about ten per cent of total. One diplomatic and one military book each alphabetical 12,500.

  Afghanistan intake five or six messages per day circulate about every other day. Circuits are between Kabul and Ankara, Bombay, London, Teheran, Meshed, Berlin, Rome, Geneva, Tokyo and Vichy, traffic with Tokyo very infrequent and purely administrative. One book which closely resembles the Persian.

  I spent several hours reading nineteen forty three output of Near East section. While the Iraq, Saudi Arabia, Afghanistan and Egypt output is picturesque and occasionally interesting I think it is a safe assumption that we can live happily without reading this mail immediately and in any event I will probably be able to send you decodes of anything of interest. Iran for the most part is dull stuff but contains some significant material in view of our present special interests in Iran and I think we should read this as soon as possible. Turkey is definitely an important line; it is disappointing in that messages from Ankara are largely administrative and do not substantially reflect leanings of the government, but traffic to Ankara contains a fair number of very important messages such as interview with Molotov, or with JIG gents in Far East; furthermore a great many messages from smaller European capitals seem quite as important as those of familiar ministers at same points whom we frequently honor with an item in the Bugle. Thacker tells me that since April all Turkish traffic has been sent to Washington, if Arlington has not been getting it please let me know promptly. Subject to your approval I propose to make photographing and mailing to you of important Turkish books matter of first priority and that of Persian books second priority and Chinese books third priority.

  ***

  Friedman and myself have seen Italian section Denniston’s show, Friedman will report on personnel headed by Mr Catty and cryptanalytic features. British coverage includes hosts at Madrid, Lisbon, Tetuan, Tangiers, Tenerife, Lobito, Dublin, Stockholm, Oporto, Seveille [sic], Barcelona, Raima, Helsinki, Bratislava, Vienna, Budapest, Sofia, Bucharest, Belgrade, Athens, Rhodes, Istanbul, Izmir, Adana, Mersina, Kabul, and Sanna now quiescent. According to my recollection our coverage of above circuits has been poor or nonexistent. I am told that since April first they have been sending us copies of all their traffic; if traffic is now reaching you I assume it can be read at Arlington as well as here if not please let me know. Cryptanalytic exchange appears to be going smoothly and no pressing problems in this quarter, however I propose to read back files of decodes on above circuits and send you anything of real interest.

  This section includes two ladies who handle Vatican. Work started on this line April 1942 after lapse of two years in continuity. Up to December first nineteen forty two British had received six hundred sixty four messages in all ciphers and read five hundred fifty or eighty two per cent. Only sixty four were found worth translating and circulating and these are made to be of little interest; I plan to read these as well as later and current traffic and send you if important. Most traffic is ecclesiastical charitable or personnel. Apparently stuff goes by bag. One system in use London, Vichy, Berlin, Madrid, Lisbon and Washington is not yet cracked because of traffic shortage.

  ***

  Herewith some miscellaneous information on Berkeley Street Brazilian and Portuguese sections, Friedman will report in detail on crypt observations.

  Section comprises ten people of whom five are decoders and clerical and five are linguists bookbreakers of whom two do cryptanalysis. No permanent employees. Head of section -ELL is a botanist who has researched on Portugal and colonies. Brazilian largely handled by his wife and one assistant. Enterprise started February nineteen forty at which time all continuity had been lost. As in case of Italian. British have been covering several Portuguese continental circuits inaccessible to us, I will run over these old decodes and send you nuggets. I am told that since April first they have been sending us all Portuguese DIP traffic, but not Portuguese colonial traffic, which can be read here. One basic book for all colonial with variety keys. Am told colonial traffic is most uninteresting but will look it over. Portuguese military use PLAYFAIR at Lisbon and Lourenco Marques, simple substitution at Atlantic islands, British read both kinds. At end of March this year the Portuguese bought some Swedish HAGELIN repeat HAGELIN machines and have them at Air, War and Colonial Ministries, some Azores traffic in Hagelin has been observed. Portuguese security police in Azores at Ponta Delgada are using ten alphabet substitution, I am told British have sent us the keys for this.

  ***

  Herein the smaller European enemies. Hungarian not read and continuity lost but Denniston expecting return of someone who will soon undertake it; however Hungarian commercials in clear are translated by member of Swiss section and circulated if important.

  Paragraph two. Roumanian section consists of Griefenhagen former diplomat and one secretary started nineteen forty having lost continuity and started reading nineteen forty one. Intake zero to twenty messages per day perhaps twelve average. Principal circulation a day perhaps twelve average. Principal circuits covered are to Lisbon, Madrid and Rome, also get some Stockholm, Helsinki, Tokyo and Buenos Aires. Highest grade code not read, is in early stage of reconstruction. Bulk of traffic in two large two part codes called Roger Forty Two and Roger Forty Four, some of the traffic is enciphered. Roumanians follow French practice of numerous and voluminous two part codes. Bulk of traffic which is circulated is commercial in nature, for instance purchases of Spanish lead and blankets and Spa
nish demands for oil in exchange. So far as is known there is no Roumanian exiled government traffic.

  Bulgarian section Woods elderly permanent and one secretary. Continuity since 1934. Small traffic intake circulate perhaps one per week. Principal circuits covered are Berlin, Kuibyshev, Roge, Vichy, Budapest and Bratislava; Tokyo which has new book can’t be read and Stockholm can’t be read. No Bucharest traffic. Coverage on all circuits very poor; for instance between Sofia and Berlin they got only twenty four out of 450 messages. For administrative purposes an old one part code is used unenciphered, pagination changes twice per month. Secret stuff in newer books used unenciphered.

  Woods also handles Jugoslav traffic; diplomatic books not yet read, Jugoslav Army Chief Cairo named Pretnik talks to London and Istanbul in double transposition but British switch the keys. Jugoslav intelligence organization headed by Piritch is housed at Istanbul and is in touch with Mihailovitch, apparently using same system. Three or four Jugoslav messages per day are procured and about seventy five per cent of the intake is circulated. Friedman will I hope supplement crypt data herein. Will look over old decodes of all this stuff. Please advise have sent dozen letters home hope some received.

  ***

  Research section Berkeley Street, four men, two women, headed by Sainsbury. Main job is to take problems about which nothing is known and penetrate far enough so that problem can be turned over to language section for bookbreaking etc, and secondary job is to handle odd sticky problems for language sections. Sainsbury speaks Finnish and his section theoretically handles Finnish Hagelin but in fact is doing little on it; Norwegian Hagelin was dropped long ago and Swedish Hagelin has been dropped as impossible because of excellent Swedish use. In fact Hagelin activity at Berkeley Street appears to be at complete standstill, possibly because best Hagelin man is at ‘Baker Peter’. Main current occupation of this research station is Free French; they appear to have made a considerable penetration of principal diplomatic system, but I was told there has been no interchange of cryptanalytic data on Free French with Arlington as yet. System in question is based on four figure code groups, the first figures are reinciphered on a table into three figures producing five figure groups, and this result is transposed on a key derived from the magazine France Libre. Bulk of traffic perhaps ninety per cent is furnished by Cable and Wireless Ltd.

  Next job scheduled for research station is Dutch about which nothing is known.

  ***

  Continuation of my message on Commercial Section number four eight seven six. Extent of shift from commercial code to clear is most surprising; excluding JIG, out of some 7,000 intake per day only about one hundred are in code all the rest in clear. French and English branch of commercial section comprises four people, intake about two thousand per day, circulate in Charlie Sugar series about two hundred per day. No translation done in this branch, British assume everyone can read French and circulates French messages in French.

  German branch ten people, circulates from one hundred fifty to two hundred per day; small amount is MOSSE repeat MOSSE but the bulk is in clear, HISROWAK traffic not handled by this branch in considerable German traffic to Far East, some of which reflects German sales of machinery; other traffic deals with transactions in grain, mohair etc with Turkey and Spain, seems to be an important line of traffic. Italian section five people, all the traffic they handle is in clear; circulate about one hundred per day, covering BARFDV* with Turkey, special articles, list of other MEW papers which can be procured on request, etc; MEW Far Eastern weekly intelligence summary classified secret ten pages covers broadcasts and other information including Charlie Sugars, arranged by geographical areas within Far East with special sections on shipping and trade with Europe; MEW monthly statement of action taken secret twelve pages arranged geographical sections special sections on relief, contraband control, enemy transactions. MEW also furnishes secret daily Lloyd’s Shipping Index showing sailings and locations of merchant ships. MEW also puts out short weekly list most secret summary of comm. Series which has very limited circulation.

  I examined March file of comm. Series covering serial numbers 7540 to 8500 about thirty per day. JIG messages averaged perhaps twelve per day, usually two to four each German, Portuguese, Spanish and Swiss, scattering French, Italian, Turkish, Belgian, Persian and South American. Also included occasional Spanish Naval Attaché messages between Berlin and Madrid sent down to Hope from Baker Peter called ‘XIP Series’ dealing chiefly with German naval supplies to Spanish Navy.

  Also hastily examined two days output of Charlie Sugars each about important trade with Spain, and most important purchases of Roumanian and Hungarian oil by Italian companies such as AXSP (Axienda Generate Italiana Petrol) and Sterava Romana. Portuguese Charlie Sugar circulation about fifteen per day, Spanish forty, miscellaneous (comprising Hungarian, Roumanian, Bulgarian, Finnish and Japanese) one hundred per day. Cryptanalytic section five people headed by Hooker, whose routine job is to handle HISROWAK KRYHA traffic but also do other commercial ciphers; for example Italian 1STCAMXI cipher based on Mengrini code used Rome to Lisbon, German MELCHERS cipher used Bremen to Tientsin and Mukden, Italian commercial attaché cipher used by Angelone, and a cipher which lay on top of JIG OKURA book was pulled off and book was turned over to Berkeley Street for breaking.

  The six thousand messages rejected by Hooker are sent to Information and Records branch of censorship which re-examines them and sends about five hundred per week to MEW.

  Relations between Hope and MEW very close and MEW gives Hope lots of stuff to assist his work as follows: MEW weekly intelligence report classified secret ten to twelve single space legal size pages contains notes covering week’s developments, long term five hundred messages. Single messages rarely meaningful and it seems clear to me that either we must rely on British and accept their intelligence summaries based on this mass of material and other related material from other sources, or else we must start a very substantial operation of our own. It would be impossible for us to process or digest the Charlie Sugar series even with several assistants, may be twenty.

  ***

  Iranian principle diplomatic book has been photographed and was sent today to Baker Peter [Bletchley Park] for transmittal to Washington. Have been discussing smaller enemies and will have further information next week.

  Reference Floradora agreement Berkeley Street has sent off two batches for you and are sending specially about eight Dublin items at my request. They are interested to know how Arlington is coming on its share of back traffic, which is to be interchanged on fortnightly basis.

  ***

  SECRET

  G- Section (‘Floradora’).

  The G Section is primarily Floradora. It started out as strictly an amateur show under the present section head, who is attractive 25year old Patricia Hartley, whom Tiltman took at the beginning of the war fresh from Oxford and trained in his school. Assisting her in the central direction of crypt work are Lt. Filby, Librarian of one of the Cambridge libraries attached to Trinity College, and 2 civilian men who formerly worked for Barclay’s Bank in Germany. The total personnel of the section is 37, of whom 3 are registers of traffic, 17 are code clerks, who are proficient in German, and the 13 not included in the above figures are the following: Key breakers, Fett Erlien [sic Fetterlein], dean of crypt people and a permanent employee of great age, who was the leading figure in this work in Russia during the last war and has been with the British ever since, Adcock, Professor of Classics at Cambridge, and Trainor, another permanent employee; three who comprise the liaison section and do practically everything except break keys, depending where the heat is on; and 7 top-flight translators. Including a schoolmaster named Pallinger, and one Potter, who before the war was the German [sic] Foreign Office expert on diplomatic English.

  Conversations with Denniston

  Certain gleanings from conversations with Denniston may be of interest to you, in case you are discussing with Arlington any of the problems involved in dipl
omatic and commercial.

  First, as to Japan, Denniston says that if Arlington wants to divert any talent from present Japanese operations to turn them to JAC (?) he is prepared to take up the full slack and to transmit finished translations of all material here plus anything that Arlington wants to send him for that purpose.

  Second, he expresses a desire to give Arlington traffic and information of every kind that has to do with winning the war, by which he means complete exchange of all enemy traffic and crypt information plus anything that Arlington wants to get out of the non-enemy field where, as in the case of Turkey, we have asked for it and stated reasons connected with the war effort.

  Third, he agrees that in the crypt field each country wants to establish for itself a position of independence so that it can get and turn its efforts toward any class of traffic that may interest it.

  Since in the European field the British have been in the game much longer than we, and can supply both traffic and solution information, he agrees that it is only fair for them to give Arlington the benefits of their experience and also traffic if they want it. The only lines that he appears to draw are these. He has express instructions from the Foreign Office which prevent him from giving cable traffic into or out of London. He is not enthusiastic about giving us traffic on areas which are considered by the British as their primary concern, such as various Near East areas, but nevertheless admits that we are now so committed in those areas in their war aspects that, if we ask for that solution information, he will have to furnish it though not including traffic to and from London by cable. As you know, the British have a cable running around Africa and of course have their own communications net over which they can permit the Shah of Persia or whoever else it may be to communicate with his representatives in London or elsewhere. Finally, he distinguishes between information as to solution methods and crypt documents (defined to mean code books and key tables obtained either by cryptography or by S.S. methods). While he is prepared to give everything that they know about methods of solution, so far as Arlington may now or hereafter want them, he does not want to commit himself to give crypt documents except where required for immediate jobs related to the war. It is not clear how much in this case he is influenced by the difficulty of copying some of the books that Arlington might want, though he stresses that point, and his facilities for copying are rather limited; and he mentions some instance of a year or so ago where he had some stuff copied for Arlington which was valuable only as library material and not for current use connected with the war.

 

‹ Prev