Alastair Denniston

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

  Everything that is circulated goes to the Foreign Office and Director, that is C, and except in very special cases everything goes to the War Office and Admiralty (will have to check further what goes to Air Ministry). The very special cases that service ministries do not get are such cases as where material shows that some British diplomat abroad has ‘dropped a brick’. Policy is to recognise that especially in time of war service ministries have their own views on matters of foreign policy and are entitles to be kept fully informed on all diplomatic developments. In case of Berkeley Street as in in case of Park, what impressed us was the pains that are taken to see that all information gets out to those who can make use of it.

  There are liaison officers from various governmental offices who are assigned to Berkeley Street and have access to the material, although this is not the full time job and most of them are high personages. In this manner the following are directly represented in Berkeley Street operation:

  C’s office; BP; Winnie himself; Foreign Office; all three services; Ministry of Economic Warfare; Treasury; War Cabinet; Colonial Office; Military Intelligence Five, which is counter espionage; India Office and Secret Intelligence Service, which is Charlie in another form but this last named liaison officer does not serve C but serves other branches of the government, including Board of Trade; Department of Overseas Trade; Petroleum Board; Home Office and Minster of War Transport. There is some form of liaison to everybody who might be able to act on information coming in.

  Through BP liaison is maintained with Washington, New Delhi, Melbourne and African Force Headquarters. Please make following correction at beginning of message: Elderly ladies do not register and record all incoming material, but only material from certain sources, when there is some reason to keep such records in order to evaluate source or keep track of source for some reason. At present they are keeping register of incoming traffic only from Mauritius, Port Sudan, Cairo, Gibraltar, Khartoum and Jerusalem. Khartoum is not mentioned above and I believe this is censorship point. Note however that complete register of all traffic by nationalities is kept by geographical sections.

  ***

  Records maintained at Berkeley Street do not include figures for total traffic and I did not ask them to go to geographical sections and add up all these figures, but they receive ‘thousands’ of messages per day. Their serial numbers of circulated diplomatic items crossed one hundred thousand early in 1942. Do not have present range of number.

  But monthly average has been running at from eight hundred to one thousand. Generally speaking I would say that their policy of what to distribute is same as Arlington’s, the difference in numbers distributed being accounted for by fact Arlington covers wider field and that here they have been going for much longer and have learned over period of time what is new information and what is not; and they also have had much more guidance as to what to put through and what to discard than Arlington gets from its distributors, naturally in view of short time that operation has been going on present scale.

  Basic principle is same however, namely, to distribute everything that might have any intelligence value, however small.

  Great difference between BP operation and Berkeley Street is that BP is essentially an intelligence organization with important operational functions whereas Berkeley Street is production organization like Arlington. However they do seem to have the intelligence side of their operation, functioning as aid to crypt people, very well set up. It centers in the distribution and reference section, headed by Earnshaw Smith, a youngish looking man who nevertheless has been in this work since 1917. His assistant is Ore Jenkins, Professor of Medieval and Modern Greek at Cambridge, who has been in it since war started. They have three girls in theory but two in actuality, one of whom has been fourteen years at Berkeley Street and six years in this section. The other has been in the section for five years. These people are in effect a reference bureau for crypt people, and in addition they determine what shall be circulated and to whom it shall go in cases of doubt, and they supervise mechanical operation of getting it out.

  Nothing goes out by teletype, other than exceptional material for Chief, the four teletype connections being with BP, Chief’s office and intercept stations at Denmark Hill and Sandridge. Earnshaw Smith’s section has most simple filing system imaginable. Material is simply laid in folders in serial number order, unbound. There are three files – one is master file and another is spare file, both in straight serial number order. The third and important file is called ‘Subject and Country’ but in fact it is only country and not subject, since all material bearing on any particular country is laid in a folder in order of dates and there is no subdivision of country files by subjects. Most startling fact about this section is that it receives daily, shows to geographical sections where interesting to them and then puts in its files all incoming and outgoing communications of Foreign Office, Dominions Office, Colonial Office and India Office, and all but purely operational material of the three service ministries. More later.

  ***

  One of the most illuminating documents in Denniston’s show is book showing net results of all British intercept activities in diplomatic field for second half of 1942. It consists of one sheet or set of sheets for each city in which traffic originated. On each sheet:

  The first column shows alphabetically the cities to which traffic was destined.

  Column two further breaks down traffic for each city according to nationality, all subsequent columns following the breakdown of column two.

  Columns three and four show by call signs the originating station and, station of destination.

  Column five shows station or stations which did Interception. Column six shows estimated number of messages sent. This is based on message numbers. In some cases exact number of messages is known; in others estimate is based on numbers intercepted up to late date in month; in others it may be guess based on small numbers of interceptions not running to end of month. In cases where senders do not number messages this column is left blank.

  Column seven is number of messages intercepted

  Column eight is number missed, being difference between column six and column seven, where six has been filled in.

  Column nine is number of messages intercepted in cases of unnumbered series.

  Column ten is the usual remarks. Remarks include information necessary to understand figures, such as note that traffic only started in June or that interception commenced in August or that diplomatic relations were broken in May and traffic discontinued. Remarks also include information to guide users of book, such as note that traffic is important and that further efforts should be made to get it, or that at actual places of interception it is hard to get legible copy, or that because only slow means of relaying are available it arrives too late to be used, etc.

  Pages are thirteen plus by sixteen plus inches and triple spacing is used, leaving plenty of room for notes. Such a book would be of great use to us and over the past year if information of this kind had been available in this readily usable form it would have saved innumerable inquiries to Arlington and would have made answers to many inquiries more accurate. From such a book it is possible to take off totals of material by origin and destination, by nationalities etc, and to see what intercept stations produce important material and in what volume, to determine at a glance what sources of information there are in each principal city and how many of them are being tapped and of course to see how much of each type of traffic is being brought in.

  Recommend that you and Corderman consider setting up this kind of a record and keeping it up month to month, which British are now starting to do.

  It is impossible to exaggerate how many surprises you would get from looking at these records or how many subjects that we have discussed it would illuminate. To mention one, you will recall our discussion of possibility of stopping Axis traffic to and from Portuguese African points. You may be surprised to learn that Portugal, which governs the colony, s
ent 122 messages to Lourenco Marques during second half of nineteen forty two whereas the Italian Embassy in Lisbon send 1295, and that figures in other direction are comparable. Had we had these figures in our mind at time of our discussion we would have seen this picture in very different light.

  Pages for Tokyo give clear and extremely interesting picture of whole Japanese diplomatic communications picture. They reveal that 498 diplomatic circulars were intercepted out of 936 sent out, but unfortunately figures for interceptions does not in any instance include traffic received from United States and therefore that side of picture is incomplete.

  Tokyo figures show that German communications to and from Berlin are much more numerous than Japanese (almost four times as large).

  I cannot carry in mind all striking examples which I noted but one other is that Italian traffic to and from Bangkok is much larger than I had imagined and that much the largest item in Irish Government traffic was to and from Vichy. I give you these to illustrate how much information you can spot quicker from such a schedule and how useful it would be in various ways, including setting up of traffic priorities (Special Friedman System)

  ***

  Japan Section of Berkeley Street is most impressive because of experience and ability of personnel. Traffic coming in is classified and registered by three girls who have had long experience and is then passed to proper crypt people who number 14. Those who actually do crypt work at present are not prewar employees but have been in this work most of war.

  One girl handles crypt side of Purple while two men in middle thirties or thereabouts do all work of recovering JIC keys and are now complaining of lack of work. Head of Section was away when we visited and name escapes me.

  One young man handles liaison with us and Australia and balance of section consists of decoders and clerks. This young man and two men handling JIG Nineteen keys are typical of selection methods here. Before war one was history teacher in public school, one was detective story writer and third was seller of odd books to American collectors. All were called up and all were selected for Intelligence Corps and after six months in Army sent to Tiltman’s school. On finishing course they had privilege of taking commission or being returned to civil life in assignments in this work. Two took commissions and are now Captains, the third chose to serve as civilian. All wear civilian clothes.

  You can see that these realistic British know that men working on JIG Nineteen are not really in Army and they do not have to pretend that they are unless they want to. JIG Translation Section is much higher form of animal life than Crypt Section and includes people who have done crypt work and might be doing JIG crypt now if it were not so easy. Head of Section is White, who at beginning of war was Consul General in Tientsin and had long service in Japan and China. All work coming from Crypt Section passes over his desk and is examined by him or whoever is acting during his absences, in original language, and is marked with one to four symbols meaning high priority, priority, unimportant, or please look over to determine whether it is worth translating. High priorities must be translated at once, priorities must be translated when translators get around to them, unimportant ones are not translated. Result is that eighty per cent of color material and fifty per cent of JIG Nineteen’s are translated. Lousy Annie material is not normally translated. It is looked over in original language and watched for items referred to in higher class material, to circulars, for identifications OK persons etc, and in certain special cases, such as Kabul, where it has proved to be source of information, it is followed with more care.

  This is based upon study of Lousy Annie* material made last year.

  Translation staff (working on DIP and not including those on Commercial) consists of retired Colonel (who was Military Attaché in JIG in twenties and has been in his present work since 1927), and seven civilians: Cunningham, formerly Consul in JIG, Roscoe, who was Consul in JIG during last war, thereafter lived and worked there and returned here for present work in nineteen thirty one after seventeen years of residence there, Hobart-Hampden who they say is probably world’s leading JIG scholar and coeditor of Standard Dictionary recently reprinted in Washington, two recently retired Consuls, who were Consuls General at two JIG commercial cities, one Consul on active list, who when war started was serving at Taiwan and, finally, one Braithwaite, who was brought up in JIG, got his advance education there during last war, went into business and resided there up to present war. As White remarked, his section’s standard of scholarship in JIG is very high.

  We did not have chance to inspect closely the files of this section or aids used in its work. While general impression is of men steeped in knowledge of their special field, and who carry most of what they need in their heads, we did see a lot of dictionaries and reference books and two complete looking files, one of personal names and one of geographical names, the latter apparently considered very important in their work. They of course keep files of their own material, and I got impression that they more or less specialize, one man dealing with material relating to one part of world and another man another part, and that these men are not merely translators but are experts on the areas with which they deal, and that they are consulted by Foreign Office people for their opinions on information coming out or material.

  All of which leads me to this suggestion: Arlington is plugging away at very difficult work which is largely for use of State Department [and] that Department must have JIG experts who are spending their time at much less important work than getting this JIG diplomatic material out when Army can very properly say that, with JAC material to handle, it cannot do State Department’s job unless such people are furnished to handle DIP translation work.

  This leads me to another thought: when you get some idea of economic warfare picture that has been built up here from continental commercial material, by process of just getting every tiny item and putting them all together by most painstaking labor, you cannot fail to think that we ought to be doing same thing for areas involved in JIG material. It would take many months of most intensive work to start building up picture, and no time ought to be lost in starting it, because at some point in this war important decisions in JIG area are going to be affected vitally by what we know and do not know about the details of economic matters in those areas.

  Lord Farrar, who sits in Economic Warfare Ministry for purpose of handling most secret information in continental economic field and in fact for whole world, told us yesterday that 85 per cent of all important information about South Central economic picture has been built up from Berkeley Street production, the contributions by censorship, foreign press, radio broadcasts and agents being trivial in comparison with our material.

  There is evidence that people here, who already are reading and translating around a thousand JIG repeat JIG commercial items per day, are beginning to build up JIG economic picture.

  Here is another case in which, unless we approach this problem with imagination and energy, they are going to steal the show.

  Problem is principally one of translation, since large part of stuff comes in clear and code problems are either not serious or can be licked. When sure that with help of State Department, with vigorous efforts to find personnel without being scared by shadows (and security problem could be taken care of easily by putting this translation operation in place away from Arlington) and with MSVCL selling job that should be easy in light of evidence of what British have been able to do, this JIG commercial picture could be built up into one of best intelligence sources available. Whole job except decoding could be handled by Special Branch if translating talent were found, and translators of ability might be willing to participate in work that would involve both translating and adding up the produced intelligence. Strongly recommend that you consider this in prayerful conference with powers, including Adolph Berle.

  JIG commercial part of White’s section describes itself also [as a] Translation Bureau and consists of six translators and two lady clerks. Principal translators are
Captain Rayment and Captain Shaw, both of Navy, and a civilian who I believe is an ex-consul. When war began Shaw was serving as Chief Cryptographer at Singapore. Rayment was officer of Naval Intelligence and JIG expert and translator for Navy. Note quality of personnel assigned to this commercial field.

  Believe their traffic comes from Point Grey and other west/coast Canadian stations and is sent over every week or ten days by bomber. Great majority of items are in plain language. Material in four figure code thought to be mercantile marine code, in which page start with 111 or 110 are of interest to Park and are sent up there. Some of material is in NYK RPT or OSK RPT OSK codes, and occasionally there are items in one of the other MITS RPT MITS. Some messages are in public commercial codes. In cases of both MITS, both old and new books are used, old ones for Europe. I understood them to say that in both cases they have old books and have made progress in building up new ones.

  Large part of traffic is inconsequential, such as small financial remittances and even greetings. Rest of it concerns mainly commercial transactions, mostly offers to buy or sell, conditions on obtaining licenses. There are however about ten or twelve items per day in ship material from Canada which are important enough to circulate. Examples of these as given were references to ships by name or routing, reports or large financial transactions or of specific completed commercial transaction, messages which give address of military or naval unit or ask for personnel for such purposes as building ships, putting up power house, or contain information or clues to progress of dock repairs, restoration work on mines, refineries or something of the kind. Their rough guess on volume of Canadian material is fifteen hundred items per week.