For the bodies charged with producing intelligence the outbreak of war nevertheless presented the opportunity, hitherto lacking, to lay the foundations for their eventual success in providing decisive intelligence on a scale undreamed of in the first months of the war; and this was especially the case with GC and CS and the photographic reconnaissance (
PR) organisation. GC and CS had established before the war began that, unlike their Italian and Japanese counter-parts, the German Army, Navy and Air Force used for all except their tactical wireless traffic different versions of the same electro-mechanical cypher machine - the Enigma. As late as July 1939, however, it could hold out little hope of mastering the machine. Nor was this an unduly gloomy judgment. The Enigma, which had appeared on the commercial market in the early 1920's, had been modified before being adopted by the German armed forces at different dates between 1926 and 1935, and had thereafter been subjected to further modifications. By 1939 the Germans believed that they had made it secure against all but local and temporary compromise in the event of capture; and they had indeed made it the basis of a set of related but different cyphers which presented formidable obstacles to the cryptanalyst.
Each of the services had different keys for the machine, that is different arrangements of its wheels and interconnecting plugs. Each of them also used different keys for different purposes or in different areas; GC and CS was eventually to identify nearly 200 keys, and at some stages in the war as many as 50 different keys were to be in force concurrently. Most elements of each key were changed regularly to give a different setting, and from the beginning of the war most settings were changed daily. Permutation of the elements - of the choice and order of the wheels and of the arrangements of the plugs - allowed millions of different settings for each key each day. But the Germans were unaware, as was GC and CS until 1939, that the Poles had reconstituted the Engima as early as the end of 1932 and had solved Army settings regularly, and other settings from time to time, between 1933 and 1938.
The Poles had achieved this success with brilliant mathematical ingenuity, but by methods they would have been unable to devise but for the fact that the French Secret Service had supplied them with material obtained from Hans-Thilo Schmidt, a German employee of the cypher branch of the German Army; it consisted of two instructional manuals for the Enigma in 1931 and, during 1932, copies of several monthly lists giving daily settings for the Army key. The Polish methods had surmounted all the changes made to the machine by the Germans up to the end of 1938, but from that date two further and more drastic modifications - a new indicator procedure and an increase from 3 to 5 in the number of wheels from which the 3 in use were selected - had put the continued exploitation of the Enigma beyond the resources of the Polish cryptanalysts.
In this situation and also, no doubt, because they were anxious about the approach of war, the Polish authorities, at meetings held in January and July 1939, divulged their results and described their methods to French and British delegations; and GC and CS received from them essential documents and, in August, a replica of the Enigma machine with its five wheels, the wiring of which had been re-constituted by the Poles. GC and CS had received from the French in 1931 a copy of the documents then given to the Poles. It had attached little importance to them, in all probability because little or no Enigma traffic was intercepted in the United Kingdom until German warships operated in Spanish waters during the Spanish civil war. Thereafter, while it had realized the possibility of solving keys by methods not dissimilar from those developed by the Poles, it had made no practical progress. But it immediately recognized what steps were necessary to apply the Polish data to the latest state of the Enigma. As soon as it received the wheel wirings it started work on developing the Polish methods.
These methods were two-fold, a hand method involving the preparation of perforated sheets and a method using a cryptanalytical machine, the 'Bombe', of which the Poles had constructed half a dozen by the end of 1938. The preparation of an improved version of the sheets, an onerous task, was not completed till November 1939 and a second copy of them was not ready for despatch to the Poles, who had moved to a base near Paris when the Germans invaded Poland, until January 1940. With this copy, however, the Poles obtained the first solution of any war-time Enigma setting on 17 January, and by 23 January GC and CS had solved settings for three further days. These four settings were those that had been in force on dates between 25 October 1939 and 17 January 1940. It was with similar considerable delays that the Poles and GC and CS, working in co-operation, solved by the end of March 1940 some fifty further settings for three keys - an Army administrative key, a GAF practice key and the general purpose operational key of the GAF. But GC and CS soon consolidated this first cryptanalytical advance of the war.
From a week after the outset of the German invasion of Denmark and Norway until the middle of May 1940 it solved continuously and almost always currently the daily settings of a key brought into force for that campaign (the Yellow key). The reasons for this success were that the key carried heavy traffic, the first to do so, and that the Enigma staff at GC and CS was increasing in size and experience. It was for the same reasons that, beginning on 22 May, GC and CS mastered the general purpose key of the GAF (the Red), which was solved daily and generally with little delay from that date till the end of the war. This success was achieved despite the fact that on 1 May the Germans had again changed the indicator system for all keys except the Yellow in such a way as to render the perforated sheets unusable: GC and CS had devised new hand- methods which overcame this set-back and enabled it to keep its grip on the Red key until the first electro-mechanical cryptanalytical machine became available in September 1940.
The Polish Bombe, like the perforated sheets, had been developed to attack the Enigma through its indicator systems. GC and CS, foreseeing that these systems would change, had from the time it received the Polish material set out to devise a more versatile machine. The first effective model, which had the power of at least 12 Polish Bombes, began working from early in August 1940. Thereafter, though their number increased only slowly, the machines, with their ability to produce faster solutions, gradually superseded the hand methods and were to be the mainstay of GC and CS's success in extending its mastery of the Red key to an ever increasing number of Enigma keys.
The regular reading of the Yellow key from mid-April and of the Red key from the end of May opened up to British eyes for the first time an intimate view of the organisation and methods of the German Air Force and, to a smaller extent, of the Army. But while the decrypts were from the outset invaluable for research purposes, they could not quickly be turned to account as a source of operational intelligence. They were full of code-names, pro-formas, and bewildering references to units and commands of which the intelligence authorities had acquired no previous knowledge. Not less important, they had become available so suddenly, not to say unexpectedly, that no provision had been made for the dissemination of their contents to British Commands with the degree of security that was essential when handling such a sensitive source. For these reasons no use could be made of them during the Norwegian campaign and they were of little operational value during the campaign in France.