Secret Weapons (33 page)

The story is a fake. The submarine U-571 was never captured by anyone, but was sunk by a torpedo dropped from a Short Sunderland flying boat from 461 Squadron of the Royal Australian Air Force in January 1944 off the coast of Ireland. No Americans ever captured a naval Enigma machine.

The British audience knows a different version. They are aware that it was actually the crew of HMS
Bulldog
who captured an Enigma machine from a German submarine, the U-110. The capture took place in the North Atlantic in May 1941, when the United States had not even officially entered the war, and the enterprising British sailors were responsible for seizing the German machine and the documentation aboard.

This allowed their intelligence officers to come face to face with an Enigma machine for the first time, thus finding out how this remarkable machine could operate. That is not true either. Although the dates (and the vessels) are correct, the belief that this was the first encounter with Enigma is a myth.

Long before World War II began, the Enigma machine was already well known to the British, and to many other nations as well. It was not invented for the war, and had begun life as a commercial encryption device that was available for several years by mail order. A German engineer named Arthur Scherbius designed the original machine. It used the now-famous system of rotors and in February 1918 Scherbius patented his device – this was during World War I! In collaboration with a funding partner, E. Richard Ritter, the Scherbius & Ritter Company was established. They immediately approached the German authorities, believing that their machine would be of value for international top-secret communications. The Foreign Office considered the design, and reported back that it held no interest for them. The German Navy were then approached, but they said they were not interested either. Somewhat disillusioned by the response of officialdom, Scherbius and Ritter eventually joined with others to form the Chiffriermaschinen Aktien-Gesellschaft (Cypher Machines Stock Company) in 1923, and began commercial production of the first Enigma machines.

The ingenious device was publicly exhibited for the first time in that same year and mail-order sales began immediately. The main disadvantage of the first machines was their large size. They were was equipped with encryption gears and a full typewriter mechanism, and weighed some 110lb (50kg). The keyboard, rather than being in the standard European QWERTY layout, had the keys set alphabetically. An improved version, the Model B, was somewhat less bulky, and in 1926 the Enigma C machine was released. Instead of the heavy typewriter mechanism, it was fitted with small lamps that the operator had to read, and in consequence was nicknamed the Enigma Glowlamp. During the following year it was replaced with the Enigma D, and this was sold to hundreds of customers during 1927–28. The Enigma was now famous, and was sold across Britain, Italy, Japan, the Netherlands, Poland, Spain, Sweden, Switzerland and the United States.

This Enigma machine was a masterpiece of design. In use, a key pressed by the operator would connect one of 26 letter circuits, passing a current to one of 26 contacts in the encoding unit. Electrical current was then passed through three rotors, each of them wired so that the letter was changed. The turning of the rotors to make a new set of contacts each time a key was pressed meant that each letter was transposed to a different character every time. The problem is that, to decrypt the message, the receiving machine would have to feed power from the keys to mirror the original action exactly. The Enigma designers solved the problem by adding a reflector system that connected each contact to another and routed the circuit back through the three rotors. The original text could then be recovered by the receiver – but not by anyone who attempted to intercept the message en route.

The first nation to adopt the Enigma for military use was Italy, who code named her Enigma machine Navy Cypher D. The same technology was used by the Fascist government of General Franco during the Spanish Civil War, and the Swiss called their Enigma D machine the Swiss K. An Enigma model T was manufactured for use by the Japanese. The codes for many of these were broken by Britain, France, Poland and the United States. But the Germans were becoming interested in how Enigma could be used to encode all their transmissions, and the British were anxious because they lacked the insight into what was going on, and were eager to keep up to date. Already the Enigma machines were being adapted from having three rotors to four, which made breaking the codes far more challenging.

The Germans now had the latest encryption, and they were confident that they could rely upon complete secrecy while sending orders and complex messages from central government to the armed services. The decoding problem was broken in 1928 by the Poles. Their German-language Biuro Szyfrów (Cypher Bureau) BS4, was suddenly informed that a large parcel had been sent by the German government to their legation in Warsaw. Ordinarily, it would not attract attention, but the Germans had dispatched it by unsecured freight mail and, as soon as the error was noticed, they contacted the Polish authorities urgently to check that it was being safely handled. The enquiry was passed to the Polish Customs Authority, who immediately tipped off the intelligence service that there was something crucially important in the box. As a result, the code-breakers at BS4 had a full weekend to work day and night analysing the machine and all its code protocols, before meticulously repacking it for the Monday morning. The Germans were meanwhile assured that yes, the parcel was intact, and it would be ready – perfectly packed – for collection as soon as the office opened for business. The Polish bureau now knew all they needed to know, while the Germans believed their secret was secure.

Once the Enigma machines had received official endorsement by the German government, the German Army finally conceded that Enigma was what they required, and in 1932 they ordered the three-rotor Enigma G machines. Within two years it had been improved and was designated the Enigma I and this soon became known as the Wehrmacht (Services) Enigma. It was used throughout World War II. This new version had the addition of a plug-board which allowed the operator to exchange letters in pairs, which added greatly to the security of the codes. The machine was much smaller by this time, weighing as little as 26lb (12kg). Eventually, in 1934 the German Navy brought into use their own version, the Funkschlüssel-M or M3, and in August 1935 the Air Force adopted the Wehrmacht Enigma for their coded messages.

What they did not realize was that the Polish intelligence service was already ahead of them. One of their leading code-breakers, Marian Rejewski, had designed a top-secret machine that could be used to decode the German messages. It recreated the settings of the Enigma machine that had originally encoded the text, and allowed the operator to read the message. Rejewski called it his
Bomba Kryptologiczna
– the ‘cryptological bomb’.

Each operator of the Enigma machines was given a three-letter code that was regularly changed. This was sent first, to indicate to the receiver which rotor setting was to be used that day. The three letter code was repeated, so it might be PINPIN. Of course, the message was encrypted by the Enigma machine, so it came out at the other end as random characters – MXZLPD, for instance. This gave Rejewski the vital clue. The techniques to work out the way the three letters were coded were already well established, and the fact that the three letters were repeated was vital. The way the letters had changed when they were repeated let Rejewski work out exactly how the rotors had been installed.

The design of the Enigma machine was being regularly upgraded, in order to add further levels of complexity, and the number of possible combinations that the cryptographer had to try was becoming impossible to manage. A better system was needed. So, in the autumn of 1938, Rejewski designed his first decoding system; it became known to the Allies as the Bombe. In principle, it had the power of six Enigma machines working simultaneously, and allowed one cryptographer to do the work of 100 working alone. Within a year there were six Bombes working at the code-breaking station in Warsaw. It meant that one room contained decoding machines that could do the work of 600 highly trained staff.

By the time the Germans invaded the Sudetenland of Czechoslovakia on 1 October 1938, the Polish cryptographers had been decoding the German Enigma messages for over six years with complete success, and the Bombe had made the process steadily faster and more reliable. The Germans were meanwhile making the machines increasingly complex, and fitted extra rotors, along with a complex plug-board, so the task of decoding became far more complex. Nonetheless, the Poles were able to decode the messages that revealed the Germans’ next intention – to cross the strip of Polish territory that separated Germany from the Baltic enclave of Danzig. This would amount to an invasion, and war now seemed inevitable. When it became clear from the decrypted messages that the Germans were ready to invade Poland, all the decryption codes and the Enigma equipment was delivered by the Polish authorities to British Military Intelligence.

January 1939 had marked an important intelligence conference in Paris that brought together officials from the British, Polish and French intelligence services. The Poles revealed the extent of their remarkable success in routinely decoding German messages. To the British there was only one possible response – they would expand the Government Code and Cypher School (GC&CS) to take up the task of decoding of the Enigma messages. In August 1939, still before war had been declared, the Government Code and Cypher School moved into its new and expanded premises. This was in the beautiful old house of Bletchley Park, near Milton Keynes in rural Buckinghamshire. The building was equipped with a top-secret telephone exchange and a teleprinter room, with a kitchen and dining room alongside. The upper floor was occupied exclusively by the Secret Service MI6. A nearby boarding school was requisitioned to house the Commercial and Diplomatic Section, and a range of wooden huts was speedily erected in the grounds to use as the establishment increased in size. There was a water tower alongside, to supply the historic house, and a monitoring radio station code named Station X was set up at the top, so that its aerials had an unimpeded view across the open landscape. As the lengthy wires of the radio aerials might attract unwanted attention, the radio station was soon moved to nearby Whaddon.

Britain’s most brilliant cryptographer was a young mathematician named Alan Turing. He had read mathematics at King’s College, Cambridge, and then studied for his PhD at Princeton in the United States. At the age of 24 he published a brilliant paper on ‘computable numbers’ that was far ahead of its time, and he is widely regarded as one of the founding-fathers of modern computing science.

When the Polish consignment arrived in Britain, it was Turing who had the task of deciding what to do next. He constructed a design for an upgraded version of the Bombe with 108 different places where the decoding drums could be fitted. Each rotor could be set to one of 17,576 theoretical positions and the machines that Turing designed could try them all in 20 minutes. Manufacture of these Bombes was entrusted to the British Tabulating Machine Company at Letchworth, near London, where Harold Keen had led a research project to introduce punched-card technology, introduced from the United States, to calculating machines. The British Bombe machines were, as one commentator said, ‘the size of a large book-case’ and measured 7ft (2m) wide, 6ft 6in (1.9m) tall and 2ft (0.6m) deep. Each weighed about 1 ton.

In March 1940 the first was completed and installed at Bletchley Park under the code name Victory. A second one followed in August. This had a more advanced diagonal plug-board and was named Agnus Dei (the Lamb of God) but became known as Agnes, or Aggie, for short. Victory was soon upgraded to match Aggie’s specifications. Work at the British Tabulating Machine Company was redoubled, and five separate decoding stations were set up around North London, in case the facility at Bletchley Park was destroyed by a German raid. From that time on, the British were capable of overhearing the German military communications. Everything was carried out under conditions of top security and total secrecy, and the staff at Bletchley Park remained famously reticent about their work for decades following the war. London did not always appreciate the importance of their delicate, private approach; once the building was operational, a group of distinguished officials from London came down to inspect the progress. They arrived in a column of official cars with motorcycle outriders, flags waving, and everyone in full uniform; ‘So much for the secrecy,’ said one of the scientists.

In July 1942 the drawings and wiring diagrams were passed to the United States Navy, and in September 1942 funding of $2 million was requested for the construction of Bombe machines in America; the project was approved within 24 hours. This new version was built by the National Cash Register Corporation (NCR) in Dayton, Ohio, with full collaboration between the Americans and the Bletchley Park team and had a greatly increased capacity. An order for more than 300 of the American machines was placed.

Alan Turing was seconded to Washington DC to advise in December 1942 and went straight to NCR. He quickly calculated how the machines could be linked together, and determined that a smaller number would be sufficient – so the order was reduced to just 96 machines. The American machines were larger than the British Bombes, weighing 2.5 tons, and could run more than 30 times as rapidly. The first came on stream in May 1943 and in June two American Bombes named Adam and Eve broke a particularly difficult message which many believed it was impossible to decode.

A different design was chosen by the US Army and their contract was given to Bell Laboratories in September 1942. It was fitted with telephone relays instead of combinations of mechanical drums, so rotor changes could be done by pressing buttons rather than physically exchanging the rotors. It was intended only for three-rotor decryption, not the four-rotor traffic that the Navy Bombes could handle, but its refined design greatly reduced the time it took to decode messages.