Seizing the Enigma: The Race to Break the German U-Boat Codes, 1933-1945

by Kahn, David

  Their success kept Dönitz from seeing that the Allies had lost much of their U-boat intelligence. And later in the year, when the U-boats returned to the central Atlantic, he could attribute their increased sinkings, not to an Allied cryptanalytic failure, but to a German success: breaking Allied codes.

  For the Kriegsmarine not only made codes, it sought to break them as well. The 900-man English-language section of the B-Dienst was headed by former radioman Wilhelm Tranow, who had been brought in to investigate the hermeticity of the Enigma. A tall, erect man, with firm features and a compelling way of talking, so bursting with energy that he seemed to skip instead of walk, he was that rarity in a bureaucracy: a man who both performed the technical aspects of his job in exemplary fashion and administered the men under him effectively. It has been said of him that “If one man in German intelligence ever held the keys to victory in World War II, it was Wilhelm Tranow.” Tranow, who had cracked Royal Navy messages in World War I, achieved an important breakthrough in 1935.

  He and his assistants had solved the Royal Navy’s most widely used code, the five-digit Naval Code. But they had had less success with the more tightly held and more important Naval Cypher, which despite its name was a code; it used four-digit codegroups that were superenciphered to provide an extra layer of secrecy. The superencipherment consisted of adding to the four-digit codegroups random-seeming numbers from tables of 5,000 number groups. In the fall of 1935, a British naval squadron patrolling the Red Sea to keep watch on the Italian invasion of Abyssinia superenciphered its Naval Code messages, using, however, the Naval Cypher number tables. Since Tranow had solved the Naval Code, he could easily determine the superencipherment. He could then strip it from the Naval Cypher. His knowledge of the names of the ships and their movements cracked the bared Naval Cypher. And these systems were still in use when the war began.

  The B-Dienst of that time employed 500 persons, some of them in its sixteen intercept posts, some at its headquarters in the main navy building, the brown sandstone structure at Tirpitzufer 72–76 in Berlin. By April of 1940, the B-Dienst was reading a third to a half of the messages it intercepted in Naval Cypher, which it called FRANKFURT. On August 20, 1940, however, the British replaced the Naval Code and Naval Cypher with new editions of each, reducing B-Dienst successes against them. In June of 1941, Naval Cypher No. 3 came into use for communications between the Royal and the United States navies. Many of the messages concerned convoys, stragglers, and the U-boat situation.

  To strip the superencipherment from a message—the first step in solving it—required a “bite” of two or more messages with overlapping superenciphering numbers. This phenomenon was not at all uncommon. Indeed, in any one hundred messages the chances were better than half that two would not merely overlap, but would start at the same point in the number tables. Tranow’s six tabulating machines prepared lists that discovered the overlaps. Like the British, the Germans exploited cribs. For example, the senior naval officer in Newfoundland radioed a comprehensive convoy report for the North Atlantic at the same time every day that invariably began, “SNO Halifax BREAK GROUP Telegram in [a number of] parts FULL STOP Situation. …” Hut 8 would have recognized the technique.

  With aids like these, the B-Dienst—by now totaling 5,000 persons, of whom 1,100 were in Berlin—mastered Naval Cypher No. 3 throughout most of 1942: by December it was reading 80 percent of the messages it intercepted. Dönitz planned nearly all his U-boat operations during this time on the information from these decrypts. On October 30, for example, the B-Dienst submitted a report that Convoy SC 107, then east of Cape Race, at the tip of Newfoundland, would steer 45°. At the same time, a U-boat sighted its exact position. Dönitz at once ordered his submarines to intercept it. “The timely arrival of the radio reconnaissance report on the route of the convoy,” he later wrote in an appreciation to the B-Dienst, “made it possible to pull the U-boat formation together so narrowly that within a few hours of the first sighting several U-boats made contact.” They sank fifteen steamers. Though this was an exceptionally striking case, the German cryptanalysts played an important role in U-boat attacks on convoys. Indeed, Dönitz estimated that 50 percent of his operational intelligence came from the B-Dienst.

  Its romp slowed on December 15, when the British began enciphering their system indicators. But Tranow tripled the manpower working on the problem by shifting staff from other systems, and by February 1943 the B-Dienst was again solving a large proportion of the traffic, sometimes reading directives to convoys ten and twenty hours before the movements they ordered took place. Its information helped Dönitz position his U-boats for what turned out to be the greatest convoy battle of the war. During three days in March, the U-boats sank dozens of ships from convoys SC 122 and HX 229. The Admiralty despaired, thinking that the Germans had virtually cut Britain’s lifeline to America. It was not until Naval Cypher No. 3 was replaced in June that the B-Dienst lost its grip on Allied communications.

  Though these successes and failures tended to screen from Dönitz the Allies’ varying fortunes in codebreaking, secrecy was so vital to the success of German naval operations that almost any suspicion of its loss demanded an investigation. Thus in March 1942, the loss of two German auxiliary cruisers—the disguised and armed merchant vessels that sailed the seas to harass and destroy enemy shipping—prompted Admiral Fricke to look once again at operational secrecy. He analyzed the evidence and found no leak from communications. The movements of enemy ships contradicted any suggestion that they were reading the messages radioed to the auxiliary cruisers, he said. Similarly, when the battleship Tirpitz sailed to Trondheim and the Gneisenau, Scharnhorst, and Prinz Eugen dashed through the English Channel and the Straits of Dover to haven in German waters, intercepts of British messages gave no hint that the British were reading German cryptograms. And there were no indications that British information on German U-boats came from codebreaking: the implication was that it came from direction-finding or air reconnaissance. Fricke said that the Enigma was sixty times better than the British system and concluded that it was “unimaginable” that the enemy could read it. At the time, for M4, he was right.

  17

  BLACKOUT ’42

  A VARIETY OF CLUES TOLD NAVAL SECTION STARTING IN MID-1941 that the Germans were going to replace their three-rotor Enigma on the Atlantic U-boat net with a new, four-rotor machine. The captured U-570 had yielded up a machine lid with four windows for the rotor key letters. Solved intercepts referred to the new machine, which had sometimes been used in error. Indeed, in December 1941 one such message, combined with a repeat enciphered in the proper three-wheel mode, enabled Hut 8 to recover the wiring of the fourth rotor even before it officially went into service.

  All this made the cryptanalysts very apprehensive. The new rotor would multiply their work by a factor of 26, and they were barely keeping up as it was. Their apprehensions were justified. With the introduction of the four-rotor Enigma on February 1, 1942, a virtual U-boat blackout descended upon Hut 8. No longer could it read messages to or from Atlantic submarines on patrol or attacking or ordered to attack convoys. The Submarine Tracking Room admitted in its first weekly U-boat situation report after M4 came into effect that “the picture of the Atlantic dispositions is by now out of focus.” It grew increasingly indistinct.

  Still, the situation was not quite as bad as during the no-Enigma-solution days preceding the captures of the München and the U-110 in mid-1941. Other sources of intelligence, such as aerial reconnaissance, were functioning better than before. The tracking room had built up considerable knowledge in the half year it had been reading combat U-boat messages. Those solutions had told the British such details as the U-boats’ average speed of advance when proceeding to and from patrol and the endurance of the various types of submarines. Home Waters and Mediterranean Enigma and Dockyard messages, which continued to be solved at the astonishing combined rate of 14,000 per month, revealed the construction of new U-boats, their preparation
and crew training in the Baltic, their addition to the fleet, their departures from their home ports. The Submarine Tracking Room thus knew the strength of the U-boat fleet and the number of boats at sea. It was further helped by radio fingerprinting, through which experienced intercept operators identified individual enemy transmitters by their distinctive methods of sending, or “fists,” and by TINA, a device that recorded and displayed on long strips of paper radio transmission characteristics to identify individual radiomen. Though not very reliable, these techniques could at least distinguish supply and minelaying U-boats from combat boats. Rodger Winn put into effect his “working fiction,” which maintained that the Submarine Tracking Room’s best estimates should be treated as facts and acted on until proven wrong.

  For information on the U-boats’ locations in the Atlantic, the tracking room relied heavily during the M4 blackout on direction-finding. When a monitor in the big intercept rooms at Winchester or Scarborough heard a warship or a U-boat replying to the land station to which he or she was listening, the operator would yell out the frequency as loud as possible. At the far end of the room a controller would repeat it to the direction-finding stations on a direct landline.

  The reported frequency went to the eight direction-finding stations from the Shetlands to Land’s End. Each had several listening posts attached to it, as well as subordinate control centers. The control center attached to the DF post of Wick, outside the Scottish village of Bower, was an H-shaped building that included a dormitory. Bearings were taken in five 10-foot-square wooden huts isolated on the Caithness moors. Two were staffed by civilians and two by members of the Women’s Reserve Naval Service, known as Wrens; the fifth was a spare. No roads led to the huts; the operators were driven as close as possible and then had to walk a mile or two across the boggy ground. In winter a sailor helped shovel paths through the snow. At all times the women yielded the paths to the menacing local cattle. If the weather was too bad, they stayed in the hut from one watch to the next, sleeping in the bunk, melting snow for water, and heating cans of food in the kettle. But they preferred the trek back to the dormitory to staying in the hut and being jolted awake by the hut loudspeaker that announced a frequency on which a bearing had to be sought. None of the hardships fazed the young women. They felt that they were directly helping their men at sea and were proud to be doing the work.

  Each operator sat at a desk listening to three telephone lines, one to the loudspeaker, one to each earphone. When the Wren heard the controller at Scarborough or another intercept post call out a frequency, she tuned the radio on her desk to that frequency. Early in the war, she would turn a wheel that rotated an antenna to reduce the signal to its lowest intensity; 90° from this was the source of the signal. Later she had a cathode-ray tube that scanned the horizon electronically. When it discovered the azimuth at which the signal was strongest, a spot in the center of the 9-inch circular tube stretched itself out toward the circumference in both directions. The operator spun the movable glass plate within a metal ring marked in degrees until a wire embedded in the glass lay above the glowing green line. When she put the tube’s cursor on the line, one half disappeared, so that it indicated a single direction—southwest, for example, rather than both northeast and southwest. The work had to be done very quickly, since many of the U-boat signals lasted mere seconds; sometimes the women got a bearing only on the last letter.

  The operator noted the bearing in a log. When the control at Bower called on the intercom, she gave him the bearing; Bower would pass this to the Submarine Tracking Room’s DF section, where the several bearings would be plotted on the chart that used the strings and pins.

  During nearly all of 1942, Hut 8 solved only three TRITON keys: those for February 23 and 24 and March 14. The effect of this lack of ULTRA was becoming all too clear. In the last half of 1941, with Enigma solutions diverting Allied ships away from wolfpacks, U-boats sighted about one of every ten convoys. In the last half of 1942, with not very many more U-boats stationed on the North Atlantic routes, but with detours controlled only by direction-finding, U-boats sighted one of every three convoys. As a result, North Atlantic sinkings, which totaled some 600,000 tons in the last half of 1941, more than quadrupled to 2,600,000 tons in the last half of 1942. And each of the nearly 500 ships sunk in those six months meant more freezing deaths in the middle of the ocean, more widows, more fatherless children, less food for some toddler, less ammunition for some soldier, less fuel for some plane—and the prospect of prolonging these miseries.

  Of all this the cryptanalysts of Hut 8 were acutely aware. But they could find no workable cribs, and without these they could not test for keys. Moreover, the three-wheel bombes were inadequate for solving the four-rotor Enigma: the three solutions of February and March, based on kisses, had each taken seventeen days on six three-wheel bombes. High-speed bombes to attack M4 were designed and put under construction. Nevertheless, at the highest levels, fears grew that the lack of intelligence, combined with the growing numbers of U-boats, would sink more ships than could be built. The tonnage left would not suffice to maintain rations and sustain industry in the United Kingdom, much less bring over troops, supplies, weapons, and ammunition to carry the war to Hitler.

  So, while doing everything they could to accelerate ship construction and protect existing vessels, the supreme commanders looked with growing impatience to B.P. for its contribution. On November 22, 1942, the O.I.C. urged B.P., with British understatement, to focus “a little more attention” on the four-rotor Enigma. The U-boat war was, it said in a silent scream, “the one campaign which Bletchley Park are not at present influencing to any marked extent—and it is the only one in which the war can be lost unless BP do help.” Unbeknownst to the O.I.C, however, that help was already on the way.

  18

  THE GEORGE CROSS

  IN SEPTEMBER 1941, HITLER ORDERED SIX U-BOATS TO THE Mediterranean. They were to sink Allied vessels transporting supplies to North Africa and so relieve pressure on Rommel. Dönitz objected. He wanted to pursue the fundamental goal of cutting Britain’s supply lines. But Hitler told Dönitz that holding North Africa would keep Suez closed and force the British to go around the Cape of Good Hope, thus in effect costing them 3 to 4 million tons of shipping. In November he sent four more U-boats. They helped: Rommel, who had been driven back in 1941 almost to his starting position, was ready in January 1942 to launch an offensive. Eventually fifteen submarines, all of them with some of Dönitz’s most experienced crews, prowled what Hitler now called the “decisive area for the prosecution of the war,” the Mediterranean.

  Among the more successful of these subs was the U-559. A standard Type VIIC submarine built by Blohm & Voss of Hamburg, she had had the same captain since her commissioning in February 1941. Lieutenant Hans Heidtmann, red-haired and bearded, was a solid, quiet man. He came from Lübeck, the ancient Hanseatic port of the Baltic, and was a typical north German: precise, orderly, no-nonsense. His first cruise, east of Greenland, resulted in no contact with the enemy but an unfortunate one with an iceberg, which damaged the periscope. His second likewise had good and bad aspects: the U-559 sank the 8,000-ton British freighter Alva northwest of Spain but was attacked by three British destroyers. Heidtmann and his crew rapidly learned what war was like: 180 depth charges exploded around them while they shivered 600 feet down; after twenty-four hours underwater, with the attackers apparently gone, they surfaced and returned to St. Nazaire.

  Then the U-559 joined the submarines transferred to the Mediterranean, sliding through the Strait of Gibraltar on the strong current from the Atlantic. At first she was based at the Greek port of Salamis, which gave her young crew members both a rumor to wonder about and an insignia to laugh about. The rumor was that 25 percent of the U-boat sailors contracted venereal disease from the complaisant Greek girls and that many of the young men were executed for their carelessness. The insignia came from the belief that in Salamis donkey meat was used in the salami the sailors ate. For f
un they painted a white donkey on their conning tower. It was German humor.

  In cruise after cruise, submerging at first light and surfacing after dark, Heidtmann and his crew sought their prey. On November 27, 1941, they torpedoed a small Australian warship, the Parramatta, off Bardia on the coast of North Africa. And twenty-four action-filled hours began just after midnight of December 23, when they detected the shadowy apparent shape of a cruiser to the west and fired a spread of four torpedoes, all of which missed. They then lost contact, dived, were attacked by bombers, and surfaced. Sighting a two-freighter convoy, the U-559 fired three torpedoes as the ships overlapped in her line of vision. One torpedo sank, and the other two ran on the surface, striking the second freighter amidships and sinking it at once. Heidtmann and his men tried to attack the other steamer but could not approach because of its three escorts. They fired a torpedo at one of the destroyers, then saw it avoid the torpedo and turn and approach their submarine. They dived. They heard a weak detonation and then a very loud one that shook the boat. After sneaking away to the northwest, they heard weak propeller noise to the west and seven depth charges not very close; at 11:40 P.M., they surfaced and breathed deeply of the fresh salt air. Three days later, in the same area off the coast of Libya, where Hitler hoped to cut off British supplies to Tobruk, the U-559 sank a Polish steamer. Cruises like this earned Heidtmann the Knight’s Cross and compliments from Dönitz such as “Decisively and well carried out operation.”

  At 4 P.M. on September 29, 1942, the U-559 sailed from the Sicilian port of Messina on her tenth cruise. She headed for grid square CP 3468, which covered part of the eastern Mediterranean just south of Turkey, close to the Syrian port of Latakia. Three days later U-Boat Command radioed Heidtmann’s operational orders: “Free hunt off Palestine and Egyptian coast. Point of main effort: supply routes Alexandria-Port Said to Beirut, Haifa, Jaffa.” Over the next several days, U-Boat Command passed on an agent’s report that a British cruiser division in Haifa was expected to sail soon, listed the ships that were in the harbor at Beirut, and notified Heidtmann that a spy said two British steamers were sailing on October 10 from Port Said for the Turkish port of İskenderun, close to his operating area. On October 17, Dönitz moved Heidtmann’s operating area south, to the west of Haifa, then farther west into the Mediterranean. But neither these moves nor the information that a German airplane had sighted a three-vessel convoy with escorts a few hundred miles to the southwest helped him sink any ships on this cruise. So, late in the evening of October 29, after asking him to report the weather if possible, Dönitz told Heidtmann to return to Messina after he had used up his fuel and armament. In the early morning hours of the thirtieth, the U-559, using its Short Weather Cipher, transmitted the requested meteorological report.