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

by Kahn, David

  The effort to lift the vessel began. With two-thirds of the compressed air supply gone, Dresky ordered the unorthodox maneuver of blowing the water out of one of the torpedo tanks, then, “Surface!” and “Air in all tanks!”

  A few moments later, the submarine breached. At 5:22 she was caught in the glare of the Gleaner’s searchlight but then was lost when the light failed, owing to broken arc carbons. Price substituted the 10-inch signaling lamps and spotted the U-boat again. The Gleaner fired five rounds from the 4-inch gun and turned to ram. Dresky, who had emerged onto the conning tower, saw that the British ship was only a few hundred yards away and shouted down into his boat: “Abandon ship! Blow her up! Report [the situation] by radio!” Schilling shouted that the diesel engines were still working, but Dresky repeated his order. As the crew members clambered out, Schilling opened the main induction flappers, heard with satisfaction the crackle of the fuse for the scuttling charge, and opened the vents of torpedo tubes 1 and 2. But he could open the latter’s vent only part way. Someone called to him from above. He went to the conning tower and reported what was going on to Dresky, who came down, tried to open the vent of tube 2 without much success, and soon climbed back up. Schilling then went to the rear control room and opened both outboard valves until a heavy stream of water flowed over his feet. Satisfied, he went out, noticing that the main vent valves in the main control room were already under water.

  On the surface, Price saw the crew tumbling out of the U-33 holding up their hands in surrender, so he had the wheel put hard to starboard and all engines put full astern. The Gleaner stopped parallel to the U-boat about 200 yards away. Suddenly a shower of sparks erupted from the conning tower: the explosive had apparently set off some signal rockets and some of the smaller munitions; the heavier munitions were already flooded. The crew abandoned the submarine, which almost immediately sank by the bow at an angle of 40 degrees.

  As the stern disappeared, Dresky led the floating men in three cheers for the U-33. It was 5:30 A.M.

  The water was bitter cold and choppy; the night was black. Rottmann pulled his three rotors from his pocket and let them sink into the sea. Vietor did the same with his two. Dresky ordered, “Stay together as much as possible.” Schilling, swimming in a large group of crew members, was often asked whether they would be rescued. As the minutes and quarter hours passed, he repeatedly told the men “They’re on the way. Stick it out and keep swimming.” His pod kept getting smaller as men vanished. Meanwhile the Gleaner had lowered boats and moved slowly into the center of the largest group of survivors. Suddenly Schilling saw a searchlight nearby and then the side of a ship and a manila line, to which he clung before losing consciousness. He and the others had been in the water almost two hours. Only the many layers of their clothing had saved them. The Gleaner and other vessels picked up four officers and seventeen men. Dresky did not survive, and his body was never recovered. He had met his doom.

  Rottmann was picked up by a cutter, which took him to the Gleaner. He passed out, then woke up in a bathtub filled with warm water. The bearded Royal Navy ensign who was guarding him gave him pants and shoes. They spoke in German, and Rottmann asked whether he could visit his crew to see who was alive. The officer took him to a room in which a clump of men was sitting on the deck, each with a blanket, and shivering. On seeing Rottmann, Kumpf, the seaman who had been given three rotors to drop into the water, said, “Herr Oberleutnant!” and tried to rise to report in a proper way. But Rottmann put his hand on his shoulder and told him to stay down.

  “Herr Oberleutnant,” Kumpf resumed, “I forgot to throw the wheels away.” Rottmann went over to the bulkhead where Kumpf’s leather pants were hanging in the hope of finding the rotors so he could himself surreptitiously get rid of them. He squeezed the pants. They were empty. The British had gotten the wired codewheels.

  9

  ROYAL FLAGS WAVE KINGS ABOVE

  THE ROTORS THAT SOME ALERT BRITISH SAILOR HAD REMOVED from Kumpf’s leather pants went to Bletchley. Two turned out to be hitherto unknown naval rotors: VI and VII. Though in theory their wiring could have been reconstructed, their capture speeded the work of the Hut 8 cryptanalysts. Later they obtained rotor VIII from another naval capture. But even possession of a full set of rotors, together with a copy of the regulations for the use of the naval Enigma recovered from the U-13 in June 1940, did not make possible regular, or even frequent, solution of German naval cryptograms. And the U-boat war was increasing in intensity.

  For, as the usage regulations booklet showed, the Kriegsmarine keying system excluded the weak points in the Luftwaffe Enigma that the clever men of Bletchley had found and exploited, coining a vocabulary of colorful specialized terms as they did so. A “crab” was the simultaneous turning of two rotors, and a “lobster,” naturally, was when all three moved together. A “crash” occurred when the same letter stood in the same position in both a crib and a cryptogram, as der and DBV—an impossibility in an Enigma encipherment. The cryptanalysts of the Luftwaffe Enigma coined the term “cillies”—either the name of the girlfriend of an Enigma cipher clerk used as a key or just a burlesque of “sillies”—for some of the foolish things that Enigma operators persisted in doing, despite regulations to the contrary. One was to use as message keys a sequence from their keyboard, such as QWE or NBV, or the first three letters of a girlfriend’s name, or an obscene word. Another form of cilly occurred when a lazy encipherer chose as his message key the position that the rotors were in at the end of the encipherment of the previous message. Once the later message was solved, the cryptanalysts had merely to count the number of letters in the earlier message and turn the rotors backward that number of places to obtain the message key of the earlier message. Other cillies were the “nearness” and the JABJAB, so named from the indicator setting in which it was first seen.

  Yet another shortcut to solution was called the Herivel tip. John Herivel, an undergraduate in mathematics at Trinity, had been recruited by Gordon Welchman. He was fascinated by physics and by the way physicists made their great discoveries; he wanted to do the same for the Enigma. So each night he would return to his digs, figuratively put his feet up, and think about how the system could be broken. One day in February 1940 a “happy brainwave” suddenly came into his head. Suppose, he thought, an Enigma operator was lazy. In setting a key, he would hold a rotor with his thumb, grasp the spring for the alphabet-ring pin, rotate the ring until the key letter faced him, then release the spring to lock the ring in place. He would then slide the rotor onto the axle and repeat the process with the next two rotors. When the assembly was put into the Enigma, the three key letters would still be approximately facing the operator and would be close to the letters showing through the apertures when the lid was closed. The lazy operator would not turn the rotors but would use those letters for his message key.

  If the cryptanalysts thought an operator was following this procedure, they would have many fewer probable ring settings to try. Herivel’s colleagues thought the idea was brilliant. But it remained only a hypothesis until the Germans invaded France on May 10, when the German cipher clerks were under such pressure, because of the volume of communications, that they did not follow all their instructions in detail. Herivel arrived at work that day to see people clustered around his colleague David Rees. They were breaking the Luftwaffe Enigma using the Herivel tip. Welchman said to Herivel, “You won’t be forgotten!”

  More help came from two characteristics of German cryptographic procedure. The air force never used the same rotor in the same position two days in a row, except perhaps from one keying period to the next. This knowledge reduced the number of rotor orders that the bombes had to test. Moreover, the Germans never connected two consecutive letters on the plugboard—never A with B, for example. This enabled the British to add to the bombes a circuit called the CSKO, which, using the German term for “plug,” stecker, stood for “consecutive stecker knock-out.”

  Another aid to solution came, par
adoxically, from a German complication. Rotors I to V differed in the point at which the notch cut in the alphabet ring caused the rotor to the left to move one space. Rotor I, for example, kicked its neighbor when it (rotor I) stepped to ring position R (from ring position Q) during the course of encipherment. Rotor II gave its kick when moving to F, III to W, IV to K, and V to A. The Germans thought that this would create a more irregular movement, and in a way they were right. But the British cryptanalysts soon saw that the discovery that a rotor had moved could, if they knew the ring position of its neighbor to the right, identify that neighbor. (This is an interesting example of a complication that defeats itself. Had all the notches been cut at the same letter, cryptanalysts would not have been able easily to identify the rotor at a particular location—a point the Kriegsmarine recognized when it notched the rings at the same letters in rotors VI, VII, and VIII.) Because the different movement points were so important, Bletchley devised a mnemonic for them that made no sense but that stuck in the minds of the defenders of the monarchy: “Royal Flags Wave Kings Above.”

  With tricks like these, and with the standard analyses of keys and cribs, Bletchley began on May 22, 1940, to regularly read the Luftwaffe’s general Enigma key, called RED by G.C.&C.S., apparently from Gordon Welchman’s using a red pencil to underline discriminants and other data when he was first working on intercepts. Excitement mounted in Hut 3 as, day after day after day, RED fell before the British onslaught. To Dennis Babbage, a tall, slow-speaking algebraic geometer from Cambridge (who was, so far as he knew, no relation to Charles Babbage, the nineteenth-century father of the computer), the moment was unforgettable. “It was marvelous weather, the world crashing about our ears, France about to fall, England about to be invaded—and we made our first break into Enigma. It was exhilarating!”

  By then, Britain had long had little need of the Poles’ help. Indeed, during the phony war, from December 1939 to June 1940, the British with their greater resources had produced 83 percent of the solutions. Together, P.C. Bruno and Bletchley Park broke more than a hundred keys used on about a hundred days to read several thousand messages. These provided the Allies with such insights into German activities as that in a message of June 13: “Concerning directive of the 2nd Air Fleet for 14 June 1940: 1. Fourth Army attacks on 14 June 1940 over the Seine towards the southwest. Left wing on Chartres …”

  But the cryptanalytic successes could not, by themselves, produce military ones. As in the campaign in Poland, strength was needed to turn intelligence into victory. And France, for whatever reason, lacked it. On June 14, Paris fell, and on that day P.C. Bruno ceased to exist. The Poles and the Spaniards were evacuated to Oran, thence to the Villa Kouba, south of Algiers, and eventually back to the part of France not occupied by the Germans. In another villa, the Château des Fouzes, outside Uzès, a Roman town in the south of France, they rejoined the French cryptanalysts, still under Bertrand, to form a codebreaking unit that continued to solve Enigma messages and to radio the keys to Britain (enciphering them, subtly enough, in Enigma.)

  Here they worked somehow undisturbed for two years until, in November 1942, the Germans overran all of France. Most of the cryptanalysts escaped. But some key people were left in occupied Europe, REX, whose activities were known to German counterintelligence, was questioned; he cracked and disclosed the identity and treason of Schmidt, who was arrested and interrogated before he was shot in July 1943. But since neither man knew of the Polish work, much less of its success, they could not betray it. REX also told of his contacts with Bertrand, who was arrested in the white-domed basilica of Sacré Coeur in Paris. The Germans sought to get the officer to act as a double agent and find out, among other things, what German ciphers the British were reading. Instead Bertrand escaped, eventually getting to England. Ciȩżki and Langer were seized while trying to cross from France to Spain. Their prewar functions were known, and they were questioned specifically about Polish success with machine ciphers. But they convinced their interrogators, whom they suspected were German cryptologists, that the complications introduced into the Enigma encipherment before the war had made solution impossible. (It may have been what the questioners wanted to hear.) In Poland, the longtime chief secretary of the Biuro Szyfrów spoke only of superficial clerical matters when questioned about its work. Others—AVA technicians, Pyry deciphering clerks—never talked about the cryptanalysis. The secret of the Enigma solution was safe.

  The cryptanalysts who escaped—all except Różycki, who drowned when his ship, the Lamoricière, went down in a storm—reached England in 1943 after harrowing trips. The Poles reaped the customary reward of the innovator whose efforts have benefited others: exclusion. The British kept Rejewski and the others from any work on the Enigma, assigning them instead to a signals company of the Polish forces in exile, where they solved low-level ciphers. It was not one of Britain’s finest hours.

  The success with RED, which ran virtually unbroken from May 1940 to the end of the war, had little technical effect on Bletchley’s work with the naval Enigma: the keying systems were too different. But that success did have a strong psychological effect: it encouraged Naval Section. If the German air force Enigma could be solved, perhaps the German navy Enigma could be as well. A solution was badly needed. Germany was building more and more U-boats, and they were sinking more and more ships. Bletchley’s hope was further kept up by the bits and pieces of German cryptography that dribbled in from time to time.

  On the blustery morning of April 26, 1940, the British destroyer Arrow was returning to Britain. She had just carried men and supplies to Norway as part of the fight against the German invaders. At 6:50 A.M., three hours out and 60 miles northwest of the central Norwegian port of Ålesund, she sighted a trawler to the south and approached her. The trawler was flying a Dutch flag and had the Dutch colors and the word “Holland” painted on her side. She stopped when ordered but did not reply to other signals. The Arrow’s skipper decided to board her despite an extremely rough sea. As the Arrow was about to launch her boat, the trawler, about 300 yards to port, suddenly raced full ahead, swung to starboard, ran up the Kriegsmarine flag, and rammed the destroyer, holing her port side in the engine room about 4 feet above the water-line. The trawler’s crew jumped overboard, and the Arrow fired sixty-one rounds, hitting the trawler eight times. But it took the cruiser Birmingham’s bigger guns to sink the trawler, with four salvos.

  Upon reporting the sinking, the Birmingham was instructed to search in the vicinity for other ships. The Birmingham replied that the trawler was apparently only laying mines, but she followed orders. By then, the two warships had been joined by two other destroyers, the Griffin and the Acheron. At 9:55 A.M. they took station 4 miles to either side of the Arrow and headed approximately southwest at 15 knots. About 10:15 the Arrow sighted what appeared to be a trawler and ordered the Griffin to investigate.

  With her men at action stations, the Griffin closed with the other ship at 25 knots. Like the first trawler, this one was flying Dutch colors and had the Dutch flag of red, white, and light blue stripes painted on her sides. She wore the name Polares and “Holland” on her sides, and on her stern was a registration number from IJmuiden, a Dutch port. She stopped when ordered to do so, putting up no fight, despite her two torpedo tubes, which suggested that she was not just a fishing vessel. The Griffin sent over an armed boarding party in a whaler. While it was struggling in the strong breeze and choppy seas, a member of the trawler crew threw overboard two canvas bags. One sank at once; the other floated. As soon as the boarding party realized the trawler was German, a prize crew was dispatched. The bag had by this time floated near the destroyer, but it could not be grappled. It was on the point of sinking when Gunner F. H. W. T. Foord of the prize crew jumped over the side of the whaler with a heaving line and caught hold of the bag. While he was being hauled back aboard in the rough water, the line parted and he went under. He came up still clutching the bag. Another line was thrown to him. He held this for a mo
ment with one hand but could not hang on and went under yet again. Finally he managed to pass a bowline under his arms and was hauled aboard, saving the bag. The vessel, the former Julius Pickenpack, a 394-tonner built in Hamburg nine years before and converted to a disguised attack ship, was taken in tow.

  The bag proved to contain documents, among which were, in addition to key tables for non-Enigma systems, notes comprising Enigma keys for April 23, 24, 25, and 26. These found their way to Bletchley Park. Foord’s courage and alertness made possible the first break into naval Enigma. The slips of paper enabled Hut 8 to read some naval Enigma messages retrospectively in May for six days in April. But though this success expanded Naval Section’s knowledge of the Kriegsmarine’s signals organization, it neither affected naval operations nor made further naval Enigma solutions possible.

  * * *

  With the naval Enigma apparently impregnable, Hut 8 attacked simpler-appearing German navy cryptograms, hoping to produce at least some solutions. The characteristics of the cryptograms led the analysts to think that they were enciphered in hand systems: ciphers that employed only letter tables and pencil and paper. One system succumbed almost at once. This was Radio Cipher H, used by merchant ships controlled by the German navy. A bigraphic system, it was solved the month the war broke out by Dr. W. H. Bruford, a professor of German who had learned cryptanalysis in Room 40. Radio Cipher H rarely carried messages of significance. But in April 1940, when its reading was becoming current, a solution by two of Bruford’s acolytes was sent to the Admiralty to show what Naval Section could do. The message merely gave Oxelösund, Sweden, as the destination of a German freighter of moderate tonnage. But the solution caused a stir because it dented the pessimism that reigned in high quarters about the possibility of solving any German naval signals in time for them to be of use. It did not dispel all doubts, however, for a few months later, with the naval Enigma still unbroken and a new high-grade hand system, the Dockyard Cipher, likewise unsolved, Denniston, the head of G.C.&C.S., remarked mournfully to Birch, “You know, the Germans don’t mean you to read their stuff, and I don’t suppose you ever will.”