The Second World War

by John Keegan

  Dönitz, a First World War submarine captain, had recognised the mathematical disadvantage at which his naval arm operated and conceived a method to overcome it. By experimentation with surface torpedo-boats, during the period when Germany was denied U-boats by the Versailles Treaty, he demonstrated that ‘packs’ of submarines, if disposed in a chain on the surface where their speed exceeded that of merchant ships, could identify the approach of convoys across a wide band of ocean, be concentrated against one by radio command from shore and inflict mass sinkings by a concerted raid in numbers that would overwhelm the escorts. Once Germany acquired the French Atlantic ports, it was these ‘wolf pack’ tactics that were to make the Battle of the Atlantic the knife-edge struggle for advantage which overshadowed Churchill’s conduct of the British war effort from mid-1940 to mid-1943.

  Convoy, which was the Royal Navy’s defence against the wolf packs, offered only partial protection to the Atlantic lifeline. The naval escorts themselves – in the early days perhaps only two or three destroyers and a corvette were available to shepherd forty freighters and tankers across 3000 miles of ocean – were little direct threat to a determined U-boat formation. Asdic, the echo-sounder used to detect submerged U-boats, was ineffective beyond 1000 yards and reflected only range and bearing, not (until 1944) depth. The depth charges used to attack U-boats, triggered by water-pressure fuses, had to be set by guess and fractured the U-boat hull only if detonated close by. Most U-boat attacks, moreover, were delivered from the surface at night, when radar was more useful than Asdic, but until 1943 radars were too primitive to give early warning or accurate ranging.

  The radio intelligence war

  It was the measures taken to route convoys away from known or suspected U-boat patrol lines which best assured their safety, together with ancillary measures – particularly aerial patrol – to force U-boats to submerge while convoys passed by. Until May 1943 a shortage of aircraft and their shortness of range left an ‘air gap’ between North America, Iceland (available as a base to Britain after the German invasion of Denmark in April 1940) and Britain itself in which U-boats operated without fear of surveillance; the gap was closed when the very-long-range Liberator (B-24) with an endurance of eighteen hours came into service. Rerouteing, on the other hand, was a stratagem employed from the very beginning of the Atlantic war, and there was always a strong sense of direct conflict between the two sides. On the German side, the officers of the B-Dienst (Observer Service) used wireless intercepts and decrypts of cipher transmissions to establish convoy positions and read their orders; on the British (later Anglo-American) side, the cryptographers of the Government Code and Cipher School at Bletchley and the staff of the Admiralty Operational Intelligence Centre monitored the signals sent between U-boats and from Dönitz’s headquarters at Kernevel, Lorient (after March 1942 Berlin), to detect the formation of patrol lines and the vectoring of wolf packs against their targets. Rerouteing was by far the most successful of convoy protection measures. Between July 1942 and May 1943, for example, the British Admiralty and US Navy Department intelligence centres managed to reroute 105 out of 174 threatened North Atlantic convoys clear out of danger, and minimised attacks on another 53 by rerouteing; only 16 ran directly into wolf-pack traps and suffered heavy loss.

  The success achieved by Captain Rodger Winn, RNVR, and later by his American counterpart, Commander Kenneth Knowles, depended ultimately upon the skill of the Bletchley Park cryptographers in decrypting the Kernevel U-boat traffic fast enough for its significance to be applied to convoy operations. That traffic was, of course, enciphered on the Enigma machine, and the ‘Shark’ key used by the U-boat service proved particularly resistant to Bletchley’s efforts; it was not broken until December 1942 and then not regularly until 1943. Much of the vital radio intelligence used by the Operational Intelligence Centre up to that time was of lower-grade, position-fixing quality. High Frequency Direction Finding (HF/DF or ‘Huff Duff’) enabled ships to detect and locate shadowing U-boats from the transmissions they sent back to U-boat headquarters, and so for convoys to be rerouted or protecting aircraft summoned. Meanwhile, because of the Admiralty’s ill-advised persistence in the use of a book code instead of a machine cipher, the B-Dienst was able to read convoy traffic and direct wolf packs on to chosen routes with sometimes disastrous effect.

  The crux of this radio intelligence war began with the move of the U-boats from the eastern to the central Atlantic after April 1941. Substitution and rationing in Britain had allowed the import requirement to be reduced from 55 to 43 million tons, but the minimum level of subsistence was approaching and had to be measured against a rate of sinking which threatened to outstrip replacement building. In February 1941 the United States had enacted a Lend-Lease law which, in effect, allowed Britain to borrow war supplies against the promise to repay after victory; and from April 1941 the United States was operating a Neutrality Patrol which effectively excluded U-boats from the Atlantic west of Bermuda, under the terms of the Pan-American Neutrality Act of 1939. However, the U-boat fleet now had over 2000 miles of ocean in which to intercept convoys and was adding to its numbers at a considerably higher rate than it was losing U-boats on operations; during 1941 the building rate exceeded 200, while the total lost since September 1939 was less than fifty.

  The eight months of extended U-boat warfare in the Atlantic in 1941 therefore proved extremely successful to the German navy. In May it suffered the loss of the great battleship Bismarck, unwisely unleashed as a commerce raider, at the end of a great chase by most of the British Home Fleet; but that defeat was offset by the sinking of 328 merchant ships of 1,500,000 tons at a time when the British yards were launching less than a million tons of new construction annually. The casualties took down with them almost every category of material of which the home islands stood in dire need – wheat, beef, butter, copper, rubber, explosives and oil, as well as military equipment.

  Apologists for the British effort could show, however, that two-thirds of the ships lost had been sunk out of convoy and that U-boat losses for the year totalled 28, suggesting that the escorts’ success was increasing. Dönitz was certainly prepared to draw that conclusion: as soon as the United States Navy became an overt combatant rather than a hostile neutral (as it had been since September 1941), he transferred the weight of his effort to the United States coast. From January 1942 onwards, up to twelve U-boats were cruising off the American east coast and in the Gulf of Mexico at any one time; between January and March they sank 1.25 millon tons of shipping, equivalent to an annual rate four times higher than that achieved in the North Atlantic during 1941.

  By May, however, convoy had been introduced on America’s Eastern Sea Frontier and sinkings at once declined in those waters. Moreover, the rate of new building, both of merchantmen and escorts, began to accelerate spectacularly, as the American shipyards revived and were mobilised for new construction. Of particular importance was the appearance of a standardised tanker, the T10, and a freighter, the Liberty ship, both of which were larger (14,000 and 10,000 tons) and faster than their pre-war equivalents, besides – most important of all – being quick to build. Three months was the average construction time; by October 1942 American yards were launching three Liberty ships a day and in November the Robert E. Peary was built from the keel up in four days and fifteen hours – a public-relations stunt, but grim evidence to Dönitz of the challenge that prefabrication techniques posed to the efforts of his U-boat captains.

  The critical point

  By July 1942, though neither side yet perceived it, the Battle of the Atlantic was approaching its crux. There were diversions from the central issue. In March 1942 the British destroyed the dock at Saint-Nazaire which offered an Atlantic coast home to the Tirpitz, Germany’s largest battleship. It was then harboured in north Norway, where it had been joined in February by the Scharnhorst and Gneisenau after their daring dash through the Channel – an incident that provoked much recrimination between the Admiralty and the RAF
over who bore responsibility for the failure to intercept them. The three heavy ships levelled a menace against the Arctic convoys for months to come and caused lethal disruption to convoy PQ17 in July. In December 1943 Scharnhorst made a brief sortie against an Arctic convoy, provoking much anxiety at the Admiralty, but suffered an identical fate to that of Bismarck; and the Tirpitz menace did not end until November 1944 when she was sunk by bombing at her moorings in Troms| Fiord. But these episodes largely summarise the contribution of the German surface fleet to the Kriegsmarine’s war. The need to mount the North African ‘Torch’ landings in November 1942 temporarily drained away Allied merchantmen and warships from the Atlantic lifeline. There was a serious interruption of Ultra decrypts of U-boat radio traffic in February 1942 which lasted for most of the year, at a time when the B-Dienst was enjoying increased success against the Royal Navy’s no. 3 book code. The British, American and Canadian convoy control systems were simultaneously experiencing problems of ‘shaking down’ into a routine of co-operation; the Royal Canadian Navy, which was undergoing an expansion from six to nearly 400 warships in service, the largest of any armed force in any country during the Second World War, found particular difficulty in matching the expertise of its larger partners. Since the outbreak of the war the Admiralty and the Royal Air Force had been locked in a quarrel over the deployment of long-range aircraft, the Admiralty rightly but vainly arguing that convoy protection produced a better return of effort than spectacular but often ineffective offensive bombing of German cities. Against this background Dönitz was working to extend the range of his U-boats by experimentation with refuelling at sea from submarine ‘milch cows’, and to equip his boats for the increasingly dangerous surface transit of the Bay of Biscay against attack by such long-range aircraft as the RAF did allocate to the Admiralty via Coastal Command. In the first half of 1942 Coastal Command aircraft mounting the new and powerful Leigh searchlight began to surprise U-boats in the Bay at night and attack them with depth charges. Two were destroyed in July, although Dönitz had already ordered that they must make the passage submerged, despite the delay entailed in reaching the Atlantic cruising grounds. The significance of the Leigh Light was that it gave aircraft ‘eyes’ in the final 2000 yards of their approach when radar did not work. In a see-saw of technical duelling, its usefulness was to be reduced when the Germans learned to develop passive radar detectors which indicated danger before the Leigh Light could be activated. Over the course of the Bay of Biscay battle, however, which lasted into 1944, the advantage consistently returned to the Allies. It was not until the deployment of the first schnorkel-equipped U-boats (which could recharge batteries while cruising submerged) in early 1944 that the danger levelled by anti-submarine aircraft began to be offset.

  By then, however, the climactic phase of the Atlantic battle had come and gone. From July 1942 onward, when Dönitz at last achieved his target figure of 300 U-boats, he redeployed his effort to the central Atlantic, where the Allied escort fleet had been weakened by the transfer of British ships to help the Americans introduce convoy on the Eastern Sea Frontier. He was also now becoming more adept at organising patrol lines and concentrating wolf packs against convoys, and he was having greater success in locating convoys because of the advantage in cryptography that the B-Dienst currently enjoyed over Bletchley. The British responded with two experimental measures which would bear fruit later: the creation of a ‘support group’ of escorts to go to the rescue of a convoy under attack and the adaptation of merchant ships to fly off aircraft, the merchant aircraft carriers. The MACs however, proved clumsy forerunners of the true escort carriers, the first of which, USS Bogue, would not appear until the following March; while the persisting shortage of escorts would compel the 20th Support Group to be disbanded after two months.

  As a result, U-boat sinkings in the North Atlantic in November 1942 reached the total of 509,000 tons, a figure exceeded only once before in the previous May, during the ‘Happy Time’ off the American coast. Vicious Atlantic weather halved sinkings in December and January; but in February 1943, despite continuing bad weather, 120 U-boats sank nearly 300,000 tons in the North Atlantic and the toll seemed set to rise. During March, in a running battle against two convoys eastbound from North America to Britain, codenamed HX229 and SC122, forty U-boats sank twenty-two out of ninety merchantmen and one escort of the twenty which were defending them. The tonnage sunk, 146,000 was the highest in any convoy battle and led Dönitz and his crews to believe that they had victory in their grasp. Altogether they sank 108 ships in the North Atlantic in March, totalling 476,000 tons, the majority lost in convoy. Wolf-pack tactics, supported by the position-finding and decrypting successes of the B-Dienst, appeared to have achieved mastery over convoy protection.

  The appearance was illusory. Not only was the shipping replacement rate rising to meet losses (by October 1943 new construction had actually made good the amount of shipping lost since 1939, and built a superior merchant fleet into the bargain); U-boat losses were also beginning to equal launchings, at a monthly rate of about fifteen. Those statistics spelt doom to Dönitz’s effort. The explanation of the shift lay in several directions: Bletchley recovered its edge over the B-Dienst in May 1943, making rerouteing even more successful; escorts were becoming more plentiful, allowing the creation of permanent ‘support groups’, five in number in April; to the existing escorts were added two escort carriers embarking twenty anti-submarine aircraft which could force all U-boats in the vicinity of a convoy to submerge, thus effectively negating their offensive potential; improved radar, Asdic and depth-charge launchers (Hedgehog and Squid) levelled a direct tactical threat against the U-boat in close combat; but, above all, the increased availability of long-range patrol aircraft for the Atlantic battle wrought a strategic shift of advantage to the British-US-Canadian side. The long-range aircraft, particularly the Liberator bomber, equipped with radar, Leigh Light, machine-guns and depth charges, was flying death to a surfaced U-boat. In the Bay of Biscay – after a short and disastrous episode of ‘fighting it out’, ordered by Dönitz – the aircraft forced all U-boats to make their passage to the North Atlantic hunting-grounds submerged, at a fourfold time penalty; in great waters they completely disrupted Dönitz’s wolf-pack tactics by dispersing his patrol lines and savaging his concentrations wherever they appeared. In May 1943 U-boat losses, inflicted at a ratio of about 3:2 between aircraft and escorts, reached forty-three, which exceeded replacement more than twice. On 24 May Dönitz, accepting the inevitable, withdrew his fleet from the ocean, conceding later in his memoirs: ‘We had lost the Battle of the Atlantic.’

  That did not mark the end of the U-boat war. In May 1944 the first schnorkel-equipped U-boat made its trial cruise. The schnorkel, a retractable air-breathing tube, allowed a submarine to cruise submerged while using its diesel engines. The device, invented by a Dutch naval officer in 1927, anticipated the development both of the closed hydrogen-peroxide system, which the Germans would bring into service in 1945, and of nuclear propulsion, in that it transformed the submersible U-boat into a true submarine, capable of operating below the surface throughout an operational mission. Misused, it could kill a crew by suffocation, and two U-boat crews are believed to have died in that way; properly employed, it would have revived the U-boat threat. Had the Germans not lost their main Atlantic ports to the American army in August 1944, schnorkel U-boats would have reopened the Atlantic battle, to the Allies’ very great cost.

  Measured across the space of the Atlantic struggle, from September 1939 to May 1943, the cost can be seen to have fallen most heavily on Dönitz’s submarine arm. Although the Allies lost 2452 merchant ships in the Atlantic, of nearly 13 million gross-register tons, and 175 warships, mostly British (a term which included Canadian, Polish, Belgian, Norwegian and Free French escorts also), the Kriegsmarine lost 696 out of 830 U-boats dispatched on operations, almost all in the Atlantic, and 25,870 killed out of 40,900 crewmen who sailed; another 5000, plucked from the wreck
s of their depth-charged boats, became prisoners. This casualty rate – 63 per cent fatal, 75 per cent overall – far exceeded that suffered by any other arm of service in the navy, army or air force of any combatant country.

  The cost was certainly not in vain. Given that the economic odds disfavoured Germany from the outset, that its industry was organised ‘in breadth’ for a short war rather than ‘in depth’ for a long war, and that Hitler’s campaign of conquest was notably unsuccessful in adding either productive capacity or raw material resources to the Reich’s war-making capacity – it failed, for example, to acquire any large source of oil or non-ferrous metal ores for the German war machine – the delaying effect of the U-boat war on the transformation of Britain into an Anglo-American place d’armes for the eventual liberation of Europe may be seen as crucial. Moreover, while Germany fed itself easily between 1940 and 1944 from its own agricultural output and the requisitions made on farming in the occupied lands to the east and west, Britain was constantly held close to the level of minimum subsistence by U-boat depredations on its food imports. Rationing, though fairly applied and beneficial to the classes nutritionally deprived before the war, created a climate of latent crisis among the British which distorted and diminished their capacity to strike back at the enemy. Britain’s military threat to Germany during the Second World War was as intense as that levelled during the First though in relative terms Britain was not much weaker in 1940-4 than in 1914-18. It was the U-boats, marginally assisted by the Luftwaffe, which made the difference.