As the menace from Flanders receded, the burden of the campaign in British waters fell to the North Sea submarines based with the German surface fleet in the Ems, Jade, Weser, and Elbe estuaries. But here too more mines facilitated countermeasures. The Allies laid some 15,700 mines offshore during 1917; and some 21,000 during 1918, mostly by destroyers from the Humber. Extending more than a hundred miles out, the mined area became so vast and unpredictable that the Germans confined themselves to maintaining swept channels through it, but even these were repeatedly blocked. During September 1918 only seven submarines were escorted out of the Heligoland Bight, in three convoys. Increasingly the U-boats used the Kattegat, the strait between Denmark and Norway, which the British also mined though never succeeding in closing: but this route added 3–4 days to their travelling time and reduced their stay in the campaigning zone. The Allies also tried to block the U-boats’ exit into the North Atlantic by laying the ‘Northern Barrage’. This scheme – an American conception – was for a minefield from the Orkneys to Norway. It would replace the Tenth Cruiser Squadron, which had patrolled the wild northern seas since 1915, but had become a target for the U-boats while assisting little in their detection. An inter-Allied conference in September 1917 approved the plan: work began in March 1918 and by the autumn the British had laid 13,546 mines and the Americans 56,571, assembled at Norfolk, Virginia, and shipped across the Atlantic in special vessels. But so many American mines blew up or drifted off loose that laying was temporarily suspended and Beatty feared the North Sea would become too dangerous to navigate. The barrage was not completed until October, when the Norwegians submitted to Allied pressure to close a remaining gap by mining their territorial waters. It may have sunk six or seven U-boats, at a cost of $40m. Like the Otranto barrage it was an enormous diversion of resources, and was finished too late to make much difference.
More important than the Allies’ offensive measures was the submarines’ continuing lack of an answer to convoys. In British home waters, as in the mid-Atlantic and the Mediterranean, the system tightened its grip. More vessels were escorted to their assembly points, and local convoys introduced across the riskiest stretches: the Irish Sea in March and the English east coast in June. Round the shoreline, moreover, aeroplanes could operate, and by November 1918 the RAF was deploying 285 seaplanes and flying boats and 272 land-based aircraft against the U-boats, more and more of them carrying radios. In the last five months Allied aircraft sighted 167 U-boats and attacked 115, although sinking only two, in both cases aided by warships. Unlike in the Second World War, aircraft did not themselves destroy submarines, but they could spot them from a greater distance and observe torpedo tracks, and served as a deterrent: during 1918 U-boats pressed home attacks against only six convoys with air escorts. Like warships, aircraft achieved more when accompanying convoys than when engaged in general patrolling.
The argument thus far has questioned the effectiveness of the Allies’ offensive measures. Although they sank 69 submarines in 1918, compared with 63 in 1917 and 46 in 1914–16,74 new building kept the U-boat numbers at around the same level and the newly commissioned submarines were superior in performance and armament. The absolute numbers might suggest that the Allies were merely containing rather than reducing the threat. Nor did the Allies in the First World War achieve any breakthrough like that of May 1943, when the Germans called off their mid-Atlantic campaign after losing forty-one submarines in a month. The nearest equivalent was May–June 1918, after which Allied shipping losses in both the Atlantic and Mediterranean permanently fell off. The naval authorities were aware of the change, which the post-war official histories highlighted. However, it owed more to unspectacular modifications in convoying arrangements than to the fourteen U-boats destroyed in May, the highest monthly total of the war. Although Holtzendorff warned Wilhelm that losses were reaching a ‘dangerous magnitude’, he correctly saw the figure as freakish, linked to fine weather and ‘mirror-smooth’ seas that made the submarines more visible, and he foresaw that it would diminish.
What mattered more was that sinkings per submarine declined relentlessly: the U-boats based in the North Sea destroyed an average of 0.55 ships each per day in March 1917 but 0.07 in June 1918, or some eight times less.81 Even in the Mediterranean, German U-boats sank 802 tons per boat per day in April 1917 but 619 in July 1918 and 417 in August; in the Atlantic convoys loss rates almost halved between May to October 1917 and May to October 1918. Whereas the U-boats’ key objective was to destroy carrying capacity and cargo, the Allies’ was to shepherd the merchantmen safely to their destinations. The monthly totals of destroyed U-boats mattered less than that more cargo was getting through, and therefore that the Allies were winning. They prevailed as much through wear and tear upon the submarines and on their crews as through sinking them. By 1918 what the Germans had originally intended as a five-month all-out offensive had lasted four times longer, and at sea as on land men and machines were worn out. At the end of the war barely one U-boat in two remained available for front-line service, as they spent ever longer voyaging to and from their hunting grounds and rarely regained harbour undamaged. In the Mediterranean the facilities at Pola and Cattaro were overwhelmed with damaged submarines, and throughout 1918 some 30 per cent of the squadron was in dock for lengthy repairs, for which increasingly the requisite materials were unobtainable. Even in the better equipped yards in Flanders and Germany, the burden of repair grew heavier, and competed with the requirements of new building. Moreover, increasing numbers of submarines were needed to train replacement crews. The British Admiralty placed great stress on the U-boat crews’ demoralization and the killing or capture of many commanders. Being depth-charged was a terrible experience, with after-effects akin to those of shell shock on land. Interrogations of U-boat prisoners during 1918 suggested that many officers were still defiant, but the ratings were mostly raw and poorly trained conscripts, much more disaffected than in 1917, intimidated by the Allies’ countermeasures, and relieved to be out of the war. Loss of seasoned commanders was even more important, as their work demanded extraordinary courage, aggressiveness, and seamanship: out of some 400 captains only twenty had been responsible for 60 per cent of Allied shipping losses, and many had now gone. When in June 1918 British representatives negotiated an agreement for prisoner of war exchanges, the Admiralty protested that sixty-five submarine officers and men could go to neutral countries whence they could easily return to Germany, and ‘In carrying on this submarine warfare the greatest difficulty with which the enemy is now faced is the shortage of trained submarine officers and crews.’
As the U-boat offensive lost impetus, Germany’s leaders disagreed over how to revive it. Holtzendorff wanted an offensive in American waters, seeking softer targets and forcing Germany’s enemies to disperse their defences. The OHL, hoping for attacks on American troop transports, supported him. Reinhard Scheer, the High Seas Fleet commander, considered long-range operations would wear out the U-boats and sink fewer ships: he preferred to concentrate in closer waters. Moreover, Chancellor Hertling and the foreign ministry objected to proclaiming a Sperrgebiet (prohibited zone) off America’s east coast, fearing it would inflame US opinion and draw the country more whole-heartedly into the war. They feared a re-run of 1917, when they had caved in over unrestricted submarine warfare, only to find the navy’s optimism to be baseless. Wilhelm shared these misgivings, and ruled out a Sperrgebiet. Yet although the navy exploited the issue to blame the politicians for the U-boats’ failure, the submarines actually lacked the capacity to do much damage. Few of them could reach the American eastern seaboard, although U-151, operating there from May 1918 onwards, sank twenty-three ships and severed two underwater cables, while in June fear of submarine-launched seaplanes blacked out New York City, and six submarines stationed off the American coast between July and November sank ninety-three ships, including a cruiser. The Wilson administration responded by introducing coastal convoys but refused to recall American warships from European waters. It had expected such a campaign, knew when the U-boats were en route, and despite the public consternation believed that random patrolling in the Western Atlantic would be as ineffective as in the Eastern.
Around Britain, where the U-boats now faced coastal convoys protected by aircraft, they returned in summer 1918 to offshore operations. Their tactics in the final months foreshadowed those of the Second World War, with more attacks far out to sea at night (to avoid aircraft) and on the surface (for higher speed). The German Admiralty Staff had envisaged that group attacks might overwhelm convoy defences, but Holtzendorff accepted that only the North Sea U-boats (rather than the smaller and slower Flanders types) were suited to such tactics, and they would need time to become familiar with them. In fact the U-boats never achieved the co-ordination needed to break up convoys. For a fortnight during May up to eight submarines were stationed in the approach routes between Brittany and Ireland but during this period 183 vessels were convoyed safely inward and 110 outward, and only five (three of them in convoy) were damaged or sunk, while two of the U-boats were lost. Even in this relatively confined area, and in favourable weather conditions, the submarines had great difficulty in finding convoys, in massing attacks against them, and in landing torpedoes on their targets. In July a second attempt focused on the northern and southern approaches to the Irish Sea, but no more successfully.
The U-boats of 1918 had poorer radio communication than their Second World War counterparts, both for co-ordinating movements and for taking guidance from the homeland. Their numbers were much smaller, and their commanders warned Holtzendorff that to achieve much they would need more vessels.99 In March 1943, at the climax of the Battle of the Atlantic, an average of 116 submarines were at sea; in 1918 the average monthly figure was forty-five. The totals in commission were 132 in January (33 at sea), 129 in February (50 at sea), 125 in April (55 at sea), 112 in June, and 128 in September. Part of the problem was that too few had been launched earlier. After war broke out the Imperial Navy Office (Reichsmarineamt, or RMA) gave U-boats priority, but still completed two battlecruisers and two battleships between 1915 and 1917. It also built merchant ships, although German seaborne trade had almost entirely halted. On commencing unrestricted submarine warfare in 1917, the RMA had been so confident that it suspended further U-boat orders for fear of being encumbered with masses of redundant submarines after victory. In addition, it sacrificed potential gains from mass production by spreading orders between different yards and a variety of types, concentrating on larger and longer-range vessels that took longer to build. Throughout the war, only twelve U-boats were delivered by the contract date, and the lags grew longer. On 1 April 1916, twenty-five were over contract: on 1 April 1918, seventy. In May 1917 the RMA expected 142 new vessels to come into service by May 1918; in fact 93 did, so that the 1918 fleet was similar to that of 1917 instead of being substantially bigger. In December 1917 the navy tried to accelerate the building programme by creating a new U-Boots-Amt – ‘U-Boat Office’ – within the RMA. By the end of the war – the hubris of 1917 now long abandoned – it had contracted for 340 more vessels, none of which entered service. As of 3 August 1918, no fewer than 473 submarines were on order.
Extra procurement made little sense until the production bottlenecks had loosened. Under the U-Boots-Amt completions rose, but only slightly. Many factors accounted for the delay, including the rival claims of repairs and of surface ship construction, and shortages of coal, transport, and non-ferrous metals, but all involved agreed the critical scarcity was of labour. Yet when at a conference on 4 October 1917 the RMA chief pleaded for copper, lead, and extra labour to build 118 submarines for delivery in 1919–20, as well as for 7,000–8,000 workers to finish those under construction, Ludendorff objected both to transferring metals from the army’s weapons programme and to providing more manpower. The army, he said, had already released 250,000 soldiers to the home economy and could spare no more: the key requirement was to raise the productivity of the workers already in place. In ‘all probability’ the war would end in 1918, and he needed to rest his troops for the spring campaigning. Implicitly he discounted the Admiralty Staff’s argument that more submarines were an insurance policy in case the war on land went wrong: ‘the great and also only reserve for the eventuality that the military direction of the war should not – or only incompletely – achieve its objective’. Although Ludendorff agreed in principle that U-boats should be the highest armaments priority, ahead of aircraft, motor vehicles, and munitions, he would not relinquish soldiers or weapons from his projected offensive, and in June 1918 the OHL again refused men from the home or front armies for U-boat construction. Matters only changed – and far too late – after the July turning point on the Western Front put paid to Ludendorff’s hopes of a land victory, and after the naval leadership was reconstructed in August. Holtzendorff was removed while Hipper, the battlecruiser commander, replaced Scheer as commander of the High Seas Fleet, and Scheer himself became head of the SKL or Seekriegsleitung, a new overall naval command that moved out of Berlin to sit alongside the OHL at Spa. The Battle of Jutland had given Scheer some of the aura of victory that Hindenburg and Ludendorff owed to Tannenberg, and he and the aggressive young officers round him at once pressed for a massive expansion in submarine output. Assuming that the U-boats were now indeed Germany’s last means of applying enough pressure to achieve a favourable peace, Scheer wanted targets comparable to those attempted for land armaments with the 1916 Hindenburg Programme. He prepared a scheme to accelerate the completion of the submarines now on the slipways and to order more, with the aim of once again sinking ships faster than the Allies could build them. Monthly completions were to rise from 12.7 in October 1918 to thirty-six by autumn 1919, continuing at that level into 1920, while the construction workforce was to double from 70,000 to 140,000, of whom 17,000 were needed at once. Ludendorff, however, though recognizing ‘what extraordinary importance a stronger pursuit of the U-boat war possesses’, again said that neither the front nor the home army could release more men. This was the position when at the end of September the OHL decided to seek an armistice, although Ludendorff agreed that Scheer should continue developing his project in case negotiations failed, and only after Germany finally recalled all U-boats to port on 21 October was the programme abandoned. Intended to rally the public and redeem the navy’s reputation, it was overtaken by Germany’s collapse.
The Germans intended that the British should get wind of the Scheer programme, and in London the prognostications were surprisingly gloomy. In late September Sir Eric Geddes, the former railway chief who had become First Lord of the Admiralty, warned the Cabinet that because of the need to escort troopships the Admiralty had neglected anti-submarine warfare, with the consequence that U-boat numbers were higher than they had ever been, a great new submarine offensive was developing, and tonnage losses would again become ‘a menace to the Allied cause’. Frustration caused by failure to sink more U-boats partly explained this despondency, which came to the Cabinet days before it considered Germany’s appeal for an armistice. Yet even if the war had continued into 1919 and the OHL had released some of the workers Scheer requested (which Ludendorff’s industrial expert, Colonel Bauer, was willing to contemplate),118 the trebling in production the SKL envisaged was not remotely achievable. Nor would extra U-boats with unskilled and disgruntled crews have made much impact on the panoply of Allied defences. The crucial elements in Germany’s dilemma – the decline in the U-boats’ destructiveness and their inability to harm the convoys – continued to apply. On sea as on land, the options had run out.