Photo 13.1 Admiral Keyes Sea Power Centre, Australia
1918 Opens in the North Sea
THE NEW ADMIRALTY REGIME SET TO WORK. Admiral Sir Wester Wemyss seemed a new broom, but “there is no easy division between the work done by the Staff under Jellicoe and that done under Wemyss.”1 The basis of much of 1918’s operational effort was laid in 1917, even if the directing personalities were very different. One shift in relationships would prove important. Wemyss got on well with Beatty and was older (and once senior in rank), but he was junior as a flag officer and did not enjoy the same prestige. This made for a very different relationship than Beatty had with his predecessor. Further, Geddes had expended political capital, even with the prime minister’s support, in removing Jellicoe. He could not afford to repeat the move with Beatty, however difficult the latter became. The C-in-C was thus in a strong position. Beatty set his face against reductions in the Grand Fleet, particularly its destroyer force, which the naval staff sought to provide more capacity for antisubmarine work. Beatty had Wemyss’ sympathy in maintaining the Grand Fleet’s strength and readiness for action, but the First Sea Lord was always loath to overrule the C-in-C if he could avoid it.2 Beatty’s desire to preserve his fleet meant final rejection of Wemyss’ attempt to resuscitate the battleship force in the Swin. The King Edward VII class had been paid off and Dreadnought was sent to the Grand Fleet to replace the broken-down Superb.
Despite his insistence on maintaining the strength of his fleet, Beatty admitted the need for caution, particularly as Russia’s collapse ended the Germans’ three-front problem. The C-in-C would not seek battle at any cost but preserve the fleet as the covering force for the Allied campaigns at sea. British offensive measures would focus on the mining campaign. Beatty had legitimate concerns about meeting the High Sea Fleet in unfavorable circumstances, particularly if the Germans chose a moment when they could deploy their maximum strength. The Grand Fleet’s light cruisers and destroyers were in high demand, often for operations that meant they were not immediately available to support the battle squadrons. This was Beatty’s main worry, but he had others. British heavy shells remained unreliable and it would be months before the new armor-piercing units became available.3 Yet some of Beatty’s fears were created by overestimation of German shipbuilding. Beatty’s claims of his battle cruiser force’s relative weakness assumed that Mackensen was operational and other ships would soon be completed. The reality was otherwise. Mackensen was launched in April 1917, but work proceeded so slowly due to labor and material shortages she would still be more than a year from completion in November 1918. Graf Spee, launched in September 1917, was no further advanced.4 The battleship program was even worse off. The C-in-C’s estimates were consistently less optimistic than the Admiralty’s, whose reports were, if inconsistent, generally closer to the truth.5
The Admiralty started the year by adjuring units at sea to increase the destruction rate of U-boats. This was because Admiral von Capelle had admitted to the Reichstag that the monthly production rate was expected to be between eight and twelve boats. The Admiralty estimated that U-boat losses since September had averaged nine a month6 and urged the navy to exceed an average of 2.25 sinkings a week, emphasizing, “the more that this is increased the sooner will the end come.”7 The Grand Fleet itself got off to a bad start on 12 January with the wreck of the destroyers Opal and Narborough. Both were driven ashore on South Ronaldsay during a heavy storm. There was only one survivor. This followed the loss with all on board the destroyer Raccoon on the coast of Ireland only three days before.8 Beatty’s flag lieutenant noted, “Destroyers and light craft have had an awful winter, poor devils.”9 Nevertheless, despite continuing poor weather, the minelaying campaign accelerated in the new year. Operation A25, launched on 25 January, included a new twist, with Princess Margaret laying dummy mines in addition to a live field deployed by Abdiel and cruisers near Terschelling.10 The idea was that such mines would be detected, probably by aircraft, and divert German effort away from the real fields, while the British also hoped their use would disguise the extent of British knowledge of the German-swept channels.11 Given that minefields, if discovered, were generally left in place if possible, the dummy fields could also provide safe channels for the British while the Germans thought they were functioning as additional barriers. The Germans, who did much of their sweeping at night and did not collect many mines for examination, only discovered the stratagem after the war.12 Beatty’s concern over the diversion of light cruisers from their primary tasks and completion of the destroyers’ modifications brought the creation of a minelaying destroyer flotilla and an end to the routine use of light cruisers for minelaying. The C-in-C was not pleased to lose more destroyers from his escort force, but this was better than nothing. Operating from Immingham, the new Twentieth Flotilla included eight ships. The group was led by Abdiel, whose captain, Berwick Curtis, was the most experienced in this work in the Royal Navy.13
On 31 January, units from Rosyth sailed into thickening fog for exercises with the Grand Fleet. Accompanying the heavy ships were the newly formed Thirteenth Submarine Flotilla, led by the flotilla leader Ithuriel and including five K-class submarines, and the Twelfth Flotilla, comprising the light cruiser Fearless and four more K-class. As the force approached the Isle of May at the entrance to the Firth of Forth, the movement was straightforward until the Thirteenth Flotilla became entangled with a group of minesweepers. In attempting to avoid them, K14’s rudder jammed; within minutes her turn brought K14 across the bows of K22. Both boats were badly damaged. K14, despite being holed forward, got clear of the remainder of the force; but the K22 did not and found herself in the path of Inflexible, rear unit of the Second Battle Cruiser Squadron. Inflexible struck K22 a heavy blow that ripped off the submarine’s external ballast tanks but did not sink her. In an increasingly confused situation, the signal reporting the first collision did not get through to the flotilla’s commander until an hour later. By this stage, the battle cruisers knew there had been an accident and detached a destroyer to assist the crippled units. Unaware of this, Commander E. W. Leir in Ithuriel decided to turn his flotilla back to assist. This was a disastrous step in the prevailing fog. The turn put the flotilla in front of the battle cruisers and the formations behind them. Australia just avoided running down K12, but the Twelfth Flotilla was not so lucky. Fearless struck K17 amidships; the submarine sank eight minutes later. Although her crew got clear of their boat, they became victims of the melee that followed as the remaining submarines tried to avoid each other. K6 and K4 did not succeed and K4 went down with all hands. K7 scraped the sinking K4 and in maneuvering forced many of K17’s survivors under, drowning them. Only eight survived the night.14
The “Battle of May Island” struck a blow at the fleet’s morale, but also contributed to unease over the K class. The undoubted tactical promise for fleet operations of these difficult-to-handle boats was increasingly masked by accidents the class experienced. K13 was lost on trials (although salvaged and recommissioned as K22), while K1 was sunk after a collision with K4 during a sweep off the Danish coast on 18 November 1917. This accident was caused by confusion between different formations and machinery problems on board K1 due to salt water entering the fuel system. Had the incident occurred on the other side of the North Sea, the submarine could have been saved, but the risk of attack was too great and Blonde sank K1 with gunfire. Some of the K boats’ defects, such as their poor sea keeping, could be fixed, but developing the experience to operate such large boats with their complex machinery as integral components of a surface fleet would take more time than the war allowed. As the captain of K12 noted, the class was “built and designed for battle only and the type … had to share the fate of the Grand Fleet battleships which never got a fair chance at the enemy.”15
February found the Grand Fleet sailing into heavy weather in the hope of intercepting a reported sortie by the First Scouting Group. Conditions were so bad that twelve men were swept overboard, including one from New York. The inability to recover the man was another demonstration to the USN of the hazards of war.16 Delaware suffered a total power failure and substantial upper deck damage; several British ships were similarly battered. Fog also remained a hazard and the Sixth Battle Squadron soon had experience of its challenges. Approaching a rendezvous of his battleships with light cruisers before meeting a westbound Scandinavian convoy on 11 March, Admiral Rodman was faced by a bank of fog. He tried to alter course, but before the order could be executed, visibility dropped to almost nothing. With uncertainty about the force’s base course, “there were some very narrow escapes in collision.” No harm was done, but in the confusion, Texas, Wyoming, and Florida, as well as four destroyers, became separated. The squadron and its escorts did not reassemble until the next morning. Much had been learned, something apparent in Rodman’s frank admissions to the C-in-C.17
Operations in the Channel
Appointed acting vice admiral, Roger Keyes took up duty as Flag Officer Dover on 1 January. Left in no doubt about his position by the new First Sea Lord, it was Admiral Oliver who made the point most succinctly, telling Keyes, “It is up to you to deliver the goods.”18 The Flanders Command would not remain passive, particularly as Allied antisubmarine measures were proving increasingly effective. January was a bad month for the U-boats, with ten sunk or missing. Nevertheless, the command had to accept that the British mining campaign in the Heligoland Bight meant Flanders no longer had a high priority in the eyes of the High Sea Fleet. Admiral von Schröder was not given promised reinforcements, while units sent to Germany to refit were not returned, Scheer viewing his own needs as greater. The situation was not helped by the Allied bombing campaign, which became increasingly effective as the year went on, severely affecting the ability to undertake repairs—and get rest.19
Despite the success of the mining sortie on 22 December, the Flanders Command could not launch another surface operation until 14 January, when fourteen torpedo boats were dispatched to attack shipping, with the contingency of a bombardment of Lowestoft and Southwold if they found no quarry. Von Schröder wanted to maintain the pressure the Royal Navy was clearly feeling after the losses of the previous months. The British were alerted to the raid and Keyes naturally assumed the Dover barrage was the target. Matters were complicated by a British plan to lay a minefield north of Zeebrugge the same night. Keyes decided this should continue, as it could be placed across the line of German withdrawal, while he put to sea in the light cruiser Attentive. Accompanied by ten leaders and destroyers, he deployed his ships in two divisions, one at the eastern end of the barrage and the other to the west. The drifters that usually patrolled the barrage were sent inshore, while the remaining units were placed on alert but continued their antisubmarine work. Tyrwhitt, newly appointed an acting rear admiral, was also ready in Harwich.
German signals intelligence had detected that the British were expecting a surface attack and had reinforced the Dover Strait and emptied the Channel of traffic. The Germans still ascribed British foreknowledge to other causes than decryption but decided to make the best of the situation. Although there was likely to be no merchant traffic to attack, they would be operating well north of where the British expected. This maintained the chance of achieving surprise, not only with the bombardments—now the operation’s principal aim—but also in intercepting any local patrols. Weather once more decided events. By midnight a gale was blowing from the SSW, with limited visibility and the short, steep seas that made North Sea operations so difficult for small craft. None of the three German formations achieved anything of significance. Group 1, assigned to attack Lowestoft, found itself instead off Great Yarmouth and made the best of things by bombarding the town for five minutes, firing approximately fifty shells and causing four deaths. The submarine H9 was in harbor and got under way within fifteen minutes but found nothing when she put to sea. Group 2, allocated to attack shipping farther north, found none. Group 3 did not locate Southwold at all. Some of this formation sighted two enemy patrol vessels but lost them in the darkness before they could attack. In the foul weather, the British units saw nothing.
The Harwich Force sailed on receiving news of Yarmouth’s bombardment, while Keyes, realizing the attack was directed elsewhere than the Channel barrage, changed his dispositions to intercept the homeward-bound enemy. In these conditions, it was a forlorn hope and neither side made contact. So uncertain was the operational picture that Keyes believed the German flotillas “had come from the Heligoland Bight, and returned there.”20 They had not, but their passage to Zeebrugge proved fraught. Several units were badly damaged in the heavy seas, losing masts and gear. At the time, Tyrwhitt’s ships were making no more than ten knots and were hard put to avoid damage themselves.21 The British minelaying operation was undertaken in equally difficult conditions, with at least two of the minelayers unable to deploy their weapons, but bore almost immediate fruit when V67 detonated a mine. This blew off her bows and killed twelve of the crew. Despite the damage, the torpedo boat struggled back to Zeebrugge escorted by G95. Never again fully operational, V67 was scuttled when the Germans withdrew from Belgium later in the year. Keyes was inclined to blame a failure of intelligence, but his accompanying complaints about the poor weather were more apt.22
An inconclusive engagement took place on 23 January when four torpedo boats encountered the net vessel Clover Bank and her escort, the small monitor M26. The Germans mistook M26 for the much more powerful Erebus, which was also at sea, and kept their distance in hazy conditions, worsened by both sides laying smoke.23 Both thought they scored hits on their opponents, but no one suffered damage. Given the likely appearance of British reinforcements, the German torpedo boats did not linger. When the destroyers Melpomene and Marksman arrived on the scene, they found nothing. Keyes felt that the captain of M26 should have been more aggressive, but it is difficult to see what more a slow monitor armed with a single 7.5-inch gun could do against a half-flotilla of torpedo boats.24
The Dover barrage’s increasing success was reflected in the High Sea Fleet’s decision to allow the North Sea U-boats to take the longer passage around the north of Scotland. The Flanders boats were also finding the Channel transit increasingly dangerous and losses were mounting. If they needed to take the northern route as well, the advantage of forward bases in Belgium no longer existed. Something had to be done and a sortie by the Second Flotilla was organized by the High Sea Fleet. Emden led the destroyers out on 13 February, but heavy fog forced Commander Heinecke, leading the flotilla, to rejoin Emden and anchor. When the ships got under way again on 14 February, visibility was at the other extreme and Heinecke became concerned his passage was being reported by one of the many fishing vessels in sight. His outlook was not improved by the need to detach the lame duck G104, which had yet again developed condenser trouble. Heinecke nevertheless decided to continue. In the early hours of 15 February the eight destroyers, divided into two divisions, made their descent. The Fourth Half-Flotilla would attack the northern side of the barrage and the Third Half-Flotilla, the south.
The barrage was protected by destroyer patrols, while seventy-two trawlers, drifters, and other light craft patrolled the deep minefields and provided the illuminations to force U-boats to dive. A monitor was stationed near the Varne Shoal as heavy cover, although on the night of 14–15 February the guard ship was M26, not one of the big-gun units. The assembly was an inviting target. Bacon had specifically warned Keyes of its vulnerability to surface attack and Keyes later stated he was fully aware of the risks. Despite Keyes’ claims of a new approach, the night dispositions had not significantly changed from those ordered by Bacon. Although he asserted he had been successful in immediately instilling a new spirit of enterprise, Keyes failed to ensure the patrol forces were mentally ready for a German sortie, even though he expected the barrage’s success would “provoke reprisals before long.”25 His hope that Room 40 could provide warning of a sortie may have become an assumption that it would. Significantly, Keyes believed that such advanced notice would allow him to put to sea to take control of operations himself. This suggests he had little understanding of the need to be positioned to coordinate the responses of what were widely dispersed—and disparate—forces.
The Germans arrived at the barrage without triggering any alarms. Part of the British unreadiness for surface action lay in their focus on U-boats. Keyes had been warned of the passage of a crack submarine he was eager to destroy, and a possible U-boat was sighted just before midnight. The idea that this had triggered an engagement colored British appreciation of the situation. The Germans began their attack at 0030 when they detected the minesweeper Newbury. The ship was quickly disabled and left in what the Fourth Half-Flotilla thought was a sinking condition. Failing to work up minor war vessels after refit raised its head once more—Newbury had not sorted out her stores and the green flares to give warning of surface attack were not at hand. Confusion reigned among the British. Units did not interpret the gunfire they heard as a surface attack, thinking that it was coming from ashore or related to an air raid—or even the detonation of mines. The barrage’s lighting was another factor. The brilliant flares and illuminations meant that few personnel had any night vision.26 Most critically, the green flares that would have indicated the nature of the attack were not fired in the first encounters. The Germans were thus able to destroy a trawler and seven drifters and damage as many more. Both half-flotillas thought they had done twice as well, being misled by explosions of the magnesium flares on board their victims, which “gave the impression that the whole ship had blown up.”27 The Third Half-Flotilla was not as active as the Fourth, largely because G103 had also developed condenser problems, limiting the force’s speed. The Third Half-Flotilla’s onslaught eventually triggered the firing of green flares, but the British response remained disjointed. M26 moved to investigate, but her search was inconclusive and her captain did not report. Some of the small craft might have made the situation clear, but they also failed to make any coherent reports. The French did not do much better. The German southern force was seen by the small torpedo boats 344 and 350, but these were not close enough to attack, while it is doubtful they recognized the contacts as hostile. Neither they nor the shore station at Cape Gris-Nez, which also saw the Germans, made sighting reports.28 Although Keyes was alarmed by the repeated sound of gunfire and started to get the standby forces in the Downs under way, he gained the impression all was well and reversed his orders.
Worse followed. The Third Half-Flotilla was seen by the destroyer Amazon, rearmost unit of Termagant’s four-ship division. She repeatedly challenged but got no response and the strange ships rapidly passed out of sight. Constrained by the struggling G103, Kolbe would not risk a night action against destroyers and had turned away. Despite the lack of response, Amazon’s bridge team were convinced the ships were British and therefore did no more than pass a message up the line that they had seen three friendly destroyers. By the time Termagant questioned Amazon’s identification—a process slowed by a signals mix-up—it was too late. The Germans were clear away, although G102 detonated a mine approaching Zeebrugge. Patched up, she followed her consorts home a few days after they returned to Germany. The Second Flotilla had not remained in harbor at Zeebrugge for a single night, perhaps due to fear of air attack. The Germans had other concerns, particularly their machinery’s increasing unreliability. While they drew considerable satisfaction from the apparent success of the sortie, it would not be repeated by High Sea Fleet units. Believing the British were vulnerable to an immediate follow-up, von Schröder dispatched the Flanders torpedo boats on another raid the following night. They found nothing, despite Keyes’ declaration that the barrage was manned and fully illuminated. It has been suggested the Flanders’ units did not proceed as far southwest as Heinecke’s forces, while the barrage lighting seems to have been considerably reduced following the raid.29 Low visibility may have limited the horizon of both sides. The Channel at night in winter remained a difficult environment.
Admiral Keyes vented his wrath on those he considered negligent. He was backed by the Admiralty, and the captains of Termagant, Amazon, and M26 were superseded—despite A. A. Mellin of M26 having received the DSO and Bar and special promotion to commander for his work in Q ships. Perhaps these rigorous measures were needed to set an example. Keyes later railed at the role played in the defeat by unfortunate legacies of the Bacon regime.30 There was some justice in his claims, but systemic problems of readiness and training extended more widely than the Dover Patrol. Destroyers did not have enough practice in night warfare, with all its challenges and uncertainties. The RNR captains of small craft were splendid seamen but had little indoctrination in many aspects of operations that bore upon their naval tasks in ways they did not understand. Newbury’s unreadiness confirmed that small ships were not receiving the unit training essential on commissioning or after a long period in refit. Nevertheless, after six weeks in command—coming with the benefit of time in the Plans Division—Keyes must bear much of the blame. An attack should have been expected within days, and it was Keyes’ responsibility to ensure that the Dover Patrol understood this. Too much of his attention may have been on preparations for the attack on Zeebrugge and Ostend that was another element of his new regime. He had not thought through the implications for the Germans of the barrage’s success, as well as the extent to which the changes exposed his ships to a surprise attack on the surface. He certainly displayed no understanding of the problems of situational awareness in the illuminations’ glare.
The Attacks on Zeebrugge and Ostend
Keyes’ morale recovered quickly from the reverse, sustained by evidence of further U-boat sinkings and a successful minelaying sortie by four CMBs on the night of 7–8 February. Although two got lost and the other two were very nearly caught, the latter pair managed to lay mines off Ostend, one of which claimed the small torpedo boat A10. The admiral also had other fish to fry. An attack on Zeebrugge and Ostend had been debated over many months. Ideas of a direct descent were put in abeyance in favor of the divisional assault on the Flanders coast planned in 1917, but the land offensive’s failure brought them back into consideration. Different schemes developed by Tyrwhitt and Bacon had been considered and rejected. There was no consensus about the practicability of gaining surprise and achieving the main purpose of any attack, destroying the ability of the German light forces and U-boats to employ the canal and lock system between Bruges and the open sea. By September 1917, with Beatty expressing support, Jellicoe had authorized new Admiralty studies. Keyes’ arrival as Director of Plans injected, arguably, a much more positive outlook. On 3 December Keyes submitted a plan, which Jellicoe soon endorsed.
Preparations were overtaken by Keyes’ appointment to Dover. He quickly moved to revise key aspects. Keyes differed from Bacon in having less faith in 12-inch gun monitors and more in block ships placed to close the canal channels. There were two elements to this difference of opinion. Bacon and Keyes agreed that an assault on the mole that sheltered Zeebrugge harbor was essential to allow the ships attacking the canals to do their work. The canal at Ostend, being much more open to the sea, did not present the same problems. Bacon had intended a specially equipped monitor strike the Zeebrugge mole head-on, disgorging troops over a brow fitted for the purpose. Keyes’ objection was that the underpowered monitor was limited by its extra gear to four knots, inadequate in the current around the mole. His preference was to put a more maneuverable and faster ship alongside and disembark the force on as many gangways as possible. Keyes’ scheme also recognized that an assault force could be rapidly overwhelmed if German forces on the mole were reinforced. To prevent this, the viaduct that connected the mole to the land would be breached by CMBs filled with explosives.
Differences in the methods of closing the canals were more fundamental. Bacon believed destruction of the lock gates by close-range heavy gunfire would be more effective than block ships alone and determined that a monitor should bombard the locks from inside the harbor. Keyes was convinced that properly positioned hulls sunk in the channels could close them if enough block ships were assigned—three for Zeebrugge and two for Ostend. Filled with cement, the sunken ships would be hard to remove. The objections to this assessment were that the combination of the width and slope of the channels, the large tidal range (fourteen feet in spring tides), and the relatively shallow draft of their torpedo boats and submarines meant the Germans would soon be able to work around whatever obstacles were placed in their way. Keyes argued that the natural silting to which the canals were prone would rapidly build around the block ships, creating a problem impossible for the Germans to solve for many weeks.
The Admiralty considered the revised plan for Operation ZO at the beginning of March. Achieving surprise and disrupting any German response required not only the right combinations of moon and tide, but weather good enough to allow every unit, however small, to do its work. It was also key to have wind blowing in the right direction to deploy the smoke screens essential to mask the last stages of the assault force’s approach. Moon and tide created windows of only a few days each lunar month; sea state and wind would limit the opportunities even further. Much was hoped from the operation. The final comments by the Admiralty attached great importance to the moral effect of the venture inside and outside the navy. Bacon’s ideas of using army troops were rejected in favor of using RN and RM personnel, including large numbers from the Grand Fleet itself. This was a deliberate effort to find an outlet for the most enthusiastic (and frustrated) officers and ratings, who flocked to volunteer; the division between ships was made as even as possible, with HMAS Australia’s contribution of eleven being typical.31 Keyes had hoped to mount the attack in mid-March, but neither all the conversions nor the smoke devices were ready. The Admiralty became particularly anxious the element of surprise would be lost, since it was difficult to conceal the preparations from German aerial reconnaissance.
Flanders Command did not detect the buildup of British forces, perhaps due to its own preparations to support the great offensive the German army was about to initiate, which included diversion of its aircraft to scouting and combat farther inland. A preliminary reconnaissance of the French coast was conducted by torpedo boats on 19 March, two nights before the offensive’s planned start. Early on the 21st, the Germans sailed in three groups and began to bombard the French coast as firing began on the western front. Five boats under Lieutenant Commander Assmann attacked Dunkirk at 0348, but rapidly became aware that there were strong Allied forces in the vicinity. The monitors Terror and M25 opened fire in response and the explosion of Terror’s star shell above his ships convinced Assmann that immediate withdrawal was necessary. Collecting two smaller torpedo boats positioned as navigational markers, Assmann made for home. Steering to cut him off were the destroyer leader Botha and four British and French destroyers that had slipped their cables in the Dunkirk anchorage on hearing the firing. The two formations made contact in mist at approximately 0440, Botha confirming with star shell the Germans’ presence on her port bow. At close range, the two sides exchanged fire with main armament and light weapons. Botha suffered a pierced steam pipe. With his speed falling off, Commander Rede turned toward the German line and rammed A19, cutting her in two. An attempt to follow up by ramming A7 failed, although Botha’s gunfire quickly reduced the little torpedo boat to a wreck. Torpedoes fired by Assmann’s units did not find their mark, but their smoke screen proved more effective. Not only did Assmann’s five bigger boats make a clean break, in the confusion the French destroyer Capitaine Mehl mistook Botha for the enemy and put a torpedo into her as she emerged from the smoke. Holed in her boiler room, Botha came to a halt. Since the Germans had disappeared, all the Anglo-French force could do was finish off A7 and escort the crippled flotilla leader back to Dunkirk. CMB20 spotted the German force and succeeded in getting a torpedo away, despite being under heavy fire. The CMB’s crew thought they hit the fourth in line, but this was not so. The other German bombardments did not achieve anything significant, although Commander Albrecht’s units covered the last stages of Assmann’s withdrawal.
Botha’s crippling was an embarrassing end to an otherwise encouraging engagement and a reminder of the conclusion drawn by a French officer, “Numbers are often a disadvantage at night.”32 As Morris had also fired a torpedo at what she thought was an enemy, it took the discovery of “a piece marked Creusot” inside Botha’s boiler room to confirm who had done what. Keyes worked hard to patch up the Anglo-French relationship.33 This was already under strain, since a week earlier the French airship AT-0 had mistaken D3 for a U-boat off Dieppe. When the airship failed to recognize the submarine’s recognition signals and closed to attack, D3 dived, but not quickly enough. In one of the very few successes of the air against submersibles, six 114-pound bombs sank D3. The submarine broached before finally sinking, leaving four survivors in the water, but all drowned before they could be rescued.34 In the meantime, the admiral was determined the enemy venture would not pass without a response, particularly as the Allied armies were retreating from the devastating German offensive. On the afternoon of 21 March Terror bombarded Ostend, putting thirty-nine rounds into the waterfront, withdrawing only when the German shore batteries became too accurate. Although no ships were hit, the accuracy of the shooting from both monitor and batteries confirmed how far fire control techniques had developed with the help of air spotting, sound ranging, and increased navigational accuracy. Over the next few weeks, despite their desire to support the land offensive, the Flanders surface forces were constrained by the need to keep the approaches to Ostend and Zeebrugge clear for the U-boats, as well as serviceability problems and continuing air attacks. Later sorties to bombard the French coast in April achieved nothing more than the first effort, although the Germans avoided contact with Allied forces at sea. For its part, the Dover Patrol was not only maintaining the barrage and preparing for the attack on the enemy ports but also protecting the flow of reinforcements as the British combed the United Kingdom for troops to shore up the front. The Admiralty, fearing the prospect of the loss of the Channel harbors—and perhaps the entire coast of northern France—to the German offensive began to prepare for the destruction and blocking of the ports concerned.35
The old cruiser Vindictive was selected for the attack on Zeebrugge mole and refitted to carry gangways and additional light weapons to suppress the defenses. The Mersey ferries Iris II and Daffodil were also requisitioned to carry the remainder of the force assigned to the mole. The old Apollo-class cruisers Intrepid, Iphigenia, and Sirius were allocated to block Zeebrugge, while Brilliant and Thetis were designated for Ostend. Instead of CMBs, two C-class submarines were assigned to blow up the viaduct, each being packed with five tons of explosives. The survivors of A7 and A19 also played a part. Their interrogations provided useful details of the arrangements at Zeebrugge.36 The combined Royal Marines–Royal Navy unit formed to attack the mole received intensive training, but they lacked battle-experienced officers and their preparation was unsophisticated, emphasizing hand-to-hand combat (bayonet drills) that seemed out of place in 1918.37 The landing force joined the assault ships in the Swin anchorage on 6 April to wait for the first attack window. This opened on 11 April and the force got under way at 1600. The ships successfully moved along the succession of buoys laid to ensure there were no mistakes of navigation; but at the position from which the attack force would separate from the covering forces, Keyes detected the onset of a southerly wind that would make the smoke barrage ineffective. As the covering smoke was fundamental to a successful attack, Keyes had to order a return to the anchorage. CMB33 may not have received this order, as it approached Ostend too closely and went aground. CMB18 also did not survive, being run down and sunk as the force turned for home. Keyes was initially undismayed, but poor weather forced him to turn back once again on the 13th, the last day of his window for April. Faced by the probability the Germans would realize what was in train, a probability that could only increase, Keyes rethought his preconditions. As he argued to the First Sea Lord when Wemyss visited him on the 14th, if the smoke were as effective as expected, even a full moon would not make much difference to the Germans’ ability to see what was going on. The attack could therefore take place on the next spring tide—only nine days away. Wemyss had intended canceling the operation outright, but accepted Keyes’ proposal and with it ultimate responsibility for the raid.38
The Germans quickly salvaged CMB33 and found material that made it clear Ostend had been the intended target of some form of attack. That a descent on Ostend could be combined with a simultaneous venture against the more difficult target of Zeebrugge did not occur to Flanders Command, possibly because of the scale involved. Furthermore, von Schröder’s resources did not permit him to put constant night patrols to sea. Uncertain what the British intended, the admiral’s response was to alert the coast defenses and rely on his artillery as the first response.39 Concerned to keep the Germans off-balance and inure them to offensive British activity, Keyes ordered a diversion for the night of 17–18 April when all three 15-inch gun monitors and the 12-inch gun Prince Eugene bombarded batteries to the west of Ostend.
On 22 April, Keyes tried again, aware that the new window almost certainly represented his last chance. More than 160 ships, big and small, were under way by late afternoon. It is possible Keyes would have gone ahead whatever the weather and conditions at first favored him, the large force disguised by mist and squalls, the wind in the right direction, and the sea low enough to allow the passage of small craft and ships under tow. The one problem was the weather did not allow the planned air raid against the enemy ports. Movement along the line of illuminated buoys went without a hitch. C1, one of the two submarines designated for the viaduct, had to drop out, but the remainder of the force went on. Deployment of the smoke screen started at 2130 and continued over the next few hours. Just before 2330, Erebus and Terror began their bombardment. This was key to the deception plan, even more important because of the absence of bombing aircraft. In addition to the damage that the 15-inch shells could do, the barrage forced the German gun crews and garrison troops into shelters, reducing their readiness for an assault, while the more routine nature of bombardment by monitors lulled them into thinking this was the main event for the night.
The British achieved almost complete surprise at Zeebrugge—the key word being “almost.” Although the smoke screens successfully concealed the force until the final stages of the approach, a wind shift just before midnight revealed Vindictive three hundred yards off the mole. The Germans, suspicious enough they had already closed up at their fighting stations, were ready to respond. Despite the efforts of CMBs to suppress the defenses of the mole and disable the torpedo boat V69, berthed on the mole’s inshore side, Vindictive came under heavy fire. Critically, positioning the old cruiser correctly in the strong current proved as difficult as some of the senior officers who reviewed the scheme had feared. Her captain, Alfred Carpenter, was a specialist navigator, but he could not avoid placing Vindictive three hundred yards farther along the mole than planned. An intense engagement ensued in which the British were hamstrung by problems disembarking from the cruiser, the increased distance the assault parties had to cover to reach their objectives, and the difficulty of suppressing the enemy’s fire. The battle included an extraordinary duel over the mole between Vindictive and V69, which culminated in the German disabling Vindictive’s fighting top on her foremast, the only weapon position in the ship with the height to enjoy a clear field of fire. In the heavy current, Vindictive was kept alongside by Daffodil pushing her bodily against the mole while Daffodil’s assault parties used the cruiser as their bridge ashore. Iris II was berthed farther along the mole but suffered so many casualties and had so much trouble getting her people onto the mole that she finally repositioned alongside Vindictive. The one clear success for the British was C3’s successful breach of the viaduct. Contrary to instructions, Lieutenant Richard Sandford and his crew stayed with their boat until she had rammed the viaduct under full power. They set fuses and got away in a small skiff under heavy fire. Although they were still within a few hundred yards when the charges detonated, the explosion’s effect on the German defenders gave the crew enough breathing space to be rescued by a motor launch (commanded by Sandford’s brother Francis).
Meanwhile, the first block ship, Thetis, was approaching. She entered the inner harbor but became entangled in the protective nets. While she succeeded in clearing a path for her consorts, the nets fouled Thetis’ propellers. Losing way, the cruiser could not enter the canal. Commander Ralph Sneyd ordered Intrepid and Iphigenia to pass ahead, waiting until they were clear before he swung the ship across the channel for scuttling. Thetis’ task had been to ram the lock gate, probably the only practicable way of achieving the blockage the British sought, but no contingency orders were in place. Intrepid therefore continued her mission of blocking the entrance channel. Lieutenant Stuart Bonham-Carter positioned his ship well, grounding her by the stern. Iphigenia, which followed, tried to fill the remaining gap between the Intrepid’s bow and the side of the canal. The effect was to push Intrepid into a position less directly across the line of the channel, while the Iphigenia drifted back to seaward before scuttling charges took effect. The result was that the ships lay at angles across the middle of the channel, impeding passage but not preventing it completely.
Action around the mole continued. The destroyer North Star entered the inner harbor to attack any targets she could find. Subjected to devastating fire as she withdrew, North Star was left stopped and burning only a few hundred yards off the mole. The destroyer Phoebe, covered by the smoke screens that were again proving effective, labored for nearly an hour to tow North Star away. She failed at this task but succeeded in taking off all but a few of the crew. This incident may have provided a helpful diversion for the withdrawal that began just after 0100. Despite the hazards of being alongside the mole, moving away from its shelter put Vindictive and the two ferries under even more intense fire from the surviving German positions. Iris II in particular suffered heavy casualties. ML558 did extraordinary work in covering her with yet another smoke screen. This was typical of the vital part played by small craft in the operation, not least in their protracted efforts to rescue survivors from the various elements of the force.
Despite the casualties, the Zeebrugge operation appeared to have achieved its goal. The assault against Ostend was less fortunate. The change of wind that disrupted the protective smoke screens occurred at an earlier stage in the operation, leaving the block ships exposed. The final stage of the approach and the turn toward the canal entrance depended on the Stroom Bank buoy outside Ostend as a marker. Although the British had designated a motor launch to stand by the buoy as an additional mark, they were unaware the Germans had shifted it as a routine precaution only the night before. Commander Alfred Godsal could not see the low-lying shore in the mix of smoke and mist and turned more than a mile short of his target. Sirius and Brilliant ran hard aground a few minutes later. Although they were lying under the German fortifications, the evacuation of both ships was achieved by a group of motor launches with relatively few casualties. Here the presence of Marshal Soult and three 12-inch gun monitors may have been critical. Anchored offshore, they exchanged fire with the German coastal batteries over several hours. Although their initial targets were the heavy batteries, they shifted to take on the lighter guns that were the principal threat to the withdrawal. The monitors did not do much damage, nor did they suffer from the German return fire, but they certainly provided some of the diversion required.
The attack on Ostend was obviously a failure, but Keyes never wavered from his claim that Zeebrugge had been successfully blocked. Room 40 rapidly developed a contrary view, while aerial photography and local intelligence also confirmed the Germans could get their coastal U-boats and smaller torpedo boats around the sunken hulls. The bigger units made a transit on 14 May, although they had to do so at high tide. Perhaps more inconvenienced was the seaplane force, which could no longer move its aircraft by rail to enter the water by crane from the mole and had to use the weather-bound beach.40 But Keyes was correct in his appreciation of the enormous effect on morale within the Royal Navy and in the country and allied nations. The operation was in the finest traditions of the Royal Navy and it came at a time when the war news had been bad for many weeks. The affair restored the navy’s lagging confidence, as well as much of the confidence of Britain.
The Travails of Flanders Command
The psychological effect on the Germans was also significant. It took time to determine the extent of the damage and von Schröder was criticized for the absence of night patrols. He pointed out how few resources he had to meet his tasks and his resultant inability to divert units to guard against such contingencies. Minesweeping had to take priority if the U-boats were to continue operating from Flanders.41 The admiral’s litany was justified. In addition to losses in the mine fields, air raids in March and April had damaged three torpedo boats, as well as UB30. The Flanders Command was slowly buckling under the combination of Allied pressure and its diminishing resources. When the Admiralstab did permit the return of the refitted quartet of the Sixth Half-Flotilla, the operation did not go well. The torpedo boats sailed on 19 April preceded by minesweepers. Their passage proved difficult, dogged by the discovery of minefields. It became clear that the force could not get far enough along the coast to make the final approach to Zeebrugge at night. Lieutenant Commander Gautier decided to turn back, but his sweeping force found itself unexpectedly entangled in yet another minefield. After M95, M39, and M64 struck mines and sank in rapid succession, Gautier’s only course was to anchor and ask for additional minesweepers. These arrived the next morning and began to clear a passage.
The Harwich Force had intelligence of German sweeping activity that was confirmed by flying boat reconnaissance. What was now designated the “Striking Force” was dispatched. The day was one of “marvellous visibility” and Gautier sighted the British early enough to get his half-flotilla underway and send the sweepers east under cover of a smoke screen. His torpedo boats proceeded to lay smoke as five British destroyers led by Thruster approached. A long-range engagement with heavy expenditure of ammunition ensued. The British labeled Gautier as “unenterprising” because he maintained the range at over eight thousand yards and withdrew as soon as the minesweepers were safe.42 The British seemed just as cautious, but the destroyers had already crossed the limiting east longitude that Tyrwhitt had ordered because of the British A36 minefield.43 The light cruisers covering the destroyers to the west could not intervene, while the destroyers did not follow the Germans farther because they believed they had “sighted their backing-up force.” This was the Fourth Scouting Group, but its advance west was halted by urgent advice from the minesweepers.44 The British were fortunate this was the case, since the destroyer Sturgeon was disabled by a 105-mm hit in her engine room and had to be taken in tow. Within the Harwich Force Tyrwhitt was highly critical of the destroyers’ “foolhardy action” in crossing A36 but covered for Thruster’s captain in his report: “He did all that could be done, the extraordinary visibility defeating any chance of surprise.”45 The naval staff’s post-action analysis did not miss what had happened, however, and the Admiralty called Commander George Gibbs to account. To his great credit Tyrwhitt closed the matter, pointing out he had already dealt with Gibbs and “cautioned” his commanding officers.46 The minefields that littered the southern North Sea had once more proven to be an overriding constraint. The incident confirmed another development. An increasingly active naval staff in the Admiralty was conducting more sophisticated analysis of operations than had been the case earlier in the war. This was both necessary and correct, particularly as it would help identify deficiencies in information flows and coordinating remote operations that had dogged the British since 1914. Nevertheless, it must have been galling to officers like Tyrwhitt, accustomed as they were to operational autonomy. A note of black humor was sounded after the war. During the encounter, the Germans were puzzled by the British advance across one of their own minefields. When they discovered the existence of dummy mines, the explanation seemed to be that A36 was not a live field, but it was.47
The Second Assault on Ostend
Keyes was determined to renew the assault on Ostend. He still had a few days of the right tidal conditions and the battered Vindictive was available for conversion to a block ship. A galvanized Dover Command and support units patched up the old cruiser and filled her with cement in time for an attempt on 27 April. Keyes understood the heightened risks of this second venture and insisted on the smallest crew possible on board Vindictive.48 Worsening weather forced another delay. This allowed the depot ship Sappho, another elderly cruiser, to be converted as an additional block vessel, but also meant the raiders had to accept a shorter period of darkness to cover the attack and withdrawal. The operation went ahead on the night of 9–10 May. Sappho had to drop out with a boiler defect soon after the force sailed from Dunkirk, leaving Vindictive to continue alone. The preliminary bombardment began just after 0140, this time supported by a bombing raid. The final approach went well until a thick fog suddenly descended, forcing Vindictive to grope her way along the coast while the German gun batteries, alerted by reports from torpedo boats picketed to seaward, searched for their target. As a last resort, the British illuminated the canal entrance to show Vindictive the way, but this also made the cruiser visible to the shore defenses. Under intense fire, Vindictive (whose captain, Alfred Godsal, was dead) approached the entrance just before 0200 and grounded by the bows on the eastern side of the channel. The plan had been to ground by the bow on the other side, allowing the tide to swing the stern across the passage, but Vindictive finished up practically parallel to the side of the canal. The blocking attempt had failed again.
Two motor launches extracted Vindictive’s survivors under a hail of fire. ML254 was so badly damaged that she only just made it to Keyes’ flagship, the destroyer Warwick, lying offshore. Keyes had brought the ship close into the coast to look for any “derelicts.” This was a risk justified by the results, but just after Warwick began her passage home, the destroyer detonated a mine. The ship’s back was broken and her stern only held on by the mine rails fitted over the quarterdeck. The struggle to save the crippled ship was an anxious one. The disabled Warwick was first towed alongside by Velox and then by Whirlwind. Even after Keyes’ ships had got out of range of the German batteries, there remained the possibility of a surface attack. The British believed a flotilla of High Sea Fleet destroyers had been dispatched to Flanders, although their intelligence may have confused the aborted return of the command’s own units, which resulted in the action of 20 April. Delayed by weather, these did not arrive at Zeebrugge until 12 May.
Keyes would not give up and developed yet another proposal to block Ostend, this time using Sappho and the stripped and lightened pre-dreadnought Swiftsure, whose armor and size would make her a much more formidable proposition for the defenses, as well as a substantial obstacle to place in the canal. The operation, however, was canceled by the Admiralty just before it was due to take place in June. The personnel concerned were “bitterly disappointed”49 but the British no longer believed the Bruges-Ostend canal justified the casualties inevitable in a third attack, while they knew the Germans had laid additional mines around the approaches to the port.50
The focus of British offensive effort was shifting in any case. Although Keyes became intensely frustrated by his reduced control over the air elements that followed the establishment of the Royal Air Force on 1 April 1918 and the incorporation into the new service of the RNAS, the bombing campaign against the Flanders bases was increasingly effective.51 One of the Zeebrugge lock gates was put out of action for a week at the end of May, while several torpedo boats were severely damaged in the same month. Further, continuing improvements in range-finding techniques meant the monitors were more effective than ever. On 9 June, Marshal Soult and Terror bombarded Zeebrugge. Their primary targets were the dredgers attempting to widen the channel, but they achieved the hit that Bacon had sought for so long when the newly repaired lock gate was again put out of action. This time the caisson was so badly damaged the canal was unusable for a month.52 Driven off by the increasingly accurate coastal batteries, the British do not seem to have realized their achievement.
The Dover barrage’s consolidation and perfection of the night illumination organization completed the effective closure of the strait to U-boats, adding to von Schröder’s woes. The Allied air offensive could be met in part by reinforcement aircraft and the reallocation of squadrons already in Flanders; but the Germans were finding it increasingly difficult to maintain mine-free passage for the U-boats, and increasingly active British and French surface forces required the submarines to have escorts. The problems with the canals and the cumulative effects of the bombing campaign meant the Flanders Command was hard put to field enough operational minesweepers and torpedo craft, while some of the small A boats had to take on minesweeping to compensate for the losses of more suitable units.
Sortie by the High Sea Fleet
The High Sea Fleet was also feeling the pressure. The light cruiser Stralsund had been severely damaged by a mine in the Heligoland Bight on 12 February, while A57 and A56 were sunk by mines early in March, with M91, M36, and M40 falling victim later in the month. On 30 March, G87, G93, and G94 were caught in a minefield laid only the previous day and sank within less than an hour of each other, highlighting the risks of supporting the minesweeping operations. Scheer determined that a new northerly mine-free route should be established. This could be done, but it would only be a matter of time before the British deduced the existence of another passage and turned their attention to it. The one active step was to embark on a new program of defensive minelaying. During April and May, five thousand mines were deployed around the swept channels, placed across the expected approach axes of the British minelayers. It would be some time before this effort paid dividends.
The navy was also under pressure to do what it could to support the Army’s offensive on the western front. The Handelskrieg was not going well. While the loss rate of U-boats dropped sharply in February and remained relatively low at six boats in both March and April, the average tonnage sunk by each unit also declined. The requirement to pass around the north of Scotland was one factor, so was the difficulty in maintaining operational availability. Finally, although there were still targets to be found, convoy was proving a significant challenge. The army’s high command had lost faith in the ability of the U-boats to end the war but appealed for a special effort against the shipping carrying reinforcements to France. This was easier said than done, but Scheer believed the High Sea Fleet could strike a heavy blow. The possibility of attacking in the south was canvassed, but Scheer found the kaiser would not allow the risk. The admiral was particularly critical of the Admiralstab’s influence in this decision, contributing to his own belief that the German navy should be put under one commander alone. Despite the closer proximity of the Grand Fleet, however, a sortie north was acceptable to the kaiser and the Admiralstab. The British had increased the interval between the Scandinavian convoys until by April they ran every five days. This made their protection less onerous, but also enlarged the convoys and thus their attractiveness as targets. Scheer’s staff had gathered intelligence about the operating cycle from U-boats, although not apparently from local agents in Norway.
The plan was for von Hipper to attack the convoy with his battle cruisers. The Germans were aware that a division or squadron of the Grand Fleet’s capital ships would be at sea as cover. If Hipper were unable to brush this force aside, he could call on the support of the complete High Sea Fleet, which would follow him into the North Sea. Poor weather forced the cancellation of a sortie planned for 10 April, but the raiding forces finally sailed early on 23 April. Although briefly delayed in the Heligoland Bight by fog, the High Sea Fleet was on its way by daybreak. German communications security on this occasion was excellent. Scheer and his admirals had at last stopped using radio to pass preparatory messages, adopting the British practice of passing such directions by hand. Security was helped by the recent issue of new call signs and a change in the cipher key on 21 April, which Room 40 did not break for three days.53 There were scraps of signals intelligence, notably orders to zeppelins to scout the North Sea, but not enough to convince the British a major sortie was under way. Five British submarines were on patrol in the Bight, disposed to detect and report such a sortie. One of the submarines—J6—operating off Horns Reef, sighted the German force on the afternoon of 23 April. Unfortunately, her captain decided—admittedly in poor visibility—that the ships he saw were forces covering a British minelaying operation of which he had been warned and neither reported his sighting nor made an attack. Perhaps it was an improvement in British procedures that J6’s captain had been informed of friendly forces, but it is clear that the information was either insufficiently precise or not well understood.
What the Germans did not understand was that their own intelligence was poor. They had timed their attack for a gap in the schedule. The irony was the greater because the covering force that returned to Rosyth on 24 April was the most vulnerable of any element of the Grand Fleet: the Second Battle Cruiser Squadron, comprising the four oldest and weakest of the type. The scouting groups also had to do their own reconnaissance. The zeppelins L41, L42, and L63 launched late on the 23rd, but were recalled in the early hours of the following morning because of rising winds.54 This forced Hipper to proceed near 60° N to search for the convoy, farther than any capital ships of the High Sea Fleet had been during the entire war.
The scouting groups’ progress was not without difficulty. Seydlitz shut down her starboard engine for two hours due to a condenser defect. The poor quality of the battle cruisers’ coal also caused problems, forcing frequent furnace cleaning and limiting speed—Von der Tann could barely manage twenty-one knots.55 Worse followed. At 0410 Moltke’s inner starboard shaft snapped due to the failure of a thrust block. The outboard end of the shaft and the propeller broke off and despite efforts to use the overspeed shutdown, the starboard high-pressure turbine ran away, shattering the turbine, and scattering debris that penetrated pipes and machinery. The midships engine room and the dynamo room rapidly flooded as the Moltke took in sixteen hundred tons of water. Since the auxiliary condenser pipes had been breached, the feed system rapidly salted. The ship could still manage thirteen knots, but it was only a matter of time before she lost motive power.56 Hipper was unaware of the full extent of Moltke’s problems when he detached her to join Scheer and was alarmed to receive the crippled ship’s 0543 signal that she was restricted to four knots. Sent at high power, this was a major breach of radio silence, but justifiable in circumstances that saw the ship at risk of being stopped and helpless in the open sea. The signal took time to get through to Scheer and Moltke’s initial text confused the situation with an incorrect position. While this was being sorted out, Scheer detached Strassburg to assist Moltke. Hipper initially turned back to support his lame duck but resumed the search for the convoy when he became aware the main body was dealing with the cripple.
Moltke’s 0543 signal was picked up by the British, their first clear indication something was amiss. At this stage, the text was not decrypted, but Room 40 was sure of the call sign. Unfortunately, the directionals placed Moltke inland in Norway, an ironic assessment given the positional error in the original signal.57 Although the Admiralty immediately informed Beatty of the intercept, uncertainty as to its credibility delayed an order to sail by nearly two hours. However, by this point Room 40 was in little doubt that something was up. This was confirmed by decryption of a Neumünster signal reporting its own analysis that the British were not aware of German forces at sea. The inference was clear. This information was sent to Beatty at 0955 and he first directed the battle cruisers and then the remainder of the fleet to prepare to sail. The Admiralty ordered the Grand Fleet out at 1047. Despite the Firth of Forth being enveloped in thick fog, a frequent condition and one of the long-standing objections to Rosyth as a fleet base, the entire fleet departed without incident in less than three hours.
Even without the problem of Moltke, the High Sea Fleet was not going to linger in the north. The force still included torpedo boats with very limited operational endurance and these could not remain out another day without eating into their operational reserves, vital to retain in case the enemy was encountered. The clear weather eventually left no room for doubt the intelligence was mistaken, even if Neumünster still claimed a convoy was due to depart Bergen on 24 April. After sighting the Norwegian coast, the First Scouting Group turned for home just after 1500. Strassburg and then the battleship Oldenburg had taken Moltke under tow, while her engineers worked to flush the boiler tubes with newly distilled fresh water. When von Hipper caught up, his ships took station around Moltke. By 1600, the battle cruiser’s portside shafts were turning and there were prospects of bringing the outer starboard shaft into operation using the low-pressure turbine. The tow was maintained, but von Hipper and Scheer could be reasonably confident they had a clear run to the Bight.
The geometry of the situation was probably obvious to Beatty as he sailed. The Grand Fleet could not get across the North Sea in time to intercept the Germans once the latter turned south. That they had done so was made clear by a signal from the Admiralty at 1625, which gave the position of the German flagship steering a course SSE. Perhaps to Room 40’s amusement, it decrypted a signal from U19 that reported “11 old enemy cruisers” in a position that matched that of the High Sea Fleet. J6 was not the only muddled submarine on patrol, although the U-boats had the benefit of warning by “catchword” signals that their fleet would be at sea—a clue Room 40 realized it had missed.58 The Grand Fleet continued its sortie into 25 April, but the High Sea Fleet was well ahead of it, helped by practically perfect weather conditions that allowed Moltke to maintain speed through the combination of tow and engines. Only at 1640, by which time Moltke could manage thirteen knots, was the tow cast off. This was timely as the fleet was well into the Heligoland Bight, trammelled by mine fields and the submarines the Germans knew to be on patrol. The minesweeper M67 had already hit a mine and sunk, while there had been at least one submarine sighting.
J6 did no better a job on the High Sea Fleet’s return. Between 0400 and 0715 Lieutenant Commander Geoffrey Warburton observed substantial components of the German fleet on their homeward course. He did not attack but surfaced and reported the sightings. However, orders that gave priority to reporting were intended for outward-bound contacts; J6 should have attacked the homeward-bound German fleet. The British had a last chance with E42, dispatched to lie across the swept channel that the Admiralty thought most likely the High Sea Fleet would employ. Forced to submerge and bombed by a seaplane, Lieutenant C. H. Allen was fortunate to spot smoke in the northeast in the late afternoon of 25 April. A dash southeast brought him in contact with Moltke’s formation. Allen fired his two bow and two beam tubes in succession at a range of over two thousand yards and was rewarded by a single explosion.59 Moltke spotted a torpedo on her port bow and turned to avoid, but it struck abreast her port engine room at 1838. More than seventeen hundred tons of water flooded in, forcing immediate counter-flooding and threatening Moltke’s ability to move. The battle cruiser responded with gunfire in the torpedo’s direction and called for help. Over the next few hours, torpedo boats and destroyers supplemented the battle cruiser’s escort as she struggled to get home. E42 was hunted, and an estimated twenty-five or more depth charges were deployed against her, but she escaped unscathed. In addition to strenuous patrols, smoke screens were deployed around the battle cruiser. These were so effective they embarrassed the late-arriving Second Flotilla, which was lucky to avoid any collisions. Good damage control and further hard work by her engineers meant that Moltke never lost steaming power, although she was once again under tow that evening. Moltke was safe in Wilhelmshaven anchorage by the morning of 26 April, but her damage meant she would not be fully operational again until mid-September.60
The Grand Fleet returned to harbor after the Admiralty assessed that the High Sea Fleet was on its way home and could not be caught. Beatty had already decided to allow the convoy cycle to resume without delay, although he strengthened the covering force. The incident confirmed the C-in-C’s misgivings over the requirement to protect the convoys with a detachment of heavy ships. What is clear is that his strictures about the potential weakness of the Grand Fleet were not justified. There were thirty-one capital ships in the force that left Rosyth, despite the absence of the Second Battle Cruiser Squadron and at least three other dreadnoughts. Room 40 carried out its own postmortem. The signals intelligence system could certainly have done better but had not done that badly. Confirmation Room 40 could not be relied upon to provide warning on every occasion heightened the concern of senior officers, but perhaps the greatest failure lay in J6’s captain’s poor assessments.
One theory sometimes raised by historians is that a previous convoy cycle had been protected by the Sixth Battle Squadron.61 It is true that the destruction of the American force, when thus isolated from the Grand Fleet, would have been both a material and a psychological disaster for the Anglo-American relationship. Von Hipper’s battle cruisers, however, would have found Rodman’s quartet of powerful dreadnoughts a hard nut to crack, even if the German problems with coal and machinery did not eliminate their speed advantage, while Rodman would have had to make a succession of unlikely errors to become entangled with the main body of the High Sea Fleet. He and all the Grand Fleet’s admirals were acutely aware of the risk of defeat in detail and Rodman drew Sims’s ire when he expressed his concerns a little too openly later in April.62 Scheer had already taken his fleet farther north than it had been since July 1914. He would have no appetite for a chase toward the Scottish coast, a chase that in any case the light forces of his fleet could not have long continued.
