There was apparently a certain tension on the German side between the requirements of Scheer for destroyers to support operations in the Bight and the demands of Flanders for destroyers to support the guerrilla war against the cross-Channel barrage. The dual control of the flotillas also produced problems. The Third Flotilla arrived at Zeebrugge on 23 March, and the Sixth Flotilla returned to the Helgoland Bight at the end of the month. Scheer considered the Sixth Flotilla as necessary to support the High Sea Fleet’s operations in the Helgoland Bight, which at that moment consisted largely of supporting the operations of the Sperrbrechers (“mine bumpers” or “barrier breakers”) and minesweepers in keeping the Wege open for the U-boats passing in and out of the North Sea bases. The Third Flotilla was now stationed permanently in Flanders with its commander, Fregattenkapitän Kahle, serving as F.d.T. (Fuhrer der Torpedoboote) Flandern, and Korvettenkapitän Albrecht’s Flanders destroyers were raised to full flotilla status and Korvettenkapitän Assmann’s Flanders Torpedo Boat Flotilla was similarly reinforced. Assmann’s smaller torpedo boats, although less prominent in the raids, had been able to claim success in largely clearing the old Belgian coast barrage by the beginning of 1917.44
The German forces in Flanders now numbered 24 to 25 destroyers supported by 14 to 15 torpedo boats. They were in a position to wage a troublesome, but not likely to be decisive, guerrilla war against the British in the Straits of Dover. The Germans launched a raid on the night of 20–21 April and came up short. They attempted a novel form of direction to solve the problem of command and control on a dark night. Kahle, the F.d.T. Flandern, directed the operation from German headquarters at Bruges where he could gather German and whatever British wireless intercepts that could be deciphered, collate them, and issue the appropriate orders. The 12 German destroyers, launched 1915–16, were new, 34-knot, 824- to 960-ton craft armed with 3.4- or 4.1-inch guns. They were divided into two distinct groups, one to operate on the Dover and the other the Calais side. The German orders called for an attack on the barrage patrol and, failing this, a vigorous bombardment of Dover and Calais. Room 40 and other intelligence sources apparently failed to provide any advance warning of the German move, and that night there was only the normal night patrol of the destroyer leaders Broke (1,794 tons) and Swift (2,207 tons) on the Dover side and four destroyers on the Calais side. The Germans bombarded first Calais and then Dover. The Swift and Broke engaged the Germans, who had bombarded the latter. In the ensuing mélée the Swift torpedoed G.85 and the Broke rammed G.42 and, with the two ships temporarily locked, fought to repel German boarders in a hand-to-hand action reminiscent of an earlier age. The two British leaders were badly damaged, the Broke had to be towed back to port, but the two German destroyers were sunk. The Broke’s captain, Commander E. R. G. R. Evans, caught the imagination of the British public and was henceforth known as “Evans of the Broke.” The run of German success in the Strait was ended and they refrained from raiding the cross-Channel barrage for ten months.45
The German destroyer raids in the Straits of Dover had a dual objective: to damage the barrage and facilitate the passage of U-boats, and to disturb communications between England and France. The raids were hazardous and the Germans depended on surprise, darkness, and speed to escape potentially overwhelming forces. The Germans failed to seriously disturb communications, and in waging a war of attrition against the patrols guarding the cross-Channel barrage they were really attacking, and the British defending, an obstruction that at the time had little real value. The barrage had been extended to the Snouw Bank off Dunkirk in February, but the great tidal pressure restricted the nets to only 40 to 60 feet in depth. There were also three lines of moored mines a half-mile to the west of the nets, with secret gaps for patrols, which were changed periodically. Unfortunately the mine nets (with two electro-contact mines in each net) proved to be difficult to maintain in Channel conditions, whereas the moored mines dragged their moorings and fouled the nets. They became so hazardous for the patrols that from May to June the entire western portion of the barrage had to be lifted, swept, and relaid with stronger moorings southwest of their original position. The U-boats may have been bothered less than the patrols. They were normally able to cross the net barrage during the hours of darkness by drifting over the nets at high tide. If they were unlucky enough to foul a net, they were usually able to break through without exploding a mine. It is estimated that from January to November of 1917 only one U-boat out of the six sunk in the Dover Strait area was actually sunk by a mine during the 253 passages through the barrage.46
In the latter part of April it was obvious the British would have to find some method of stopping their huge shipping losses. There was always the possibility of another “offensive” action—a direct assault on the Flanders submarine bases themselves. The Dover Patrol had conducted periodic bombardments with big-gun monitors of the German-held coast since the beginning of the war. The Germans had expended considerable effort in fortifying this area, making it one of the most heavily defended zones in the world, and the coastal bombardments had become dangerous for the navy. Bacon’s bombardments of Zeebrugge in the spring of 1917 were an attempt to destroy or damage the lock gates of the canal to Bruges, thus bottling up the U-boats. The Zeebrugge bombardment was followed by a bombardment of one of the dockyards and workshops at Ostend. The attempts were not successful.47
Jellicoe’s pessimism over the submarine war helped make the capture of the Flanders submarine bases one of the most important initial objectives of the British army’s offensive in the summer of 1917. Jellicoe was convinced that “a naval bombardment alone will never turn the Hun out of Zeebrugge and Ostend.” Even if the bombardment succeeded in destroying the workshops at both ports, the Germans would only have been driven up to Bruges where the navy could not get at them.48 Jellicoe’s insistence on Haig’s offensive, which ended in the horrendous losses and futility of the Third Ypres (or Passchendaele) battles, undoubtedly earned him the ill will of the prime minister. Lloyd George was much more attracted to the idea of the bombardment than to a costly land offensive. Admiral Bacon also had developed a scheme for a landing near Westende on the Belgian coast by means of huge pontoons pushed by monitors. The objective was to turn the German flank from the sea, but the plan was stillborn when the army failed to advance far enough.49 Perhaps this was fortunate. The slow-moving monitors and huge pontoons would have made easy targets, and the Germans were confident they could defeat a landing from the sea.50 However, these problematical operations were in the future and would do nothing immediately to stop the hemorrhaging from submarine losses at sea. Furthermore, even if the Allies were successful, the Flanders U-boats accounted for only a portion of the Allied sinkings, and their bases could have been shifted back to the German coast.51
Zeebrugge and Ostend remained attractive targets, but the dramatic attack on them did not take place until the following year (see chapter 13). By then the Admiralty had stumbled on the most effective counter to the submarine, the general convoy system. It was not a new idea, convoy was actually a very old and traditional idea, but the Admiralty for a variety of reasons resisted its widespread introduction for a long time.52
THE CONVOY SYSTEM
The Admiralty were finally driven in desperation to the introduction of convoys at the end of April. There were actually a series of preliminary experimental and small convoy systems instituted before the oceanic convoys that eventually frustrated the submarine campaign. These were the Dutch, the French coal trade, and the Scandinavian convoys. The little-known Dutch convoys on the Hook of Holland route were the first. They were more in response to the threat from German surface flotillas than the submarine danger. The Dutch route was vulnerable to sudden dashes from either the Bight or Zeebrugge on its flank, such as the one resulting in the capture of the Brussels and Captain Fryatt in June 1916 (see chapter 10). The first Dutch convoy sailed on 26 July 1916, and for the remainder of the war British vessels going to and from Holland proceeded si
multaneously in convoys of four to nine ships at intervals of two to three days. Tyrwhitt’s Harwich Force was responsible for the convoy’s protection and referred to them as the “Beef Trip” after an agreement with the Dutch that supposedly diverted Dutch foodstuffs from Germany to England. They became, according to Tyrwhitt’s biographer, “a principal—and often blasphemously execrated—activity of the Harwich Force.”53 The details of protection changed, but the convoys were extremely successful, despite the fact the route had German submarine and destroyer bases on both flanks. The British lost only one escorted ship—a straggler, which was captured—in the interval between the introduction of the convoys and unrestricted submarine warfare.54 Afterwards there were only 6 losses in convoy out of a total of 1,861 sailings, or a loss of .32 percent. Moreover, the losses were all prior to the introduction of a close formation in June 1917, in place of the long straggling columns that had been the rule up to then.55
A skeptic might argue that the Beef Trips represented too small a portion of the volume of British and Allied trade and the conditions too restricted for these convoys to be a true test of the system. The French coal trade convoys represented a truer precedent for the general system of convoy. The French coal trade convoys were a response to the heavy losses inflicted on the colliers carrying British coal to France during the last quarter of 1916. Roughly half the colliers were neutral ships, largely Norwegian, which could not be detained if they wished to sail and had not received the special warnings or routings provided to British and Allied vessels. The French were heavily dependent on the British for coal, and the detention of British and Allied colliers when submarines threatened the route was tantamount to a successful German blockade. A special representative sent by the chief of the French naval staff to London pointed out on 2 January that these detentions represented an effective blockade for 30 to 40 percent of the last two months of 1916. The French presented a detailed plan for convoys based on an apparently meticulous study by the French naval staff of statistical data pertaining to the coal trade. The Admiralty balked at the term convoy, which they considered might embarrass neutrals, and proposed substituting the term controlled sailing. They apparently feared the word convoy might represent protection and that the Germans might respond by sinking all neutral shipping on sight, a threat that might paralyze sailings to Scandinavia. Nevertheless, the First Sea Lord approved the arrangements on 22 January, and the first convoy sailed from Mount’s Bay to Brest on 10 February. There were three routes: Mount’s Bay to Brest, Weymouth to Cherbourg, and Weymouth to Le Havre. There was also a fourth very short route from Southend to Boulogne and Calais. The crossings were made every twenty-four hours, and the slow colliers proceeded either in groups steaming in rough formations without direct escort but following special routes assigned by the Admiralty or under direct escort by armed trawlers of the auxiliary patrol. The escorts were weak; in the first quarter of 1917 fewer than 30 trawlers protected more than 4,000 crossings. The results were startlingly successful: from March to May 1917 only 9 of the 4,016 ships convoyed were lost. The entire loss in convoys on the four French coal trade routes during the war was only 53 out of 39,352 sailings, or .13 percent.56
The heavy losses suffered in 1916 by Norwegian shipping trading on behalf of the Allies on the Scandinavian routes led the Norges Rederforbund—the Norwegian shipowners association—to ask the Admiralty in September for some form of relief. They had been prompted by their insurers and were supported by the British Board of Trade. The Admiralty balked at the idea of convoys or any special forms of protection, and by mid-November Norwegian crews were threatening to refuse to sail without more effective protection. There was seemingly endless discussion between the Admiralty, the Foreign Office, and the Board of Trade, and the Admiralty answered criticism by pointing out that routing and patrol was complicated by the dictates of Norwegian neutrality and British men-of-war could not enter Norwegian ports. Admiral Oliver, the chief of staff, peevishly remarked: “It had never been contemplated that we should have a big enough Navy to protect the whole of the world’s commerce.” After apparently interminable discussion, and under the threat of the resumption of unrestricted submarine warfare, the British finally adopted a scheme whereby vessels on the Bergen-Lerwick (Shetland Islands) route covered as much as possible of the journey during the hours of darkness and armed trawlers escorted them on the so-called daylight stretch from a rendezvous approximately 50 miles from Lerwick.
The absolute and ruthless nature of the German declaration of blockade may have helped to resolve the problem, for it removed from the Norwegians the last vestiges of protection their neutrality might have given them, and they gave up their scruples against accepting British orders in regard to routes. The British government also agreed to take over the reinsurance of all Norwegian steamers carrying contraband for the Allies in the war zone. The first British steamer to be escorted left Lerwick on 29 January, and the first Norwegian left on 10 February. The first convoy to be escorted through the 50-mile “daylight stretch” consisted of one British and eight Norwegian steamers and sailed on 24 February, but convoys remained a temporary measure in February and March when vessels were escorted singly or in pairs.57
Norwegian shipping in the North Sea was hard hit by unrestricted submarine warfare in March when twenty-seven Norwegian ships were lost. The Norwegians lost confidence in the Shetlands-Norway route; they complained they never saw a British warship except near the naval ports. Moreover, German submarines were active off the east coast of Scotland, and the Norwegians considered the stretch from Lerwick to Peterhead as even more dangerous. The British also had intercepted German wireless messages from Bruges to U-boats reporting the departure of Norwegian shipping from Bergen. Once again driven by the crisis, Vice Admiral Sir Frederick Brock, Admiral Orkneys and Shetlands, called a conference at Longhope in Scapa Flow. The first meeting on 30 March considered local measures for protection of traffic between the Shetlands and Norway. Certain specific rendezvous were established and would be changed at intervals of three to ten days. Eastbound traffic from Lerwick would be escorted to the rendezvous, then dispersed so as to arrive on the Norwegian coast at daylight. Westbound traffic would leave the Norwegian coast daily so as to arrive at the rendezvous at 6:00 A.M. where they would be met by an escort to bring them into Lerwick. On 4 April a second conference at Longhope unanimously recommended a system of convoy from Lerwick to the south along the east coast of Scotland. Beatty, now commander in chief of the Grand Fleet, approved with the observation that as it was impossible to provide an escort for each individual vessel, the only alternative would be a system of convoys.58
Something had to be done, for the Bergen-Lerwick route was suffering heavy losses. Between the 10th and 15th of April, U.30 sank nine ships between the Norwegian coast and the rendezvous, or in the vicinity of the rendezvous for the “daylight stretch” escort into Lerwick. The Admiralty approved the proposals of the Longhope conference on 24 April, although Oliver still doubted they would be successful, as in his opinion convoys invited torpedo attack and the available escorts were too few. The convoys sailed south from Lerwick and north from Immingham on three days out of four. The first southbound convoy left Lerwick on 29 April.59 By the time the Scandinavian and east-coast convoys went into effect, the issue of local convoys, which the Admiralty regarded as experimental, had merged into the larger question of a general system of oceanic convoys to protect the great majority of trade coming into the British Isles from the rest of the world. At this point it would help to examine exactly why it was so difficult to institute a system of defense of trade, which had worked so well in preceding centuries.
The First Sea Lord in November 1916, Admiral Sir Henry Jackson, and the chief of the war staff, Admiral Oliver, had rejected the idea of convoy when it was mentioned in the War Committee on the grounds that convoys had only been successful when it had been possible to allocate a separate escort for each vessel. Convoys also presented too big a target and it wo
uld not be possible to keep merchant ships together. In the opinion of Jackson and Oliver, the fitting of merchant ships with defensive armament offered the most effective means of protection. Oliver also was inclined to believe, as he has already been quoted in regard to Norwegian shipping, that however necessary and effective escorts might have been for troopships and special cargoes or under certain circumstances, it was simply impracticable to attempt to use them for the maritime trade of the entire world. Many naval officers were loath to employ warships merely to convoy trade when, in their opinion, their proper function was offensive, that is, attacking and combating enemy warships. The popularity of so-called hunting flotillas was the result.60
One would have thought that the experience of the war might have changed this view, but at the beginning of 1917 it had not. In January 1917 the War Staff issued a paper entitled “Remarks on Submarine Warfare,” which concluded that the system of sailing ships in company as a convoy was not recommended in any area where submarine attack was a possibility. The larger the number of ships forming a convoy, the greater the chance that a submarine would be able to deliver a successful attack and the harder the task of an escort in preventing that attack. It was also preferable for a defensively armed merchant ship to sail singly rather than be placed in a convoy with other ships. A submarine also might be able to remain at a distance and fire a torpedo into the middle of a convoy with a good chance of success. The War Staff at the end of January delineated specific reasons for rejecting the idea of convoy: it would be impracticable to supply the numbers of escorts convoys would require; it would be difficult for convoys to meet at a rendezvous; there was danger of attack while convoys were assembling; numbers in a convoy would have to be limited or they would have to sacrifice the protection of zigzagging; it was difficult to form convoys of equal speed; it would be disadvantageous to sail faster ships with slower ones; congestion at ports would result when numbers of ships arrived at the same time, which would result in a decrease of available tonnage; there would be increased risk from mines; and masters would not be able to maintain station in convoy. In the words of the study made by Lieutenant Commander Waters of the Naval Historical Section after the Second World War, “Not one of these objections was valid.”61
A Naval History of World War I Page 61