"Each and every convoy now involves the Naval Staff in intricate operations, all carefully planned and brilliantly executed".
W. S. Churchill. Extract from statement at a Conference of Ministers, 20th April, 1943.
At the beginning of the year the three great operational bases of the Western Approaches Command (Liverpool, Greenock and Londonderry) controlled twenty-five groups of escort vessels totaling some seventy destroyers, eighteen sloops, sixty-seven corvettes and ten ex-American coast guard cutters.1 In addition to these ships American destroyers were still escorting certain North Atlantic convoys, and were using Londonderry as their base. The British groups were divided into four categories. Firstly, the 'Special Escort Groups', composed of short-endurance destroyers, looked after WS and PQ convoys2 during the first part of their passages, met the 'monster' liners when they started to bring American troops across the Atlantic, and undertook any unusual requirements which might arise. Secondly, there were the groups of long-range destroyers and corvettes, which provided the ocean escorts for the North Atlantic convoys from the Western Ocean Meeting Point right home to Britain, and also escorted the Gibraltar convoys. There were now sufficient long-range ships to avoid, save in exceptional circumstances, the complication of having to send escorts to Iceland to refuel while on passage, as had been necessary in 1941.3 Indeed, the importance of Iceland as a refueling base for the Atlantic escorts had declined just when its importance as an air base came to be fully exploited. These first two categories of escort groups were based on Liverpool, Greenock and Londonderry. The third class of escort
group was composed of long-range sloops, destroyers and ex-American cutters. There were six of these groups, and they usually worked from Londonderry to escort the SL/OS convoys to and from the rendezvous with the Freetown local escorts. Lastly, there were anti-aircraft groups, which consisted mainly of auxiliary A.A. ships and worked in the Irish Sea, besides escorting the Arctic and Gibraltar convoys. The Greenock, Londonderry and Liverpool groups were theoretically interchangeable, but every endeavour was made to keep each individual group intact and to employ them on one route, thereby gaining the advantage of familiarity not only with each other but with the problems peculiar to that route.
At Gibraltar there was a force amounting to about two groups to provide local escorts. The South Atlantic Command possessed one destroyer flotilla, five sloops and some two dozen corvettes in the Freetown Escort Force, while in the Western Atlantic there were fourteen destroyers and about forty corvettes of the Royal Canadian Navy (or lent to that service from the Royal Navy) forming the Newfoundland Escort Force, plus about a score of escort vessels for western local escort purposes.
While transfers of complete groups from one command or station to another occurred fairly frequently, the same general organisation, based on the broad principles described above, continued in the Western Approaches Command throughout the period covered by this volume. Appendix G shows the strength available on different dates, and how changes in allocations were made to meet the varying needs of the war.
Not long after this phase started - to be precise in March - the British and American authorities reviewed their needs of escort vessels. The table below shows the results.
Table 6. British and American Escort Vessel Requirements, March 1942
British American Required 725 590 Available 383 122 Shortage 342 468
It was agreed that new ships should be allocated in proportion to each nation's deficiency. There were now 300 escort vessels building on British account in the United States (known as British Destroyer Escorts or B.D.Es, and later called Frigates by the Royal Navy). Delivery of 200 was expected by the end of 1943; but they did not actually enter service fast enough to meet the ever-rising calls for escorts, and the severe shortage from which we had suffered since the beginning of the war continued. It remained indeed a permanent feature of the Atlantic struggle almost to the end. The main factor
'Convoy Air Cover', by Norman Wilkenson. (National Maritime Museum)
German Type IXB (1.050 ton) Atlantic U-Boat
U 71 under attack by Sunderland U. of No. 10 Squadron R.A.A.F. on 5th June 1942. The U-Boat dived when sighted, and was forced to the surface by depth charges. The Sunderland then engaged with machine guns, as can be seen in the photograph.
which delayed completion of the new escort vessels was the overriding priority for labour, steel and engines given in America to the landing craft needed for the cross-Channel operation, which they still hoped to launch in 1943 . It was at British insistence that this was finally postponed, and American history has since accepted that our judgment was sound4; but the priority given to landing craft certainly contributed greatly to the continuing weakness of our Atlantic escorts, and to the heavy shipping losses we suffered at this time. In July it was agreed with the Americans that all 'Destroyer Escorts', as well as the new twin screw corvettes and minesweepers, whether building in America, Canada or Britain, should be 'thrown into a common pool' and 'assigned between us according to our strategic requirements'. But difficulties soon arose in deciding which 'strategic requirement' should have the highest priority.
During the winter of 1941-42 our convoys were, in general, using the northerly Atlantic route in order to gain the greatest possible protection from the air bases in Iceland. Furthermore, by keeping close to the 'great circle' track between Newfoundland and the North-West Approaches our escorts were able to conserve their fuel as much as possible. But the northerly route was extremely trying to the merchantmen and their escorts in mid-winter, and the Admiralty much desired to move the convoys further south. Not until nearly the end of the phase now to be described did this general shift of the shipping routes become practicable; and even then it had the disadvantage that it reduced the effectiveness of the cover afforded by aircraft working from Iceland. Throughout the whole of the winter and spring our convoys had to contend with the very severe conditions of the high latitudes.
Long before the entry of the United States into the war the German Naval Staff had realised that our shipping was far more vulnerable in the Western Atlantic than in the more easterly parts of that ocean. Although the HX and SC convoys, and also their outward counterparts, now had continuous anti-submarine escorts throughout their journeys, it was inevitable that, after reaching the Canadian coast, many ships from west-bound convoys had to be routed onwards independently. These, and the great flow of American shipping off the east coast of the United States, which was still sailing in peacetime fashion, offered inviting targets to the U-boats. Well before Pearl Harbor the Germany Navy had been watching the increasing American participation in the Atlantic struggle with concern, and had repeatedly, though vainly, asked to be allowed to retaliate. Although U-boats had, in the previous September, been sent to work off southern Greenland and in the Straits of Belle Isle,
and had there been involved in incidents with American warships5, Hitler had then refused to allow them greater freedom. To Admiral Dönitz and his staff the entry of the United States into the war brought not only the welcome lifting of such restrictions, but an opportunity again to find virtually undefended targets, such as they had not enjoyed since 'the happy time' of July-October 1940.6 They knew, moreover, that they had somehow to improve greatly on the results achieved in 1941 if they were to succeed in bringing Britain to her knees. Their estimate was that if sinkings averaged 700,000 tons per month we should soon be defeated. Yet in 1941, although they believed the figure to be much higher, they had actually only managed to sink an average of 180,000 tons every month. America's entry into the war therefore brought new hope as well as new opportunities to the U-boat command, especially as it occurred at a time when its strength was at last beginning to increase rapidly. The decision to send U-boats to the American seaboard was taken on the 12th of December - the day after Germany declared war on the United States - and the movement was given the somewhat histrionic name of 'Paukenschlag' (Roll on the drums). But because Japan's intentions had not been known in Berlin, the sudden arrival of an opportunity which they had long desired took the Germans by surprise, and they were not able at once to make the most of it. They had suffered heavy losses in the Mediterranean and off Portugal in the last month of 1941, and still had a number of boats caught in what Dönitz had accurately described as the 'Mediterranean trap'.7 Moreover, the German Naval Staff still clung to their determination to help avert a collapse in Africa by keeping a number of U-boats in the Mediterranean. The consequence was that when Dönitz proposed to send twelve of the large (1,100 ton) U-boats to American waters the number was reduced by his superiors to half that figure. Those six boats sailed on the 2nd of January. Three were finally ordered into the Mediterranean to replace losses, while the other three were kept on patrol between Gibraltar and the Azores.
A month later the enemy realised that the Western Atlantic offered far better prospects than the Gibraltar area, where his recent experiences had proved unfortunate. In mid January Convoy HG 78 had been attacked by three U-boats, and one of them (U-93) was sunk by the escorting destroyer Hesperus; and when the troopship Llangibby Castle was torpedoed on the 16th and put into Horta in the Azores to effect temporary repairs, U-581 was sunk by the destroyer Westcott, one of the three escorts sent to bring the damaged ship home. Between January and March about five U-boats worked between the
Gulf of St. Lawrence and Cape Hatteras, and three more soon reached the waters off the Chesapeake. They immediately achieved such substantial success (forty-four ships were sunk in the Canadian Coastal Area in January and February) that the enemy decided to send every available boat to follow them. The 'Roll on the drums', whose start had been somewhat muffled, was now reverberating menacingly around the western ocean.
By mid January seven of the medium-sized (750 ton) U-boats had arrived off Newfoundland. There they found conditions much less to their liking. Most of the shipping was strongly escorted, air cover was better and the weather was intensely cold. In February they were therefore moved to the south of Halifax, where it was again found to be both cold and dangerous. In March they started to cruise off New York. Dönitz was pleasantly surprised to find that the endurance of the smaller boats permitted them to stay two or three weeks on patrol in these distant waters.
From the foregoing account of the enemy's policy it will be seen that the Americans were granted about five weeks' grace (7th December 1941 - mid January 1942) before attacks started in earnest. In assessing the reasons for their slowness in starting convoy and getting their anti-submarine defences in order, it is important to remember that many of their escorts were employed in the Pacific until after the Battle of Midway had removed the Japanese threat to the life-line between the west coast of America and Australia. Yet it seems undeniable that even the simplest arrangements for the better control of coastal shipping were very slow in being adopted. As an example of the favourable conditions encountered by the U-boats, we may quote from the war diary of Lieutenant Commander Hardegen (U-123), who spent a very happy night on the 18th - 19th of January off Cape Hatteras. In his diary he wrote: 'It is a pity that there were not ten to twenty U-boats here last night, instead of one. I am sure all would have found ample targets. Altogether I saw about twenty steamships, some undarkened; also a few tramp steamers, all hugging the coast. Buoys and beacons in the area had dimmed lights which, however, were visible up to two or three miles'. It appears that the actual result of that night's work by U-123 was three ships of about 17,000 tons sunk and one damaged.
Not until the 1st of April was even a partial convoy system started; and in the meanwhile innumerable factors such as the lighted channel marks, the complete lack of a coastal 'blackout' and the unrestricted use of ships' wireless, gave the U-boats all the help they needed. Thus began what the American historian has ironically described as 'a merry massacre'8, and the enemy called his second
'happy time'. The U-boats generally lay submerged not far offshore by day, and moved inshore to attack with guns or torpedoes on the surface by night. Few north Atlantic convoys were sighted, let alone attacked at this time. The enemy had, in truth, no need to invite retaliation from their escorts; for he could find all he needed with much less risk to himself elsewhere. An exception did, however, occur at the end of February when convoy ON 67 was sighted by a U-boat 600 miles north-east of Cape Race. Five U-boats (four of which were outward-bound from their Biscay bases) were called to the scene and in a three-day battle sank eight ships, six of which were large tankers, without loss to themselves.
One of the most surprising facts regarding the havoc wrought off the American coast in the early days of 1942 is that there were never more than about twelve U-boats working in those waters at any one time-no greater strength than the enemy had sometimes mustered to make a 'pack attack' on a single one of our convoys in 1941. Yet in the first two weeks of the new campaign they sank 13 ships of nearly 100,000 tons, and in February the sinkings in the American Eastern Sea Frontier Command alone exceeded that tonnage.9 In the following month twenty-eight ships of 159,340 tons were sunk in the Eastern Sea Frontier and fifteen more of 92,321 tons in the Gulf and Caribbean commands; and no less than 57 per cent of the tonnage sunk was tankers. Not until April were defensive measures started by the Americans in earnest.
From the 12th of December 1941 until the following 17th of January the First Sea Lord himself was in America to review the whole maritime war with our new Allies, and to coordinate with them the many fields in which the naval forces of the two nations would be working together. Shortly before Admiral Pound's return to London discussions between the Admiralty and the U.S. Naval Mission were started on this side of the Atlantic. The first U-boat attack off the American coast had just taken place. Arising out of these discussions the American mission in London asked the U.S. Navy Department the following questions:
- What help could be given by Britain to provide escorts from the American Atlantic ports to Halifax or Bermuda, or to the Western Ocean Meeting Point? The reply was that 'no effective assistance' could be given.
- What was the Navy Department's estimate of the effectiveness of their anti-submarine surface and air defences on the Atlantic seaboard? The reply was that they were 'inadequate'.
- Whether it would be possible to keep a coastal lane reasonably free from U-boat attack? The reply was negative.
In spite of these discouraging replies the British Government had not forgotten what Admiral Pound called 'the two great pre-war gestures' made by the Americans, namely the provision of fifty destroyers when we were at our lowest ebb for escorts, and American participation in Atlantic convoys.10 Admiral Pound had, while in the United States, offered to turn over ten of our corvettes to them. Early in February the British mission in Washington suggested offering two dozen of our anti-submarine trawlers, and the Admiralty promptly put the proposal into effect. By the end of March most of them had arrived on their new station, and Professor Morison has recorded that 'these rugged little coal burners . . . were a great help'.11 Loan of some of our bombers was also discussed, but the Air Ministry found itself unable to spare any. However, in June, with Admiralty agreement, No. 53 Squadron of Coastal Command went to Rhode Island and then on to Trinidad, off which sinkings had become serious. The First Sea Lord hoped that we had thus `responded with equal generosity when the American need for reinforcement became urgent'.
None the less to the British authorities it seemed that the Americans were both slow and unwilling to start coastal convoys. On the 19th of March the First Sea Lord told Admiral King that he 'regarded the introduction of convoy as a matter of urgency', and that convoys with weak escorts were preferable to no convoys. Mr Churchill telegraphed his 'deep concern'12, but the President suggested palliatives such as reducing the British import programme; while Admiral King considered that 'inadequately escorted convoys were worse than none' - the exact opposite to all that our experience had taught.
The release of escorts from the mid-ocean groups was constantly discussed, and on the 16th of April the Halifax convoy cycle was opened out from six to seven days so that two escort groups could be sent to the east coast of America. Mr Hopkins was at that time in London, and on the 14th of April he telegraphed to President Roosevelt that in the preceding three months Allied losses had totalled 1,200,000 tons, over half of which had been tankers. The First Lord had, in fact, just told the Prime Minister that for the preceding week our tanker losses 'had been frightful'. 'We are', said Hopkins prophetically, 'going to need all these ships desperately in the next few months . . . I doubt very much that anything short of convoy is going to do this job . . . They [i.e. the British] whose island is so dependent on imports realize full well the significance of
these sinkings to the future of the war'. The Admiralty's Director of Anti-Submarine Warfare (Captain G. E. Creasy) and the Air Officer Commanding No. 19 Group of Coastal Command (Air Vice-Marshal G. R. Bromet) had meanwhile gone to the United States to advise on the formation, training and organisation of air and surface antisubmarine forces.
In March, when Admiral Stark was appointed to command the American naval forces in Europe, the First Sea Lord had a review of the whole Atlantic battle prepared for him. Our anti-submarine experiences 'over the past thirty months' were there summarised under the following four headings:
- The comparative failure of hunting forces.
- The great value of aircraft in convoy protection.
- The supreme importance of adequate training and practices.
- The value of efficient radar.
As to the first the Admiralty said that 'this is one of the hardest of all the lessons of the war to swallow. To go to sea to hunt down and destroy the enemy makes a strong appeal to every naval officer. It gives a sense . . . of the offensive that is lacking in the more humdrum business of convoy protection. But in this U-boat war . . . in the oceans the limitations of hunting forces have made themselves very clear'. We ourselves had travelled a long and hard road, and had wasted much effort in hunting for U-boats since 1939.13 One may hope that the conclusion quoted above will prove the final epitaph of the U-boat hunting group.
The Americans, however, certainly seem to have been slow in putting much of our experience into practice. They first tried every conceivable measure - except convoy and escort. Even 'Q Ships' were sent out, and one cannot but agree with Professor Morison's description of them as 'the least useful and most wasteful of all methods to fight submarines'.14 Yet the most surprising thing about American unreadiness is that, ever since Admiral R. L. Ghormley, U.S.N., arrived in London in August 1940 with a strong mission composed of some of his service's ablest officers, the policy of the Admiralty had been to give to the American Navy virtually the whole of our knowledge and experience, not excepting our latest radar developments. The writer of this history happened to be one of the members of the Naval Staff instructed to put that policy into effect; and he is confident that Admiral Ghormley himself, and the individual officers who accompanied him to deal with their own specific subjects, will agree that nothing of importance was ever
withheld from them.15 In fact Ghormley soon told the Washington authorities that he 'was obtaining information fresh from the laboratory of war, of priceless value to national defence'.16
In July 1942 the First Sea Lord wrote to Sir Arthur Salter, then head of the British Merchant Shipping Mission in America, reviewing 'the critical situation in the Battle of the Atlantic'. Admiral Pound said that he 'hoped and believed that we have taken them [i.e. the U.S. Navy] fully and frankly into our confidence and have given them all the information available, both on tactical and technical matters'. It is therefore plain that British and American records agree that all we had learnt from more than two years of war was given to the Americans; and it is also the case that numerous publications containing British doctrine and experience were issued by the Navy Department to United States' ships and establishments well before their country was at war. Whether our organisation for the control and protection of shipping could have been imitated earlier by a country which suddenly found itself plunged into a great maritime war may remain a matter for dispute. It is, however, justifiable to quote the German post-war comment that 'the U.S. Navy failed to profit from Britain's war experience'; and to remember that the cost of that failure (if such it was) in terms of tonnage sunk and lives lost was certainly not light.
The American historian has stated his conclusions regarding this disastrous period in such forthright terms that there is no need for us to dwell on them further here.17 From the British point of view the position was, however, very serious. Not only were the American escort vessels and aircraft, which had so recently begun to take a real share in the Atlantic battle, now needed in their own coastal waters, but British-controlled merchant vessels were being sunk in waters where the Admiralty's writ did not run, and after they had loaded for the long haul to the east or survived a westward passage of the Atlantic. As Admiral Noble put it to the First Sea Lord, 'The Western Approaches Command finds itself in the position today [8th of March 1942] of escorting convoys safely over to the American eastern seaboard, and then . . . finding that many of the ships thus escorted are easy prey to the U-boats . . . off the American coast or in the Caribbean'. Even though much of the shipping was not our own we could not remain passive in face of such a holocaust.
Returning now to the enemy's onslaught, the time which U-boats could spend in the more distant waters was greatly extended by the use of U-tankers (or 'milch cows' as the German Navy called them). The first of these (a converted 1,000-ton ex-Turkish boat) left Lorient on the 14th of March and fuelled three operational boats in the Western Approaches. She was followed a fortnight later by U-459, of 1,600 tons and the first proper U-tanker. Between March and August 1942 no less than six 'milch cows' made fuelling trips, and sometimes two or three were engaged on supply operations at the same time. These measures had the effect of about doubling the endurance of the 75o-ton boats. They were now worth nearly the same to the enemy as the large 1,100-tonners. Torpedo capacity and expenditure, rather than fuel supply, now became the limiting factor in the length of U-boat cruises in the Western Ocean.
After the initial onslaught off the east coast of America the enemy's intention was to send all his larger boats to the Caribbean, and five of them had arrived there by the middle of February. They sank many ships, especially tankers, in the first few days; one worked off Aruba, another penetrated into the harbours of Port of Spain (Trinidad) and Castries (St. Lucia), while U-126 sank nine ships in fourteen days between the Windward Passage and the Old Bahama Channel.18 Happily no large boats were available to relieve those of the first wave. Meanwhile the enemy tried to prevent us concentrating our counter-measures in the west by renewing his assaults off Freetown - an area which the U-boats had not visited since the abortive operation of October 1941.19 Early in March two U-boats arrived there and sank eleven ships (64,391 tons); but the American coast was seen to be the more profitable theatre, so no more boats were sent to west Africa for a time.
One of Hitler's 'intuitions' now caused a fortunate relaxation of pressure in the western Atlantic. At his conference on the 22nd of January he announced his conviction that Norway was 'the zone of destiny', and demanded 'unconditional obedience to all his commands and wishes concerning the defence of this area'. He decided that every warship and U-boat would be needed off that country's coast. Next day, however, in a manner typical of Hitler's desire always to have things both ways, he ordered that operations off the American coast were none the less to continue. On the 25th Dönitz received a totally unexpected order to send eight boats to the waters between Iceland, the Faeroes and Scotland to protect Norway from the anticipated invasion; and the final German defence plan envisaged the disposal of no less than twenty of the medium-sized boats
for that purpose. Though Dönitz himself protested vigorously against the diversion of his U-boats, the German Naval Staff seems to have made no serious attempt to counter Hitler's obsession by a reasoned argument against its probability. Nor did they even represent what the consequences would be in the Atlantic. Inevitably the weight of the offensive off the American coast declined, just at the time when it had proved highly profitable. In actual fact, the U-boats stationed between Iceland and the North Channel accomplished little in February and March, though two homeward convoys (SC 67 and HX 175) and two outward ones (ON 63 and ONS 76) were attacked in those waters. In passing, it is of interest to remark that in April 1942 Mr Churchill did tell the British Chiefs of Staff to examine the feasibility of a landing in Norway, with the object of relieving enemy pressure on our Arctic convoys.20 His proposal, however, never reached the stage of serious planning, because it conflicted with the basic Allied strategy, which was to strike first in North Africa.
After the successes of the first two months of 1942 it was natural for Dönitz to want to send every U-boat he could find to the American seaboard; but the vacillations among the higher German authorities diverted a substantial proportion of his strength to other waters. None the less the months of March and April saw the climax of the U-boats' successes in the west, in spite of the fact there were rarely more than six to eight boats actually operating at any one time. A new crop of U-boat 'aces', similar to the one which we had successfully harvested in March 194121, sprang into being. Between mid-March and the 10th of April Hardegen (U-123) sank eleven ships and Mohr (U-124) nine, while Topp (U-552), Muzelburg (U-203) and Lassen (U-160) each had five or six to his credit. They found most of their victims between New York and Cape Hatteras, but it was off the latter that, on the 14th of April, the American destroyer Roper achieved the first success for his country by sinking U-85. Gradually the U-boats now began to be driven from the shallow, profitable coastal waters; more and more did they find it necessary to retire further to seaward, especially on moonlight nights, to recharge their batteries.
While the enemy was achieving enormous, and one may feel largely avoidable, destruction in the west, three thousand miles away to the east his experiences were very different. British antisubmarine tactics and weapons, both surface and air, were improving rapidly; and, unknown to the enemy, radar had arrived in a form capable of being fitted in escort vessels and aircraft. The activities of our aircraft over the Bay of Biscay transit routes and the counter
blows of our air and surface convoy escorts were causing the enemy serious losses and much anxiety. The surface escort of convoy OS 18 sank U-82 on the 6th of February, that of the troop convoy WS 17 dealt similarly with U-587 in March, and in April U-252 was destroyed by the escort of OG 82.
The increasing importance now attached to Coastal Command's anti-submarine activities, and the slowly rising tide of their success, are indicated by the fact that in March, when the Chief of the Air Staff asked for statistics of the latter, the Admiralty replied that 'the steadily increasing efficiency of air attack' had been brought about chiefly by changes in tactics, by improved training and by the shallow-set depth charge. 'This increased potency' continued the Admiralty, 'will not only be maintained but should be further increased'.
After the first three months of 1942 had passed, American antisubmarine measures at last began to make themselves felt. This led to the U-boats being allotted to specific areas, instead of being allowed to rove where they wished in search of targets. On the 20th of April the first 'milch cow' (U-459) arrived 500 miles off Bermuda. There were at the time about a dozen boats in American and Caribbean waters. Refueling them started on the 22nd; twelve medium and two large submarines were soon replenished. It was thanks to this measure that, early in May, the enemy's strength reached a peak of some sixteen to eighteen boats operating between Cape Sable and Key West.22 The degree of success which Dönitz had expected from them was not, however, achieved-chiefly because a partial convoy system had by now been introduced. Only off Florida could the U-boats still stay in the shallow coastal waters, and there three U-boats sank ten ships. Further north the aggregate results accomplished by a much greater number of enemies were no larger.
The reasons for this long-awaited change for the better are not far to seek. On the 20th of May Admiral King, U.S.N., the Chief of Naval Operations, wrote to the First Sea Lord to say that 'because of your recent addition to [the] Caribbean Escort [Forces] and the inauguration of coastal convoys, I have hopes that matters can be got in hand to a better degree'. But he added that there were, as yet, no escorts for 'the vital Gulf of Mexico part of our common oil transport, which is now seriously threatened'. He therefore asked for fifteen to twenty more British corvettes to be lent for use on the American east coast, so that their own flotilla vessels might be released to the Gulf of Mexico. This request was carefully considered in London, and it
may be appropriate to review the employment of our corvettes at this time.
There were now two hundred of these little ships in service, including those of the Canadian Navy; but ten of them had already been permanently transferred to the Americans. Of the remaining 190 corvettes:
47 were at Gibraltar, at Freetown, in the Mediterranean, in South Africa, on the American Pacific Coast or in the Indian Ocean. 6 were working with the Russian Convoys. 37 were with the United Kingdom-based Atlantic Escort Groups. 78 were with the Atlantic Escort Groups based on Canada and Newfoundland. But two groups from these bases had already been allocated to the U.S.N.'s coastal escort forces, and four other corvettes were working with the special tanker convoys then running between the Dutch West Indies and Canada. 8 had been lent to the U.S.N.'s Caribbean Escorts. 14 were working with the Gibraltar Convoys.
It will be seen from the foregoing how widely our escort forces had to be dispersed to deal with the U-boat threat, and that the new commitments in the western Atlantic had already absorbed a substantial proportion of the corvettes. None the less Admiral Pound decided that he must try to do something to meet King's request. At the beginning of May we had, in response to an earlier American request, lent a British group for the Trinidad-Aruba convoys by reducing our mid-ocean groups (that is the Western Approaches ships which worked between the Western and Mid-Ocean Meeting Points) from twelve to eleven. Admiral Pound now proposed to meet the new American request by robbing the mid-ocean escorts of one more group. This, of course, meant that the remainder would have to be driven even harder; they would now average twenty-four out of every thirty days at sea, and we should have no margin to allow for foul weather or diversions. The First Sea Lord made the offer conditional on 'U.S. and Canadian escorts being equally hard worked'. A conference was held in Washington, and it was then found that by adjustments to the Caribbean convoys the diversion of further strength from the Atlantic could be avoided, provided the group already sent to those waters remained there.
To return now to the Caribbean, after the first attacks a month passed before it was again visited by the U-boats. In mid-April three were once again cruising in the profitable waters between Trinidad and Curacao, and off the Lesser Antilles.23 On the 18th one of them bombarded the oil storage tanks at Curacao; and by the 10th of May some nine boats were concentrated, chiefly off Trinidad. There shipping still moved in peacetime fashion, and a lot of it got sunk.
In May the German Naval Staff reviewed the U-boat war as a whole. Sinkings off the American coast had started to decline, while in Arctic waters poor results had been achieved. In the attack on PQ. 16 for example, five U-boats had been damaged for a return of only one ship sunk.24 Donitz wished to call off such operations, but his views did not prevail-unhappily for our next Russian convoy, the ill-fated PQ 17.25 The severe winter of 1941-42 had delayed U-boat training in the Baltic, with the result that only thirty-nine boats had been made ready for operations in the first quarter of 1942 and only thirty in the second quarter. Of these sixty-nine, twenty-six were sent to the far north and two to the Mediterranean. Twelve were lost in the first six months of 1942, so the net gain in the Atlantic fell far below German hopes, and was actually only twenty-nine boats. After the middle of the year the enemy's operational strength was increasing at the formidable figure of about twenty boats per month26; but the small total available early in the year, combined with diversions to unprofitable purposes, now seems to have been a decisive factor in the Atlantic battle. Dönitz's staff reviewed the Allied shipping position as well as this time. They believed that the Axis powers were between them more than accomplishing their aim of destroying 700,000 tons a month, the total considered necessary to bring about our defeat. But in truth the exaggerated claims put forward, especially by the Luftwaffe and the Japanese, misled them badly. The true rate of loss inflicted on us is shown in the table below.
Table 7. British, Allied and Neutral Merchant Shipping sunk by U-boats in all theatres January-July 194227
No. of Ships Gross Tonnage January 62 327,357 February 85 476,451 March 95 537,980 April 74 431,664 May 125 607,247 June 144 700,235 July 96 476,065 TOTAL 681 3,556,999 Monthly Average 97.3 ships 508,143 tons
Between May and July the U-boats gradually deserted the American seaboard, and concentrated in the Caribbean, though one entered the St. Lawrence estuary and sank two ships on the 12th of May. A few attacks were made in these months on Atlantic convoys, and three supply U-boats were now at sea. On the 14th and 15th of May the first north and south bound convoys sailed between Hampton Roads and Key West.28 This measure sounded, at long last, the knell for the U-boats' offshore operations on the American coast; but to accomplish it the entire carefully dovetailed British and Canadian Atlantic escort system had to be recast, and most of the recently gained help from American ships sacrificed.
Since easy torpedo targets had virtually disappeared, minelaying was tried by the U-boats off Boston, and in two other areas further south in July; but only three ships of some 19,000 tons were lost in the minefields. A second penetration into the St. Lawrence took place in June, but convoy had now been introduced there as well, and air escorts were much stronger than earlier in the year. By mid July three enemies had been sunk and several others damaged, and on the 19th the last two were withdrawn from the American coast.
Unfortunately, this success to the Allied anti-submarine forces was not the end of the story, for sufficient enemies were thereby released to make the Caribbean and Gulf of Mexico highly dangerous. About a dozen U-boats, supplied by 'milch cows', worked there in May and June, and they were soon reinforced by others which were sent down from the north. Sinkings became very widespread in those two months, but declined when the convoy system was extended to those waters. By early July most of the U-boats were working cautiously on the perimeter of the Caribbean, but there were still four of them in the Gulf of Mexico. It was plain that, just as the introduction of convoy and improved anti-submarine methods had forced the U-boats away from the American coast, the same process was now taking place further south.
In March and April Italian submarines appeared off the coast of Brazil, and early in May they were joined by three German boats. Two new supply U-boats were ordered to those waters, but they did not prove very profitable and the German boats were soon diverted to the Caribbean.
While these events were taking place in the west, the North Atlantic had been fairly quiet. No planned pack attacks had taken place since the previous November, and our convoys were only molested by U-boats based on Norway and by those on passage to the American coast. Early in May, however, a group of eight boats
allocated to the western Atlantic was ordered to move north and attack the convoys. The enemy had found out, from our wireless traffic, that we were again using the shorter 'great circle' route across the North Atlantic, and this meant that the U-boats need only shift their operational area a short distance. On the 11th of May convoy ONS 92 was sighted on a south-westerly course, and seven of its ships were sunk during one night. Contact was then lost by the U-boat pack. After refuelling from a supply U-boat the group continued operations and, on the 11th of June, attacked convoy ONS 100. A corvette and four ships (19,500 tons) were sunk. In the same month, with the idea of preventing us from further reinforcing the anti-submarine forces on the American coast, a group was sent to the waters off Gibraltar. Convoy HG 84 was attacked on the 14th, and lost five ships in one night; but the strength of the escorts and the effectiveness of our long-range air cover were greater than the enemy had expected, and he soon called off the U-boats.
On the 21st of June Hitler ordered that a U-boat group should be held ready in case we seized the Atlantic Islands - a project which had been as often discussed in German circles as in the British War Cabinet.29 Dönitz protested, though once more in vain, against this dispersal of his forces from what he knew to be their most profitable theatre; nor does there now seem to be any doubt that, but for Hitler's frequent diversions, the tonnage sunk by U-boats in the first half of 1942 would have been substantially greater. In spite of this the sinkings in June continued, from the enemy's point of view, to be highly satisfactory. There was a decline in the Gulf of Mexico, but this was offset by heavy sinkings in the Caribbean and off the eastern approaches to the Panama Canal.
In July the convoy system, started between Key West and Hampton Roads and between New York and Halifax in May, was extended further south. The 'Convoy and Routing' section under the Commander-in-Chief, U.S. Fleet, which was analogous to the Admiralty's convoy-control organisation, had now been made responsible for the entire United States Strategic Area.30 These initial steps towards the 'interlocking convoy system', which finally covered the western Atlantic routes as completely as the British coastal system ringed these islands, comprised the convoys shown in Table 8, page 107.
Table 8. Western Atlantic Coastal Convoys, July 1942
Convoy Remarks Halifax-Aruba and reverse (AH-HA) There were only four of these convoys. They were superseded in September 1942. Trinidad-Key West via Aruba and reverse (TAW-WAT) These were replaced after two months by Guantanamo-Aruba-Trinidad (GAT-TAG) convoys. Panama-Guantanamo and reverse (PG-GP) These were renamed GZ-ZG later. Trinidad-eastwards (TE) There were 17 of these convoys. They stopped in September 1942.
The above convoys, and also many others introduced later, linked in with the main flow of coastal shipping proceeding to and from New York. When the 'interlocking system' reached its final form the termini of the trans-Atlantic fast (HX-ON) and slow (SC-ONS) convoys were shifted from Halifax and Sydney respectively to New York, which then became the greatest entrepot of shipping in the world. The main north-bound coastal convoys, which had been fed from many secondary routes in the Gulf and Caribbean, and from as far south as Rio de Janeiro, were timed to reach New York on the day before the trans-Atlantic convoys to which they were dovetailed were due to sail. But at the end of the phase with which we are now concerned only the first steps had been taken in this direction, and many ships were still sailing independently, especially off the coast of central America.
In June an important measure which had long been desired by the Admiralty was introduced; convoy escorts began to be refuelled by tankers sailing in the convoy. It had been slow to arrive because, firstly, we had suffered (and still were suffering) from a chronic shortage of tankers; and secondly because special gear had to be supplied to the tankers and to the escorts, and their crews trained in its use. But as soon as tankers could be spared and the equipment provided, it was started, though at first in a small way. It simplified the organisation of escorts enormously, because it saved the wasteful process of sending groups out to overtake convoys at meeting points which, in bad weather, might be missed, and of the relieved groups either returning to base without a convoy or having to wait at a rendezvous for their next commitment. It also reduced the likelihood of escorts having to leave a delayed convoy for lack of fuel, and enabled us to begin again to use the more southerly, and shorter, Atlantic routes, which had been barred to us because they took the escorts too far from their halfway fuelling base in Iceland.
In spite of favourable trends such as those described above, the ebb and flow of the Atlantic battle was still, in the middle of I942, evenly balanced between the contestants. Though we now had short wave radar, high-frequency direction finding, the Leigh Light, and
air depth charges fitted with a new and more powerful explosive, and were using larger and deeper depth charge patterns, the enemy's strength was increasing rapidly.31 It was obvious to the Admiralty that the U-boats were still completing far faster than we were sinking them. To Dönitz and his staff the time appeared ripe to renew the pack attacks against our Atlantic convoys on an even greater scale. In reaching this decision he was much influenced by the need to send his U-boats to waters where air cover was still lacking, or was only spasmodic. The Gibraltar route had proved expensive, and the American coastal waters had become untenable. The central Atlantic 'air gap' plainly offered the best prospects, and plans were made to keep at least two U-boat groups permanently at sea in those waters. Thus was restarted the ding-dong battle between the U-boat packs and our Atlantic escort groups.
Convoy OS 33 was first attacked on the 18th of July, and lost five ships of some 32,000 tons. But the escorts sank U-136. The next outward Sierra Leone convoy was found by the same pack and lost two ships; but in this case the enemy noted with concern that aircraft were still with the convoy when it was nearly 800 miles out. In actual fact this was an exceptional accomplishment by Coastal Command, which had managed to scrape together only one squadron of American Liberators. Air escort at such ranges was not to become a common practice for another nine months. The month of July thus brought no marked success to either side on the Sierra Leone route, and attacks on two outward North Atlantic convoys (ON 113 and 115) produced similar results; in each case a few ships were sunk and the escorts destroyed one U-boat. In that same month the U-boats did, however, find one soft spot. Much of the traffic across the central Atlantic from Trinidad or New York still sailed independently, and two U-boats, having failed to do any damage off Freetown, sank a number of ships 500 miles further west.
As was inevitable the organisation for the protection of the Atlantic convoys was modified as the battle swung to and fro. By the middle of 1942 the usual practice was for American or Canadian groups to provide the first escorts of a convoy starting from New York. They would take it to the Halifax Ocean Meeting Point (HOMP) in about 61° West. There other Canadian groups took over duty and escorted the convoy eastwards to the Western Ocean Meeting Point (WESTOMP) in about 49° West. Then the mid-ocean groups, which might be Canadian or British from St. John's, or American from Argentia, took over for the deadliest part of the journey. During this passage an American-escorted convoy from Iceland might be met at the Iceland Ocean Meeting Point (ICOMP) in about 23° West; but the mid-ocean groups continued to the
Eastern Ocean Meeting Point (EASTOMP) near Oversay Island in the North-Western Approaches. There the last of the five escorts, the British local groups, took over; and the mid-ocean group went into Londonderry to fuel and, perhaps, to rest before taking out an outward convoy.
The Atlantic convoy cycles in use at this time meant that, on an average, four HX and two SC convoys sailed homeward every month, while an equal number of their outward counterparts (ON and ONS) left British ports: the number of ships in these convoys averaged about fifty, and each convoy needed at least seven escorts. To provide escorts for these twelve convoys as and when required at each stage of their journeys was a very intricate problem32; and often the carefully worked-out schedules were wrecked by unforeseen developments, by diversions, or by bad weather. The number of escort groups needed to fulfill the requirement is best shown in tabular form.
Table 9. The Organisation of North Atlantic Escort Forces, June 1942
(The full strength of an Escort Group was 9 ships and the average composition of groups was 3 destroyers and 6 corvettes).
Escort Force Zone of Responsibility Strength needed (on basis of 6 Homeward and 6 Outward convoys per month) Nationality Usual Bases Western Local Departure Port to 61° W (HOMP) 8 Groups British or Canadian Halifax or Boston Western Local 61° W (HOMP) to 49° W (WESTOMP) Mid-Ocean 49° W (WESTOMP) to 22° W (EASTOMP) 8 Groups British, Canadian or American St. John's Iceland Iceland ports to 25° W (ICOMP) 2 Groups American Argentia Eastern Local 22° W (EASTOMP) to arrival port 8 Groups British Liverpool or Clyde
To turn to the parallel development of air cover over the Atlantic, Central Command squadrons were now working regularly from Ireland, from bases in the west of Scotland and from Iceland. On the other side aircraft of the R.C.A.F.'s No. 1 Group were based on Yarmouth (Nova Scotia), Halifax, Sydney and
Gander (Newfoundland)33; U.S. naval aircraft worked from Argentia and Iceland. But the range of the Catalinas, Wellingtons, Whitleys and Hudsons did not enable continuous air escort to be provided, and the Atlantic 'air gap' was little, if at all, smaller than it had been in the previous phase.34 Not for another year could it be closed by shore-based aircraft. On the American seaboard a mixed and inadequately trained collection of U.S. Army, U.S. Navy and even of civil aircraft under the Commander, Eastern Sea Frontier, tried to counter the U-boats - and at first made very little impression on them.35 Not until mid-1943 did the American Chiefs of Staff order the withdrawal of army aircraft from anti-submarine duties, and place the whole responsibility on the Navy.36
On the eastern side of the Atlantic the Organisation of the Coastal Command's Groups had not changed since the middle of 194137, but their strength had increased and some new types of aircraft had entered service. It will be convenient to show these in tabular form.
Table 10. The Strength and Disposition of Coastal Command, June 1942
(Squadrons on loan from other commands included. Meteorological, Photographic, Air-Sea Rescue and Training Squadrons omitted)
Station No. of Squadrons Duty Type of Aircraft Gibraltar 1 Flying Boat Catalina and Sunderland Iceland Part of 1
Part of 1
Hudson and Whitley
Catalina and Northrop
15 Group (H.Qs. Liverpool) 3
4 (1 forming)
Sunderland and Catalina
16 Group (H.Qs. Chatham) 2
Torpedo-bomber and Minelaying
Liberator and Fortress
Beaufighter and Blenheim
18 Group (H.Qs. Rosyth) 3
Beaufighter and Blenheim
19 Group (H.Qs. Plymouth) 1
Hudson, Whitley, Wellington
The actual strength of the operational squadrons at that time was fifty-four flying boats and 490 reconnaissance and fighter aircraft. Of these, the average numbers available on any day were twenty-eight and 201 respectively. The Photographic, Meteorological and Air-Sea Rescue squadrons (about nine squadrons in all) totaled 137 aircraft and there were, in addition, four naval squadrons of thirty-six aircraft on loan to Coastal Command.
It will be seen that the growth and development of our surface and air Atlantic escorts, described in the preceding paragraphs, followed logically on the change in conditions brought about by the entry of the United States into the war. But whereas the British Admiralty had, under the old arrangements, carried the whole responsibility for the control of Atlantic shipping, there was now a clear necessity for the Americans to take a full share of that heavy burden. Planning for this eventuality had, in fact, started in January 1941 when the first British-American staff conversations took place in Washington.38 A year later-to be precise in February 1942 - it was plain to both parties that clearer definition of their strategic responsibilities was necessary. Further conferences took place and, as a result, on the 1st of July a 'Change of Operational Control' line (or 'Chop Line' as it came to be called) was established in the Atlantic. At first it corresponded approximately to the division between the U.S.A.'s and the British strategic areas - in general following the meridian of 26° West.39 Before a convoy sailed the anticipated time of crossing the 'Chop Line' was worked out; this was included in the sailing telegram addressed to all the authorities concerned in its movements. The routeing authority on the arrival side took over control at the time stated in the message, regardless of whether the convoy had or had not then crossed the 'Chop Line'. The line was altered from time to time as the requirements of the war necessitated; but the principle on which control was passed back and forth across the Atlantic remained unchanged to the end.
To sum up the results of the first seven months of 1942, the enemy had reason to be satisfied over the achievements of his U-boats. They had sunk an enormous tonnage of Allied shipping (681 ships of 3,556,999 tons in all, of which 589 ships of over three million tons were sunk in the Atlantic and Arctic theatres) - and at astonishingly small cost to themselves. Only 3.9 per cent of the U-boats at sea had been destroyed, and the rate of sinking inflicted on the Allies had been kept at the high figure of some 300 tons per U-boat per day throughout the period. A great proportion of the tonnage had, however, been sunk in American waters, and it was hardly to be expected that the very favourable conditions found in the western
Atlantic and Caribbean would continue. In fact, as has already been told, American anti-submarine defences and methods had begun to improve well before the end of the present phase.
In other waters the performance of the U-boats had been indifferent, and the outlook for the enemy was not favourable. In particular the fitting of radar sets in Coastal Command aircraft was now general, though the Germans remained sceptical of this achievement until many months later, when night attacks proved that they must be so fitted. The pressure of our patrolling aircraft, particularly in the Bay of Biscay, became far heavier in the spring. The first Leigh-Light squadron (No. 172) was formed early in March and by the beginning of June had four aircraft ready for work.40 The first 'dark night' attack quickly took place and the squadron's success was as rapid as the enemy's surprise was complete. A second squadron (No. 179) was formed at the beginning of September. Whereas for the first three months of the year no U-boat casualties had been caused in the Atlantic by British aircraft and only two by the Americans, in June three boats were severely damaged in the Bay. Dönitz's diary for the 11th of that month contains a remark that 'there being no defence against aircraft in the Bay of Biscay, the R.A.F. can do as it pleases'. The Germans guessed, correctly, that it was some new British development which was causing them surprise and discomfiture, and then set about devising countermeasures. It was not, however, until the next phase that U-boats were fitted with search receivers which could detect the approach of an aircraft using radar. In countering our wireless direction-finding the enemy showed less imagination. He knew that we listened to and measured the direction of the 'homing' signals which, under Dönitz's centralised control system, the U-boats were obliged to transmit; but no counter to our methods was devised. Actually it was in July that we first fitted high-frequency direction-finding sets in our escort vessels, and by the end of the year it was a standard item of their equipment. It made the location of U-boats much more effective than by listening from shore stations only. None the less, and in spite of the success of these new measures, it is disconcerting to find that after two-and-a-half years of war the enemy was still able to read many of our cyphered convoy control signals.
Though the enemy was, we now know, very disturbed by the increasing effectiveness of our defences, the Admiralty was far from satisfied-particularly over the small number of U-boats sunk by our aircraft in relation to the numbers sighted by them. However, by the middle of 1942 certain technical developments, to be referred to more fully later41, were at last beginning to give Coastal Command's
aircrews what the First Sea Lord's adviser on anti-submarine warfare called 'the means whereby they can deliver a deadly attack', and it was plain that better results would soon be achieved. The handling of our surface escort groups was also improving steadily. The Tactical School at Liverpool, in which all escort commanders underwent brief periods of training, contributed a great deal to this process. It was there that U-boat methods were studied, and counter-measures devised; by the introduction of standard procedures to be followed in the event of attack the flexibility and cohesion of our groups, and the speed and confidence with which individual ships reacted to any sudden emergency, all benefited greatly.
As to the enemy's strength, he started the present phase with 249 U-boats in commission; ninety-one of them were operational (sixty-four based on the Atlantic, twenty-three in the Mediterranean and four in the Arctic theatre). By the 1st of March the operational total had increased to 111, eighty of which were in the Atlantic; and by the end of June he had 140 at work. And, throughout the phase dealt with in this chapter, British, American and Canadian forces had only sunk thirty-two U-boats - approximately the equivalent to one month of the enemy's current production.42 It could not be doubted that the real crisis of the battle was still to come.
Table of Contents ** Previous Chapter (III) ** Next Chapter (V)