Shortly before the start of the phase now to be considered the First Sea Lord reviewed 'the present critical situation in the Battle of the Atlantic'. It was plain that the institution of convoy on the American eastern coastal routes had 'produced the anticipated decrease in losses' and that the same measure was now producing identical results in the Caribbean. Admiral Pound 'confidently predicted that the bulk of trade will now pass through these waters in safety;' but he considered it equally certain that we should 'have to face a heavy scale of attack on the focal area to the east of Trinidad'. Measures were in hand to deal with this probability. But the Naval Staff's reasoning and instinct saw even deeper than this into the enemy's mind. 'It is firmly believed', wrote Admiral Pound, 'that another turning point in the U-boat war is approaching'. He considered that at some stage, and probably quite soon, Dönitz would decide that the defences in the western Atlantic had become so strong that 'attack in those waters ceased to pay a return commensurate with the risk and with the lack of economy in this use of his U-boats.' For every boat which Dönitz could send to the western Atlantic, he could keep three at work in our own Western Approaches; for every convoy he could attack in those distant waters, he could deploy four or five times the strength against one in the eastern Atlantic. Meanwhile our own position was none too happy, because we had sent reinforcements to the American side just at the time when attacks on our arctic convoys were increasing; and we needed every escort vessel we could find on that perilous route as well as for the main Atlantic traffic.¹ The events now to be

recounted will show how extraordinarily accurate was the Naval Staff's prescient reasoning, and how well-founded were its apprehensions. The outlook was, in Admiral Pound's view, made even more foreboding because we and the Americans had not yet managed to make 'our dual control .., function as a single control based on a single, unified strategy'. No declaration of such a policy had yet been made and he felt grave doubts whether, if his forecast of the enemy's intentions proved correct, 'the United States authorities ... would be prepared to surrender some of their own forces, or even the British forces which have been helping their own efforts, to meet the new situation'. 'Clearly', continued the First Sea Lord, 'this was a lot to expect', but unless we could expect it we could not 'claim to be fighting on the basis of a unified strategy'. In the event, although the necessary readjustments were not made fast enough or in great enough strength to anticipate the enemy, the general purpose urged by Admiral Pound was carried out between ourselves and the Americans.

On the 27th of July, shortly after the paper summarised above had been produced, a broadcast by Dönitz gave a clear indication of his intentions. The Admiralty called this 'a tip straight from the horse's mouth'. The broadcast and succeeding Press interview were carefully scrutinised in the Admiralty, and the real motives behind them analysed; a great deal was found to confirm the impressions already gathered from other sources. The U-boats had not had a good month in July. Total sinkings had fallen to ninety-six ships of 4.76,065 tons compared with 144 of over 700,000 tons in June; and eleven U-boats had been sunk in that month (six in American waters and four in the Atlantic, all of them at the hands of the convoy escorts), compared with a total loss of only four in May and three in June.² Dönitz public warnings about the harsh realities of the U-boat war and the certainty that his forces would suffer heavier losses, might therefore reasonably be taken to indicate his intention to attack where losses were most likely to be incurred-namely around the Atlantic convoys.

On the British side it was realised that our escort forces were still too weak, and most of them too slow, to deal as we should wish with a renewed onslaught on a greater scale than in 1941. For lack of numbers the little ships were repeatedly deprived of the possibility of forcing a decision. They simply could not wait to conduct a prolonged and patient hunt, because it would deprive their charges for hours on end of perhaps a quarter of their protecting shield. Thus, when convoy SL 118 was attacked, the senior officer of the escort reported that 'again and again encounters which might have been pursued to a successful conclusion had to be prematurely broken off

---

in order to maintain a safe minimum escort with the convoy³. As to the speed of our escorts, it was perhaps now that we felt most acutely the slowness of the corvettes. Their margin of speed over the ships they were protecting might be only about four knots, which meant that if they dallied to hunt an attractive contact they might take hours to catch up the convoy again. True, slower convoys allowed them more time to hunt, but then the losses caused to the slower convoys were always far heavier (about 30 per cent) than to the faster ones. The real need obviously was for faster as well as for more escorts-and that could only mean more destroyers, of which we had always suffered a chronic shortage.

To enable us to reinforce the escorts of threatened convoys and then to hunt the U-boats to the death, it was in this phase-to be precise in September 1942 - that Support Groups were first formed. The earliest was the 20th Escort Group, of ten flotilla vessels and an oiler, under Commander F. J. Walker.' Some of its ships sailed on the 22nd of that month to reinforce the escort of ONS 132, but they were not allowed the chance to work for long together as an integrated group. In the following month the overriding need to provide for the safety of the troop and supply convoys to North Africa led to another postponement of a plan which the Admiralty had long cherished and repeatedly tried to introduce. The opening of the North African campaign also deferred the use of our few escort carriers with the mercantile convoys, for they too were diverted to help guard the invasion forces.⁴ The great days of the support groups and escort carriers on the North Atlantic routes were not to come for another six months.

Because the escort carriers were so slow in entering the Atlantic battle, the Admiralty decided at this time to fit a number of merchant ships with a flight deck so that, while still carrying normal cargoes, they could operate a few aircraft in defence of the convoy in which they were sailing. Two types of ship, grain-carriers of 8,000 tons and tankers of 11,000 tons, were chosen for conversion, and as a first step six of each type were taken in hand in October. They could carry three or four Swordfish each, and it was hoped to complete half of them by the early spring of 1943. In actual fact none was ready until May of that year, so that these 'Merchant Aircraft Carriers' (M.A.C. ships) had no influence on the battle during the period covered by this volume.⁵

At the start of this phase in August 1942 the enemy had a group of twelve U-boats to the south of Greenland, one of six boats off the Azores and another of the same strength off north-west Africa. There were four or five U-boats off the Canadian coast, some fifteen in or near the Caribbean and half a dozen off Brazil. Finally twenty U-boats were on passage to or from one or other of the theatres I already mentioned, and ten new ones were outward bound around the north of Scotland. His total strength had now passed the three hundred mark, of which approximately half were available for operations.⁶

The month of August saw, as the Admiralty had predicted, the last considerable U-boat forays in the Caribbean and Gulf of Mexico. Though air cover had now been greatly improved, the conformation of the islands forced the convoys to use certain well defined channels, such as the Windward Passage between Cuba and Haiti.⁷ It was there and to the east of Trinidad that most successes were now scored by the U-boats; the enemy had, early in August, discovered the focal waters in which the east-west shipping to and from Trinidad and the north-south coastal traffic intersected. Between the 20th of July and the end of August fifteen ships were sunk in the Caribbean and Gulf for the price of three U-boats; but the enemy's successes were steadily declining, and the surviving submarines were withdrawn early in September to the waters around Trinidad, and later to the mouths of the Orinoco. The former yielded a rich harvest in September (29 ships of 143,000 tons, to an average of about eight U-boats at work) and, contrary to the enemy's expectations, remained very fruitful in the two succeeding months. Seventeen ships of 81,742 tons in October and twenty-five of 150,132 tons in November there fell victims to the U-boats, and these remote waters temporarily gave the enemy his greatest successes of the time. The great majority of the ships sunk were still independently routed. It was not till October, when south-bound convoys from Trinidad (TS convoys) were started, that the southern sections of the Americans' 'Interlocking Convoy System' made their influence felt in this area⁸; and by that time the U-boats had begun to move elsewhere. The squadron of Coastal Command Hudsons mentioned earlier (No. 53) did good patrol work from Trinidad at this time and reported many sightings⁹; but no U-boats were destroyed.

The southward extension of the coastal convoy system was greatly facilitated by Brazil's declaration of war against the Axis powers on the 22nd of August. Although it is true that since the early days of

1942 the Brazilian Government had shown itself to be favourably disposed towards the Allies, the Germans brought its active hostility on themselves by typically callous actions. Brazilian ships had been sunk by U-boats at various times since the beginning of 1942, and tension had been rising. But on the 16th and 17th of August U-507 sank five in rapid succession close off Bahia, and this led immediately to a declaration of war. This may be considered an outstanding example not only of the Germans' political ineptitude, but of their lack of strategic insight. It was of course true that, measured in terms of ships, aircraft and fighting men, Brazil's assistance to the Allied cause was comparatively small; but the enormous length of her coastline and the fact that it juts far out in the South Atlantic were of inestimably greater advantage to us than her material aid. The Allied shipping control organisation could now be extended almost to the great focal area off the River Plate, defence of which was always one of Britain's major anxieties.¹⁰ But an even greater advantage was the stronger strategic control of the whole South Atlantic gained from the use of Brazilian bases. Natal and Pernambuco (Recife) were the closest points on the American continent to our African bases at Freetown, Bathurst and Takoradi¹¹; and so our watch was greatly improved over the narrowest part of the ocean, through which all our Middle East troop and supply convoys and a great stream of mercantile traffic still had to pass. U-boats would now find these waters less healthy, surface raiders were almost certainly debarred from them, and enemy blockade runners would be more easily intercepted. The importance to our cause of this development cannot be better demonstrated than by glancing at the maps in our first volume which show the depredations of the enemy's commerce raiders during the first two years of war in the waters from which they were now finally driven.¹²

 

Though it was an American responsibility and has been fully described in the U.S. Navy's history¹³, it may be desirable here to give an outline of the way in which the 'Interlocking Convoy System', already mentioned, worked. In essence it was the same as that organised long before in British coastal waters to feed to and from the main ocean shipping routes, at regular intervals, the traffic which started from, or was destined for ports beyond the ocean terminals. But in the western Atlantic the problem was a good deal more complex, because of the number of subsidiary routes involved.

The whole system was governed by two cardinal principles. Firstly the north-bound coastal convoys had to arrive at New York shortly before the Atlantic convoy which its ships were to join sailed for Britain; and secondly the lesser local routes were all linked into the two main coastal routes between Key West or Guantanamo and New York or vice-versa (called KN-NK and GN-NG convoys respectively).¹⁴ These 'trunkline' convoys ran on four or five day cycles, and the subsidiary routes generally ran at double those intervals, so that local convoys joined every alternate main coastal convoy. The first of the 'trunkline' convoys sailed in both directions at the end of August or early in September. Concurrently with these new measures the western termini of the trans-Atlantic convoys were shifted from Halifax (HX-ON) and Sydney (SC-ONS) to New York; and the Boston to Halifax convoys (BX-XB), thus rendered redundant, were stopped. It is relevant here to mention that the immense concentration of shipping thus funnelled into and out from New York became more than even that port could manage, and six months later the SC-ONS convoys were therefore transferred to Halifax.¹⁵

It is unnecessary to detail the many subsidiary convoys which were linked into the main coastal lines already described, but they are shown on Map 11. Professor Morison has given their full particulars and has stated that 'the inter-locking system proved its worth immediately. During the last three months of 1942 the Eastern, Gulf and Panama Sea Frontiers suffered no loss from enemy submarines'.¹⁶ Only off Trinidad (Caribbean Sea Frontier West) did sinkings continue at that time. Professor Morison further records that only thirty-nine ships were sunk between the 1st of July and the 7th of December 1942 out of the 9,064 which sailed in western Atlantic convoys-a proportion of less than one half of one per cent -and concludes that 'this record justified the convoy system'.¹⁷ British historians will no doubt agree with his conclusion; but posterity may well ponder on the combination of circumstances which prevented that achievement being realised many months earlier.

---

Meanwhile Dönitz had started to re-dispose his forces, very much in the manner foretold by the Admiralty. But before we turn again to the ocean convoy routes it is necessary to make a digression into the technical field, and review the new anti-submarine measures now becoming available to Allied surface and air escorts, and the improvements in his defences which the enemy was concurrently designing. Among our own developments the Leigh Light used in conjunction with airborne radar was of great importance. Together they placed the advantage of surprise in the hands of the attacking aircraft. In July and August the enemy lost three U-boats, all commanded by experienced men to Coastal Command's Bay of Biscay patrols. Furthermore the more powerful depth charge fitted with the new shallow-firing pistol, which entered service in mid-1942, at last enabled our aircraft to exploit their inherent tactical advantage of surprise with deadly effect. The Germans started to fit search receivers in their boats in August. The design was somewhat crude, but they sufficed to give U-boats warning of the approach of our aircraft, which were still equipped with the one-and-a-half metre radar set. Much of our advantage was thus temporarily lost, and it was at once realised that it could only be restored by giving our aircraft the new ten centimetre radar set already being developed.¹⁸ By October the Bay offensive, which had recently seemed to offer such great promise, had come to a halt. To recover the advantage was made more difficult by the fact that production of the new radar set for Coastal Command clashed with manufacture of a set for Bomber Command, from which greatly improved results in bombing Germany were anticipated. The latter command refused at first to forego any part of its claim on the new instruments; but the collapse of the air offensive against U-boats crossing the Bay of Biscay was regarded so seriously that the Air Ministry ordered the diversion of the first forty sets to the Leigh-Light Wellingtons. This, however, could only be a stop gap and was unlikely to be wholly satisfactory, because the set had been designed for a different air craft employed on a different function. The only adequate solution was to get the new sets from the U.S.A., where they were now being made and fitted to Liberators. The American authorities realised the acute nature of our need, and in October single Liberators began to come across equipped with the ten centimetre set. Inevitably some modifications had to be made on this side, and it was not till the end of January 1943 that No. 224 Squadron began to receive its new

equipment. There, for the present, we will leave the Bay offensive, since it was not until the next phase that the initiative was regained by Coastal Command's aircraft.

While Coastal Command's No. 19 Group was trying to deal with the U-boat traffic to and from the Bay of Biscay bases, the aircraft of Nos. 15 and 18 Groups were conducting a parallel offensive against the U-boats which were passing from German ports out into the Atlantic round the north of Scotland. Conditions in this 'northern transit area' were, however, more difficult. Not only were there fewer targets, but they were able to vary their routes far more widely than in the Bay of Biscay; weather conditions were generally far worse, and wireless communication often proved exceedingly unreliable. By the middle of 1942, however, patrols were being flown on a wide arc stretching from the passage between the Shetland Islands and Norway in the east to a line joining Iceland to Ireland in the west.¹⁹ The first success gained from this wider patrolling was the sinking of the valuable 'milch cow' U-464 by a U.S. Navy Catalina on the 20th August. As more searchlight-fitted aircraft became available night patrols were intensified. Many contacts failed to produce results, but on the 15th of September a Whitley of No. 58 Squadron sank U-261, and in the following month a Leigh-Light Wellington accounted for U-412. The Admiralty realised the need to introduce a two-pronged offensive in the waters north of the Shetlands by making surface vessels available to cooperate with the Coastal Command aircraft, but for a long time shortage of ships prevented this being done. At the end of October, however, Admiral Tovey was able to allocate three destroyers, but by then the outward flow of new U-boats had declined and no results were obtained. Next many of No. 18 Group's aircraft were sent south to reinforce the air cover for the invasion convoys for North Africa, and patrols against U-boats in transit declined.

Outside the Bay of Biscay and the 'northern transit area', on the main convoy routes, our surface escorts were now receiving a centrimetric radar set, with the result that the U-boats never felt safe when on the surface and within its range. Moreover they quickly found that our Iceland-based aircraft were reaching further south, thus narrowing the 'Greenland air gap' - the waters in which the U-boats greatly preferred to work.²⁰ In fact Dönitz's plan was to locate our convoys before they reached the air gap, then to concentrate against them while they were traversing it, and finally to withdraw when air cover returned to the convoys. Although in exceptional circumstances temporary air cover could be given at a distance of 800 miles from our bases, Coastal Command only had one

squadron of Liberators (No. 120) able to reach to such distances. The normal range of air cover was still only about 450 miles from the shore bases.

Though the Greenland gap was the most important 'zone of no air cover', there was a similar gap to the east of the Azores, which affected the Gibraltar and Sierra Leone convoys, and another in the neighbourhood of the Canary Islands.²¹ These too were used by the U-boats to their advantage. The enemy called the Azores air gap 'the black pit'. In it many ships were sunk; and he was often able there to replenish his U-boats from 'milch cows' as well.

The Germans felt, with good reason, that their own developments were not keeping pace with Allied improvements in antisubmarine tactics and weapons. Dönitz put great faith in the totally new design of submarine known as the Walter boat. In addition to normal means of propulsion these were to have turbines driven by gases produced from the combustion of diesel fuel and hydrogen peroxide. They would be capable of very high under-water speeds for short periods. But this revolutionary design suffered from long delays and troubles, and no Walter boat actually operated against us during the war. Meanwhile their radar lagged far behind our own. To give U-boats a better chance if caught by our aircraft on. the surface, heavier anti-aircraft armaments were fitted; and unsuccessful attempts were made to get efficient long-range fighters from the Luftwaffe to protect the U-boats. As to under-water weapons, the enemy was developing acoustic and zigzag-running torpedoes, besides improved magnetic torpedo pistols; asdic decoys, which could be released from a submerged submarine when being hunted, were also tried out, though without any marked success. The greatest improvement given to the U-boats was, without doubt, the ability to dive much deeper. The latest models could dive to 600 feet, or even deeper in emergency. But we were also setting our depth charge patterns to explode at greater depths, and releasing them in greater numbers.

Though the U-boats suffered therefore from tactical and technical handicaps at this time, there were still several important factors which acted in their favour; and of them Dönitz was able to take advantage in planning his new assault on the convoy routes. Firstly he now had a number of 'milch cows' available to refuel his boats and so extend their time on operations. Secondly, our shortage of escorts, and the acute fuel problems with which they were still beset, forced the convoys to keep close to the shortest ('great circle') route across the Atlantic. In the autumn the enemy commented on the way in which this inelasticity in routeing acted in his favour. Lastly his wireless intelligence was still working at a high pitch of efficiency; he was

once more able to read many of the cyphered and coded signals passing between our shore authorities and the convoys, and so deduce or anticipate their movements. Readers of our first volume will recollect that the initial successes of the German cryptographers were checked when, in August 1940, the Admiralty changed our cyphers.²² German records leave no room for doubt that, in spite of the change then made, by 1942 the enemy had achieved another substantial penetration of our cyphers; nor was it until the end of that year that our counter-measures began to take effect. Though its runs ahead of the stage now reached in our story, it is relevant to mention that it was not until May 1943 that the discomfiture of the highly skilled German cypher-breakers was made complete and final. The reader should not, of course, assume that we British were meanwhile idle in achieving the opposite purpose. None the less the successes of the enemy, and their long duration, will doubtless surprise those who believed that British cyphers were invariably secure against such encroachments.

Looking back to-day at the enemy's various endeavours to correct an adverse trend of which he was fully aware, one cannot but realise that British scientists had put into our fighting men's hands many developments of inestimable value, and that their accomplishments outstripped the enemy in many directions. But one of their achievements-the centrimetric radar set-stands out above all the others, for it returned to us the initiative in attack by night or in low visibility. Though it was, at this stage, only our surface escorts which were benefiting from it, a similar advantage would soon be placed in the hands of Coastal Command's aircrews, and a renewal of the Bay Offensive in greatly improved conditions would then be possible. It is, of course, the human factor rather than any technical development which is ultimately decisive in war; yet the effect on the fighting man of knowing that he possesses the tactical initiative is immense. And it is precisely that knowledge which, at this critical juncture, the scientists and technicians gave to our anti-submarine escorts and patrols.²³

Though technical developments were vitally important, they were by no means the only contribution made by scientists to the Atlantic

---

struggle. By mid-1942 a large number of them were working with the Admiralty and Western Approaches operational staffs, studying the results achieved by both sides and recommending strategic and tactical changes which could be deduced from them. 'Operational Research' had indeed been a wholly new development, but under the brilliant leadership of Professor P. M. S. Blackett it had now become a recognized element in planning our moves and disposing our forces.

The first Atlantic convoy to feel the weight of the enemy's new offensive was SC 94. By the time it was attacked on the 5th of August about 450 miles south of Cape Farewell, it consisted of thirty-three ships and had seven escort vessels in company: foggy weather and the distance from our air bases had for the time deprived the convoy of air cover. Next day a series of actions took place.²⁴ The Canadian destroyer Assiniboine rammed and sank U-210, but so injured herself that she had to return to base. Two other U-boats were damaged, and the depleted escort successfully held off all attacks, including those by the substantial reinforcements sent by Dönitz, until the afternoon of the 8th. Then five ships were lost. In the resulting confusion three more crews abandoned their ships under the impression that they had been torpedoed; two of them quickly returned on board, but the third refused to do so and their ship, though still undamaged, had to be left abandoned. She was sunk by a U-boat later. It was a rare event for British merchant seamen to act in such a manner.

The corvette Dianthus rammed and sank U-379 on the 8th, and another enemy was damaged. Again the escorts completely foiled many attacks, or forced the enemy to fire at such long ranges that the torpedoes missed. On the 9th Dönitz ordered yet more reinforcements to the scene, but that afternoon Liberators of No. 120 Squadron from Northern Ireland met and escorted the convoy at nearly 800 miles from their base, while the U.S. Navy's Catalinas from Iceland reached south towards the convoy as far as they could. The surface escort was also reinforced, and together they temporarily gained the upper hand. But the advantage was only temporary. Next morning, before the first Liberator had arrived, four ships were sunk; but from noon till dusk air escort was almost continuous and no more ships were lost.²⁵ Though many enemies were attacked by the Liberators and Catalinas, none was damaged; but it was largely the watchful pressure of the long-range aircraft which forced the U-boats to abandon the operation. On the 13th the surviving twenty-two

ships reached British ports. Eleven of 53,000 tons had been lost; but considering that all but one of the eighteen U-boats taking part had at one time or another been in touch with the convoy, and that two of them were sunk and four others damaged, the results of the five day battle were not unfavourable to the Allied cause.

Early in September the outward convoy ON 127 suffered heavily when outside the range of air cover. Seven of its ships and the escorting R.C.N. destroyer Ottawa were sunk, and four other merchantmen damaged, without any retribution having been exacted. It is, however, to be remarked that none of the escorts of this convoy had been fitted with radar. The same month saw heavy attacks on two Russian convoys (PQ 18 and QP 14), but we shall tell their story in another chapter.²⁶ Late in September the enemy failed against HX 209, and lost two U-boats to air attacks south of Iceland. Of the twenty-nine ships sunk in convoy during the month, twenty were lost in the North Atlantic.

On the Sierra Leone route our experiences were very similar. The U-boats waited in the 'Azores air gap' and tried to attack the convoys before they could be reached by shore-based aircraft. For example SL 118, which sailed on the 14th of August, lost three ships between the 16th and 17th. Then it was met by a Liberator from Cornwall, 780 miles out, a U-boat was promptly damaged and only one more ship was sunk. Gibraltar-based aircraft also helped to protect the SL and OS convoys while they were passing within their range.

That Dönitz was by no means happy over the first fruits of his new offensive is shown by entries in his war diary at this time. 'The number of British aircraft in the eastern Atlantic', wrote the Admiral, 'has increased and a great variety of them is seen. They are equipped with an excellent location device. U-boat traffic off the north of Scotland and in the Bay of Biscay is gravely endangered . . . by patrolling aircraft. In the Atlantic the enemy's daily reconnaissance . . . forces us to dispose U-boats far out in the centre of the ocean . . . There are also some aircraft of particularly long-range which are used as convoy escorts. They have been met 800 miles from British bases'. All of which was a true and accurate summary of the capacity and employment of our air escorts and patrols.

It was in this same month of September 1942 that an incident took place which had lengthy repercussions. Four U-boats and a 'milch cow' left Lorient in mid-August to work just south of the equator. There, on the 12th of September, U-156 torpedoed the troopship Laconia (19,695 tons), which had 1,800 Italian prisoners on board. Dönitz ordered other boats to go to the rescue, and the Vichy Government was asked to send help from Dakar. While U-boats

were collecting survivors they were bombed by American aircraft, and this led to the issue by Dönitz of the order subsequently known as the 'Laconia order', directing that survivors of ships sunk were not to be rescued. At the Nuremberg trial of Dönitz this was held to have been a violation of the Protocol of 1936²⁷, even though it was not proved that he had actually ordered the killing of survivors.²⁸

We have so far considered only the fortunes of our Atlantic mercantile convoys; but it was during the present phase that the 'monster' liners first started to carry troops (most of whom were American) on the same route, and it is to them that we will now briefly turn. There were six such ships under British control - the Queen Elizabeth (83,675 tons), the Queen Mary (81,235 tons), the Aquitania (44,786 tons), and the Mauretania (35,739 tons), all of the Cunard-White Star fleet, the French ship Ile de France (43,450 tons) which had been requisitioned in 1940, and the Nieuw Amsterdam (36,287 tons) which was on charter from the Dutch. They had already done a prodigious amount of steaming between Australia, New Zealand or India and the Middle East, and from the west coast of America to the Antipodes; and they had carried thousands of troops of many nationalities safely to their destinations.²⁹ Now the need to move American troops to Europe was so urgent that it was decided not only to accept the risk of employing them in the North Atlantic, but also greatly to increase the numbers carried on each such voyage. Thus the 'Queens', which had carried 6,000 men each formerly, had their carrying capacity increased firstly to 10,500 and then, in June 1942, to no less than 15,000 men. The risks were severe, for one torpedo could bring disaster on an appalling scale. Their safety lay only in their own speed of about 28-1/2 knots; but this itself brought danger, for it prevented them being escorted except at the start and finish of their journeys. No destroyers could maintain such a speed long enough to provide continuous escort right across the Atlantic. Their passages were known as 'operational convoys' and, when in the British strategic zone, they were always controlled by the Admiralty. Special routes were devised for each journey, and diversions from those routes were ordered as soon as any sign of U-boat activity occurred on their tracks. Their progress was continuously and anxiously watched from the Admiralty's Operational Intelligence Centre. A monster liner might thus be routed from New York far south into mid-Atlantic, thence almost due north towards

Iceland, and finally approach the Clyde down the sheltered waters of the Minches off western Scotland.

The liners were escorted out from Halifax or New York by American or Canadian destroyers, but after the first few hours these left, and they then remained entirely on their own until they were met by the Western Approaches escorts, consisting perhaps of an anti-aircraft cruiser and six destroyers, in about 12° West. The Queen Mary started work on the north Atlantic route on the 7th of August 1942. A month later the Queen Elizabeth joined her sister, and she had made ten Atlantic crossings before the end of the year. The other four great liners continued meanwhile to work on the more distant routes in the south Atlantic, the Indian Ocean and Pacific. Between July and December 194,850 American troops were safely carried across the north Atlantic to Britain.

'Bearing her load of lives, over and back,
The great Queen passes, scorning the deep-sea pack
Snarling below; in crimson, gold and rose
The skies salute, waves curtsey as she goes'.³⁰

Only one mishap, though a serious one, marred the great liners' accomplishments at this time. On the 2nd of October, just after the Western Approaches escort had joined the Queen Mary, she rammed and sank the anti-aircraft cruiser Curacoa. This old ship was slower than the liner, and the accident happened while she was escorting from ahead on a steady course, with the Queen Mary zig-zagging across her wake. Unhappily 338 lives were lost. On the issue of responsibility, which was taken much later to the House of Lords, it was finally held that blame was attributable to both ships in the proportion of two-thirds against the Admiralty and one-third against the Cunard-White Star Company.

To revert now to the Atlantic trade convoys, after the failure against HX 209 some of the U-boats refuelled, and then two long enemy patrol lines were established, one on each side of the north Atlantic. SC 104, originally of forty-seven ships, was sighted on the 12th of October by the eastern U-boat group, and in the two following nights one of them sank seven of the convoy, including a large tanker. The mid-ocean escort consisted of two destroyers and four corvettes (the latter all Norwegian-manned) ; but the westerly gale and heavy seas at first gave them a very difficult time. On the 15th the weather moderated and the escorts found their task easier. The destroyer Viscount rammed and sank U-619 that night; next afternoon in low

visibility the senior officer's ship, the Fame, destroyed U-353. As a Coastal Command Liberator of No. 120 Squadron sank a third enemy (U-661) not far from the convoy's track, and only eight of its ships in all were lost, the battle did not go wholly in the enemy's favour.

As October drew to a close the pressure on the convoy routes increased. HX 212 lost six ships, and a few days later a lucky wireless interception enabled the enemy to make a heavy concentration against the slow convoy SC 107. Fifteen ships of about 88,000 tons had been sunk before the air escorts arrived and forced the attackers to desist. Further south SL 125 was attacked off Madeira by ten enemies. In a seven-day battle thirteen of its ships went down, and no U-boats were destroyed. But the ill fortune which overtook this convoy appears to have benefited the Allied cause, quite unexpectedly, in another direction. The first military convoys for North Africa were passing through adjacent waters at the time when the U-boats were occupied in attacking SL 125.³¹ Had the enemy not been thus engaged he might well have detected the great movements of troop and supply ships, have attacked them or guessed their purpose and their destinations, and so deprived our landing forces of the important advantage of surprise.

As soon as the enemy realised that we had launched an invasion in North Africa, Dönitz re-deployed a large proportion of his strength off the disembarkation ports. Fifteen U-boats were sent to the Moroccan coast, but they arrived too late; the Allied air and surface defences had been given time to organise themselves, and the enemy inflicted few losses.³² Early in November one group of U-boats was sent into the Mediterranean to work off Algiers and Oran, and other reinforcements arrived off Gibraltar. There our traffic was heavy, but the air and surface defences were strong and the enemies were kept well in check. We lost a few valuable ships, but the U-boats themselves suffered severely. Three were sunk and six badly damaged to the west of Gibraltar in the second half of the month; and those inside the Mediterranean also fared ill. At the end of the month the U-boats in the approaches to Gibraltar were withdrawn further to the west, in order to catch the troop and supply convoys coming from America direct to North Africa. In this too they failed. None the less our total losses to U-boats in November were very high - 119 ships of 729,160 tons; but a great proportion of these, no less than 70 ships, were 'independents', and of that number the majority were sunk in the two 'soft spots' which the enemy had found, off the Cape of Good Hope and in the waters around Trinidad.

We shall recount later the story of the great movements by sea which preceded the successful launching of operation TORCH.³³ Here it is only necessary to consider the effect of those movements on the Atlantic struggle. The British and American Governments and the Combined Chiefs of Staff were all determined that the success of this first major Allied offensive must take priority over all other needs. The demands which it made on the Royal Navy for escorts were, inevitably, very heavy; not less than 125 flotilla vessels and fifty-two minesweepers had to be found. This could only be done by temporarily stopping the Russian, Gibraltar and Sierra Leone convoys, by holding back all reinforcements destined for other theatres and by stripping the Home Fleet and the British coastal convoy routes almost bare of flotilla vessels.³⁴ Parallel demands were, of course, made on Coastal Command to provide special protection to the TORCH convoys which sailed from Britain.

But the opening of the North African campaign did not eliminate the need for large numbers of merchant ships to sail between Britain and the South Atlantic, even though the convoys in which they would previously have sailed were suspended. This need was met by recasting the routes taken by the whole of this traffic, a feat which would have been quite impossible but for the centralised control of shipping exercised by the Admiralty. Homeward-bound ships from the Cape and from ports in West Africa, and those starting from South American ports north of the River Plate were now routed independently to Trinidad, whence they would join convoys to North America, and ultimately cross the Atlantic in HX or SC convoys. Fast ships of adequate endurance were allowed to miss Trinidad, where congestion was in any case serious, and proceeded direct to the American eastern seaboard. Lastly ships from the River Plate and a proportion of those sailing from South Africa were routed through the Magellan Straits, up the west coast of South America and then by the Panama Canal to the Atlantic convoy assembly ports. After the OS convoys were stopped, outward-bound ships from Britain to the south started in certain ON convoys which

were taking a southerly route, and broke away from them in the vicinity of the Azores. Thence they sailed independently to the west, to South Africa, or to South America.

It will readily be understood how this great re-organisation increased the length of the journeys, and so slowed down the turnround of the shipping on which the British war effort entirely depended.³⁵ But the risks and difficulties had to be accepted for the sake of the success of TORCH.

It now seems surprising that a heavier price was not exacted from the northern convoys for the successful lighting of the TORCH. Their surface escorts had certainly been temporarily weakened, but this may have been balanced by the enemy's diversion (too late) of much of his strength against the overseas expedition.

Air escorts were less affected than the surface escorts, because Coastal Command had sufficient strength in medium-range aircraft to meet the new requirement, and No. 120 Squadron, which possessed the only Liberators in the command, continued to meet emergency calls for long-range air cover. The most important consequences were, perhaps, that all the eight escort carriers (four British and four American) were diverted to meet the needs of the offensive, and that the employment of Support Groups to aid threatened convoys had again to be postponed. At the beginning of November the enemy had forty-two U-boats between Greenland and the Azores, sixteen in the eastern Caribbean and the 'Atlantic narrows' between Africa and Brazil. Seven were off the Cape of Good Hope and six off the Central African coast; ten were dispersed after attacking SL 125³⁶, and about twenty-eight were on passage homeward or outward.

Convoy SC-107 had been reported off Newfoundland on the 30th of October. The first attackers were sternly handled by the Royal Canadian Air Force, which sank U-520 and U-658. But seven U-boats made contact on the 1st of November after the convoy had passed beyond the range of air escorts. In two successive nights

fifteen ships were sunk. Then aircraft from Iceland joined, a Liberator sank U-132 and the attacks were called off. A little later the enemy located ONS 144 when it was out of range of air cover. On the 17th and 18th of November five ships and one of the escorts were lost, but the Norwegian-manned corvette Potentilla sank U-184. By the end of November more U-boats were available to throw into the battle on the convoy routes. Early next month HX 217 was pursued by no less than twenty-two enemies; but it had powerful air protection at a critical time and only lost two ships for an equal number of U-boats sunk by the air escorts. The next convoy attacks were substantial failures, and it was not till nearly the end of December that the enemy again achieved any great success. Then ONS 154 was attacked and lost thirteen ships as well as the special service ship Fidelity.³⁷ The latter, like a good many of our more important merchant-men, had the Admiralty's net defence against torpedoes. This protection was fitted to 768 merchant ships in all, and it certainly saved some of them; but it slowed the ships down and was difficult for the crews to manage in heavy weather. In the Fidelity's case it took five torpedoes to sink her.

Towards the end of this present phase an important change took place in the command of the British forces engaged in the Atlantic battle. On the 19th of November Admiral Sir Max Horton, who had commanded our home-based submarines since the early days of 1940, succeeded Admiral Sir Percy Noble as Commander-in-Chief, Western Approaches. Admiral Noble had been Commander-in-Chief since February 1941, when the Western Approaches headquarters were moved from Plymouth to Liverpool.³⁸ His period of command saw immense progress made in the formation and training of the escort groups, and in the full integration of our sea and air forces. But he accomplished far more than the conquest of many tactical, technical, and administrative problems. He recognized from the earliest days that the Battle of the Atlantic would ultimately be won by the side whose morale was the higher; that to achieve a morale which would overcome all difficulties, and would rise above all tragedies and set-backs, demanded that the Captains of the escort vessels and aircraft should have complete confidence in his shore organisation. So he constantly went to sea in the little ships and flew in the lonely aircraft of Coastal Command, sharing their dangers and their discomforts. Thus the crews came to learn that their Commander-in-Chief understood their difficulties and their problems; and links of mutual confidence of inestimable value were forged.

Though the strength which he had been able to deploy had never been sufficient to gain and keep the upper hand over the U-boats and the bombers, he had brought the country safely through the first great crisis, and he turned over to his successor not only the scores of ships which had been commissioned and trained, but also a smoothly running operational organisation in which his own staff and that of No. 15 Group of Coastal Command worked together in intimate harmony. His next appointment was head of the British Naval Mission in Washington and representative of the First Sea Lord on the American side of the Combined Chief of Staff's Organisation. Admiral Horton brought to his new command exceptional experience of submarine warfare dating back to the 1914-18 war, in which he had proved himself an outstanding commander. Moreover, he possessed a deep grasp of all the intricate human and technical problems involved in submarine warfare. There was no living officer who better understood the U-boat commander's mind, nor could more surely anticipate what his reactions to our countermeasures would be. Though the British submarine service to a man deplored his departure from its headquarters, all knew that he had been called to carry even greater responsibilities, and in a crisis which was becoming ever more plain. With his knowledge and insight, his ruthless determination and driving energy, he was without doubt the right man to pit against Dönitz.

During the closing days of 1942 the Admiralty reviewed yet again the problems and prospects of the Atlantic battle. 'Our shipping situation' reported a senior member of the Naval Staff, 'has never been tighter' ; and our surface and air escorts were still far too few. In spite of the success of the North African landings, grave anxiety was felt that future offensive plans might be delayed or even frustrated for lack of shipping. In particular, fuel stocks had fallen to a very low figure.

In mid-December there were only 300,000 tons of commercial bunker fuel in Britain, and consumption was running at about 130,000 tons a month. The Admiralty held another million tons which could be used in emergency, but if the naval stocks were allowed to run down the fleet might be immobilised. 'An ample reserve of fuel on this side of the Atlantic is the basis of all our activities' reported the Admiralty; and when the Prime Minister was given the figures quoted above, he minuted on the paper 'This does not look at all good . . .' To expedite and increase fuel imports it was proposed to open up the North Atlantic convoy cycle from eight to ten days, and to use the escorts thereby released to bring across forty-ship convoys of tankers direct from Aruba in the Dutch West Indies on a twenty day cycle. These proposals were put into effect in the next phase, but as we then suffered more heavy

losses, it was many months before our stocks of fuel had increased appreciably. The Admiralty also reviewed at this time the principles on which we should defend our convoys. One member of the Board summed up the problem to the First Sea Lord in these words. 'Experience shows quite clearly that surface escorts without air co-operation cannot give sufficient security to convoys, unless they are in overwhelming strength. It is also clear that air escort unaided by surface vessels is not sufficient. The most effective and economical use of our resources requires a careful balance in the combined use of surface and air escorts'. We had indeed travelled a long way since 1939.³⁹

As to the losses we had suffered during the year, it was beyond question that the enemy had done us great damage. At the time he believed that he had destroyed over seven million tons of shipping, and had therefore nearly achieved the target which he considered necessary to bring us to our knees. In fact the U-boats sank, in all waters, 1,160 ships totalling 6,266,215 tons; but his other weapons increased our total losses to no less than 1,664 ships of 7,790,697 tons.⁴⁰ To offset this enormous total, just over seven million tons of new Allied shipping had been built. A further deficit of about a million tons of shipping had thus been added in 1942 to the unfavourable balance shown in each year's accounts since the start of the war. British imports fell below thirty-four million tons-one-third less than the 1939 figure.

The U-boats had accomplished their share of this prodigious destruction with less strength than the enemy had hoped to receive; for only seventeen new boats had entered service each month instead of the hoped-for score or more. Yet he had started the year with ninety-one boats operational out of a total strength of 249, and ended it with 212 out of 393.⁴¹ Eighty-seven German and twenty-two Italian submarines had been sunk or destroyed during the year - an insufficient figure to offset the new construction.⁴² To the British Admiralty it was plain that the Battle of the Convoy Routes was still to be decided, that the enemy had greater strength than ever before, and that the crisis in the long-drawn struggle was near.

Table of Contents
Previous Chapter (VII) ** Next Chapter (IX)