There is substance in Lieutenant-Commander D. W. Waters’s assertion that ‘virtually every surface and air anti-submarine lesson of the first submarine war had to be, and ultimately was, re-learnt in the second at immense cost in blood, tears and treasure’. One would have thought that the entire Service knew that the most important lesson of the First War was that the U-boat attack on the merchant fleet was Britain’s most serious danger, and that it was only the introduction of convoy in 1917 that had saved the day. But the anti-submarine lessons of the war, which had never been fully understood anyway, were quickly forgotten after the war because there was no serious attempt to study the larger meaning of the U-boat campaign of 1917–18. During the interwar years, consequently, the convoy system was understood imperfectly at best. Although Captain Roskill was off the mark in stating that in 1919–39 there was not a single exercise in the protection of a mercantile convoy against air or submarine attack, the fact is that the Navy paid all too little attention to convoy work between the wars.
Ignorance was doubtless the chief explanation of the indifferent attitude towards convoy during much of the interwar period. The Admiralty’s German Navy expert, who was in charge of the captured German naval archives, has written:
A point that has emerged with startling clarity from all our researches into British and German records since the end of the Second World War is that no historian writing between the two wars (either British or German) drew the full and accurate conclusions from U-boat operations of 1917–18. The principal reason for this omission was that in those between-war years the full records of both sides were never available to any one historian, as they are available today. In this country the fact that we had eventually defeated the U-boats, and the advent of asdics shortly after the end of the First World War combined to produce in many officers an attitude of overconfidence in regard to any resurgence of the U-boat menace. In Germany, on the other hand, the researches of Admiral Spindler (the historian of the 1914–18 U-boat operations) were never completed. His work only went as far as 1917, and therefore did not include many of the lessons of U-boat operations against convoys.
Contributory causes of the failure to profit fully from war experience were (1) the old obsession with the battleship and fleet actions, which will be dealt with below; (2) an over-confidence, particularly in the 1930s, in asdic, the device that had been developed since 1917 as the answer to the problem of locating submarines; (3) the antipathy of many senior officers to what was falsely regarded as a defensive, to say nothing of a generally dull and monotonous, measure. Concerning the last, Admiral of the Fleet Sir Caspar John (among many others) bears out my contention: ‘You are very correct in writing that Convoy protection was regarded with martial antipathy by the Navy-it was too defensive in outlook for peacetime training–and, anyway, unlike battleships, there was never a visible convoy to “protect”.’ This attitude is borne out by the fact that in general the commands of fleet destroyers rather than of convoy escorts were regarded as the plums. Consequently, although, of course, there were some brilliant exceptions, the best officers were with the Fleet and the second team with the convoy escorts. Nor did it help that until the last prewar years it was the assumption that Japan would be Britain’s principal enemy in a war, not Germany, and the problem here was how to get at Japan across the world, not how to escort merchant ships across the Atlantic.
I do not want to leave the impression that progressive thought on convoy was entirely absent. The President of the Naval War College at Greenwich during 1934–7, when over a hundred officers went through the war course, recalled that ‘neither staff nor courses had any doubt on this subject. It was in fact Common Doctrine that convoy had rescued us in the first war and that it would be necessary in the future. So I cannot understand the Financial Secretary’s speech. It certainly had no effect on our teaching and as we were in close touch with the Admiralty we should have known if they thought differently.’ All that I maintain is that there was always a body of naval opinion which, through a failure to analyse the U-boat war of 1914–18, or for one or more of the other reasons mentioned above, preserved an anti-convoy outlook. The remarks of the Parliamentary and Financial Secretary of the Admiralty, Lord Stanley, speaking in the House of Commons on 14 March 1935, sum up the views of the Board and the Naval Staff at that date and are a reiteration of all the standard objections of the anti-convoy school of thought (or prejudice):
I can assure the House that the convoy system would not be introduced at once on the outbreak of war. Even the right hon. member for Swindon [Dr. Addison] would admit that the convoy system has very great disadvantages, and it certainly would not be welcomed by the trading community until conditions had become so intolerable that they were prepared to make the necessary sacrifices. In the first place, you would get delay at each end. You would get delay while the ships assembled at the starting point to be taken up by their convoy. You would get delay by the ships arriving at the same port at the same time. You would also have the difficulty of the faster ship having to go at the same pace as the slower one. Therefore, the convoy system will only be introduced when the balance of advantage is in its favour and when sinkings are so great that the country no longer feels justified in allowing ships to sail by themselves but feels that for the protection of their crews the convoy system is necessary.
(Dr. Addison:) Am I to understand the Noble Lord to suggest that the Admiralty would wait before instituting the convoy system until so many ships had been sunk that the country would not stand it any longer? Surely, they are not going to wait until such conditions arise as occurred on 17th April, 1917, when 34 ships were sunk one night. Are they going to let us get to that pitch before they start the convoy system?
(Lord Stanley:) Certainly not, but it will not be introduced in the first place. You will not know in the first place whether the ships are going to be in any great danger. It may be that it will be safer for them to sail by themselves. They will be a smaller target. The enemy ships would not know where they were to be found. If raiders were about we should have to institute the convoy system at once. It is simply a matter of expediency. We should be ready to put the scheme into operation but we should wait until we thought that the proper moment had arrived. Having got to the point when it is considered that the ships ought not to sail by themselves but should be protected by an escort, we have to decide what is the best form of protection for the convoys, and I think it is agreed by everybody that what is known as the general convoy is the best system. That is the convoy which has an escort ready to protect its ships from surface attack, from submarines and possibly from the air….
Therefore, we must put the provision of sloops into its proper order of priority. In doing that, I would ask the House to remember two things, first, that our anti-submarine defences and devices for finding out exactly where submarines are are so very much better than they were during the War that we should want fewer protective vessels in the convoy. Secondly, that as convoys will not be needed immediately on the outbreak of war it will give us time to improvise protection by destroyers and trawlers whilst orders are given to build the sloops which we shall eventually require.
The Naval Staff did not realize that, due to the closure of dangerous routes for days at a time, independent sailings had entailed even longer delays in 1917–18, while convoys guarded by escorts steamed directly to their destinations. Although it is true that Naval Staff officers had by 1935 come to favour convoy in principle, they did not think that it would be needed, at first, anyway, since the enemy, afraid to alienate neutral opinion as in 1917, would not launch unrestricted air or U-boat attacks on shipping. I should also mention that Germany was a signatory to the Submarine Agreement of 1936, which prohibited unrestricted submarine attack. Of course, we now know that Hitler’s word was worth nothing, but that could not be assumed at the time, at least openly. To proclaim a convoy system would have been to imply that the treaty was being, or would be, deliberately broken! The Air Staff, on the other hand, opposed convoy, using the discredited argument of 1917 that the massing of ships in convoy would only invite air attack and heavy losses. Criticism forced a modification of policy. In 1937 the Naval and Air staffs came to an agreement that convoy should be instituted at the outbreak of war. In March 1938, to satisfy naval opinion, the Admiralty undertook to make all preparations for convoy (for instance, Naval Control Service Officers were dispatched to all shipping ports), but not necessarily to institute it in the event of restricted submarine warfare. As the Deputy Director of Plans observed early in the war: ‘Our pre-war A/S plan was to attack U-boats with hunting groups until it became necessary to go into convoy …’ Ships were to continue to sail independently, if the enemy confined himself to restricted warfare -that is, stopping prospective victims and giving them time to evacuate passengers and crew. Having made this decision, the Admiralty neglected to provide the necessary convoy escorts for unrestricted warfare, under which ships were sunk without warning. All doubts were cleared up almost immediately upon the outbreak of war: Athenia torpedoed (against Hitler’s orders) on the first day, 3 September 1939; first convoy sailing, 6 September.
However, despite Britain’s stronger navy, assisted by Canada and the United States, it took nearly four years (i.e. not until May 1943) to overcome the German submarine menace. There was an insufficiency of escorts, unsuitable types, and inadequately trained groups, a diversion of anti-submarine vessels in the early part of the war from escorting convoys to futile offensive action by ‘hunting groups’, and a lack of air power on the convoy routes, particularly very long-range aircraft and escort carriers. (It was 3½ years after the outbreak of war before there was a single true escort carrier on the North Atlantic convoy route.) All this was in part a reflection of the low esteem in which convoy was generally held between the wars and indeed into the early stages of the Second War. It should be pointed out that Western Approaches Command did a great job with the materiel and personnel available, and that its C-in-C (1941–2), Sir Percy Noble, was against the ‘hunting group’ concept.
As regards air power, forgotten in the interwar years was the highly successful role of naval aircraft as a convoy escort in 1917–18, when a mere five ships were sunk in convoys with a surface and air escort. There were virtually no aircraft available for convoy when war came, since the responsibilities of naval aircraft did not include the protection of merchant shipping. One cause of this deplorable state of affairs was the fact that the last volume of the official history of British airpower in World War I (The War in the Air), which clearly showed the importance of aircraft in commerce protection, only came out in 1937, much too late to influence policy. Similar results to the First War were obtained in the Second War once suitable aircraft were made available for use as convoy escorts and supports, but this was not until 1943. It can be argued, and has indeed been vociferously argued by the Navy ever since, that the RAF’s obsession with the ‘wasteful and largely discredited’ policy of bombing Germany indiscriminately deprived the Fleet of the aircraft required for convoy and other sea work, while achieving no significant reduction in Germany’s war potential. The issue is not of a black-and-white sort, however. The bomber offensive, the airmen have replied, was not always what it should have been (this was the first real air war and much had to be learned), yet, in the words of the Official Air Historians, ‘both cumulatively in largely indirect ways and eventually in a more immediate and direct manner, strategic bombing and, also in other roles strategic bombers, made a contribution to victory which was decisive’.
Valuable experience of 1914–18 was disregarded in other respects as concerns convoy. Until 1943, when Professor P. M. S. Blackett produced some interesting statistics about ocean convoys and changed the staff view on convoy escort, it was Admiralty gospel that ‘the larger the convoy the greater the risk’. Had the convoy statistics of 1917–18 been analysed after the war, and the printed results of the mathematical research on comparative escort strength by an acting commander, RNVR (Rollo Appleyard) early in 1918 been studied, the Admiralty would have been aware of ‘the law of convoy size’: ‘The escort strength requires to be measured, not in terms of the number of vessels in convoy, but in terms of the total area comprised within the boundary formed by lines connecting all outer vessels.’ Appleyard went on to prove mathematically that the ratio of the torpedo attack area around the convoy perimeter to the number of escorts directly watching it is ‘a more correct numerical measure of the escort strength of a convoy than is the ratio of the number of ships in convoy to the number of close escorts’. It is sad that operational research was not understood in the interwar years; it needed someone of the standing of Blackett to show what could be done in this field.
Another instance of how the postwar failure to study with care the U-boat campaign of 1917–18 exacted a heavy penalty was the refusal of the Admiralty in the interwar period to believe the U-boats would make surface night attacks. Although by the end of the First War nearly two-thirds of all submarine attacks were being made at night and on the surface–to be sure, they proved unrewarding–the Second War found the Navy unprepared for a repetition of these tactics, this time successfully. The evidence was available, but it took the Admiralty a year (August 1940) to realize that the majority of the ships sunk by U-boats since the start of the war had been sunk at night–by, of course, surfaced U-boats. When, in 1940, the U-boats in the Atlantic, organized in ‘wolf packs’, attacked convoys at night while on the surface, the Admiralty had no immediate answer. It was, as a joint Admiralty-Air Ministry statement of 1946 misleadingly claimed, ‘a new and unheard of German tactic’. The problem was not mastered until 10-centimetre radar was fitted generally to convoy escorts. The turning of night into day with ‘snowflake’ flares and other pyrotechnics also played an important role in the defeat of surface attacks. There is no excuse for the Admiralty not having learned by 1939 that U-boats might attack on the surface at night. Whether the use of ‘wolf packs’ could have been foreseen from a study of the First War is another matter.