Notes

Introduction

1. After the First World War, Admiralty and fleet papers were collected to form the basis for the official history. They form the ADM 137 series (about 5000 entries) in the Public Record Office (The National Archive of the UK), most but not all of which are British official documents. A few are German and even fewer are Austrian, often in the original language. Most of the British official histories, for example of minesweeping, are in the ADM 186 series (formerly Confidential Books); it also includes many handbooks of various types. Other First World War records are in the ADM 1 (Admiralty dockets) and ADM 116 (Admiralty cases, meaning collected papers) series and a few are outside these series. Other papers are held by the Naval Historical Branch in Portsmouth and the Covers in the National Maritime Museum Brass Foundry outstation at Woolwich Arsenal include many documents referring to policy decisions. In some cases US papers in the National Archives include important British material.

2. The ADM 167 series. Pre-1917 books exist, but they are useless except as rosters of the Board.

3. When Arthur J Marder, the pre-eminent historian of the First World War Royal Navy, arrived in London to begin research on what became his history of the pre-war Royal Navy, The Anatomy of British Seapower, the Admiralty still retained all of its historical files and only those from 1885 and before had been opened under the existing fifty-year rule (later relaxed to thirty years and recently to twenty). Marder’s considerable achievement was to convince the Admiralty Librarian to allow him access to documents, apparently mainly Admiralty printed papers, as recent as 1904. He did not realise that access had been strictly controlled. The relevant file (ADM 176/207 – an Admiralty case held outside the usual ADM 116 series) was released a few years ago. In November 1935 Marder’s request for access to post-1885 papers was rejected altogether on the ground that the Admiralty would reap no special advantage by releasing them to him. Marder later provided a list of specific subjects. In September 1938 he was specifically denied access to his Subjects 4 (commerce protection in the event of war), 10 (naval aspects of the Anglo-Japanese Alliance of 1902), 14 (naval war plans, particularly blockade policy, against France and Russia) and 21 (Admiralty concern over the Balearics – Port Mahon), the implication being that more than forty years later the Admiralty still considered them sensitive. The Anglo-Japanese issue was clearly sensitive because the Admiralty considered Japan the likeliest enemy and the Balearics figured in the ongoing Spanish Civil War. Trade protection was interesting both because so many writers (wrongly) assumed that the Admiralty had not been interested in it and because the cost of the cruisers needed for trade protection was a driving force in Admiralty and therefore British government policy around and after the turn of the twentieth century, probably even the reason for the alliance with Japan. Because he was denied access to the mass of papers on the subject (now mostly at the Naval Historical Branch), Marder accepted published and personal claims that the Admiralty had neglected trade protection in the decades before 1914, which was emphatically not the case. The blockade aspect of naval war plans was also sensitive, because it involved the changing role of underwater weapons, as understood at the turn of the twentieth century. The Admiralty put Marder off by citing the sheer volume of papers that would have to be sifted in order to meet his requirements. Later in 1938 the Admiralty Librarian was specifically told not to show Marder any papers concerning policy (few now seem to remain). They included the First Sea Lord’s Minute of 15 June 1892 (not reproduced in the file), which presumably referred to the effort to build support for the 1893 supplement to the 1889 Naval Defence Act. In reviewing Marder’s manuscript in 1940, the Admiralty was particularly anxious that Marder not write that he had had access to most of the relevant papers for 1885–1905. What is interesting in this story is just what the Admiralty considered sensitive.

4. Marder relied heavily on papers left by many key officers (and reminiscences) as a substitute for closed official papers. His references to Admiralty manuscripts are often to the answers generated to specific questions by the staff of the Naval Historical Branch. Because no one on that staff had already done extensive research on the topics which interested Marder, no one could have suggested further questions, even had they been allowed to do so (apparently they were not so encouraged).

5. Krieg zur See is very nearly without footnotes and there is no way to be sure that its representations of either pre-war plans or wartime thinking are unbiased and accurate. It was published as German nationalism (ultimately Nazism) grew and some of its assessments may well have been shaped to make wartime naval leaders with preferred post-war politics look aggressive and therefore attractive. Thus the account of Jutland is used to justify Tirpitz’ policies and Scheer is clearly the only heroic High Seas Fleet commander. The seven North Sea volumes provide an excruciatingly detailed account of all (even minor) naval operations in the North Sea, in most cases correlating German records with the initial volumes of the British official history. The full series amounts to seven North Sea (Nordsee) volumes (six of them translated by the Admiralty, carrying the story to the spring of 1917), three Baltic volumes (of which the two translated by the Admiralty complete the story only to 1915; the third volume carries the story to the end of the war), three cruiser warfare volumes (the first written in 1928 by Captain Erich Raeder, the future Kriegsmarine commander-in-chief), two volumes on the war in Turkish waters (the first on the Mediterranean Division, the second on the Dardanelles and related matters), one volume on the East Asian cruiser squadron (commanded by Graf Spee), one volume on the war in the German colonies (in two parts), five volumes on the U-boat war against trade (of which Volumes 4 and 5, which covered 1917–18, were not published until after the Second World War) and a volume on wartime surface warships. This combination is far more massive than the unclassified British official history, but to some extent it combines the roles of published official history and staff history. It also incorporates considerable editorial comment on command decisions and on tactics. I have used the translations of volumes of Krieg zur See produced by the Admiralty (except as noted, page number references are to those volumes, not to the German originals). These volumes are held by NHB and most of the translations were published in the inter-war issues of the navy’s Monthly Intelligence Report held by NHB and, except for a few years, PRO. Other navies also translated the German history; the US Navy Library holds the Italian versions of some volumes, as well as some English translations produced by ONI. The final volumes of the North Sea, Baltic and U-boat series were not published until the 1960s and therefore they were not translated by the Admiralty (reportedly copies published during the Second World War were not distributed and were mostly destroyed). The US Navy Library holds the final North Sea and Baltic volumes, but not the two final U-boat volumes.

Chapter 1. A Maritime War

1. For an excellent account of the interaction between banks, customer and ship designers, see David Topliss, ‘The Brazilian Dreadnoughts 1904–1914’ in Warship International No 3 (1988). Dr Topliss took his degree in history after studying economics. At the time of his early death, Dr Topliss, who was then in charge of the Brass Foundry, was working on a book about the warship export market, the 1988 article being his main legacy.

2. For this writer, the great symbol of small government is to be found in surviving films of Britain during the 1914 crisis: the Prime Minister arrives at a meeting in a taxi (he is seen paying the driver).

3. As an example of the way capital markets think, about 2000 the author participated in a US Navy-sponsored discussion of globalisation and its consequences. Someone thought to ask a major Wall Street firm about its vision of the future, including the prospect of various wars. After all, Wall Street put its money where its mouth was – as the City did before (and after) 1914. The view on Wall Street was that the consequences of even a small war were so bad economically that war was nearly obsolete.

4. The best-known is Norman Angell’s The Great Illusion, first published in the United Kingdom in November 1909 as Europe’s Optical Illusion and then reprinted as The Great Illusion in 1910. By 1914 it had sold two million copies. The Daily Mail called it ‘the most discussed book in years’. Angell later said that he got the idea of the prohibitive financial consequences of war from The Meaning of Money, by Hatley Withers, the financial editor of The Times. Angell received the Nobel Peace Prize in 1933. In his book on First World War economic warfare, Planning for Armageddon: British Economic Warfare and the First World War (Harvard University Press, 2012), Nicholas A Lambert cites many other such prophets. For example, in April 1910 secretary of the Liverpool Stock Exchange Edgar Crammond concluded that the catastrophic effect of war would guarantee the peace. The President of the Institute of Bankers (a director of the Bank of England) pointed out that unfortunately the decision for war does not rest with the mercantile community. This paper was brought to the attention of the Prime Minister and of the CID. Perhaps the earliest such book was Ivan Bloch’s massive analysis of the changing character of war, initially published in French in 1898 as La Guerre Future. Bloch, a Polish banker and railway financier, was a Russian delegate to the 1899 Hague Peace Conference, where he met Admiral Fisher. Six months later an abridged version of his book appeared as Is War Now Impossible?, with a preface by W T Stead, the journalist who had worked with Fisher to put the Royal Navy’s case for modernisation in 1884–5. Bloch’s work, which predicted the sustained mass warfare of 1914–18, was widely read in Europe, but, as might be expected, it had no impact on European armies. Britain was anomalous, as its army was the weakest in Europe and had the least influence on government policy.

5. Lord Hankey, The Supreme Command 1914-18 (London: George Allen & Unwin, 1961), I, p 136.

6. Lambert, Planning for Armageddon, argues that initially two of the six divisions earmarked for the British Expeditionary Force were kept home due to fear of instability, but it had also been argued that this was due to fear of invasion, the Territorials being untrained.

7. The only major departure from the navy’s belief in blockade seems to have come during the short tenure of Admiral A K Wilson as First Sea Lord in 1910–11. Wilson argued that the Germans could use neutral ports to evade any blockade. That can be taken as a reference to an emerging Admiralty position that a wider economic attack, which would deny British ships to German cargoes, might well be effective.

8. Lambert, Planning for Armageddon, describes a drawn-out fight over the validity of the navy’s statistics showing German vulnerability.

9. As understood by the British about 1911, Rotterdam was second only to Hamburg as a centre of German trade. About 20 per cent of their seaborne trade passed up the Rhine through Rotterdam, including two-thirds of iron ore imports and half of imported grain. No other neutral port had the necessary infrastructure. The Admiralty argued that Rotterdam and Antwerp were so tightly integrated into the German economy that they should be considered German ports. Admiral Bethell told the CID subcommittee in 1911 that the Admiralty would grasp the slightest pretext to close these ports to transoceanic traffic. By this time the War Office (army) understood that the Germans would almost certainly attack France through Belgium, leaving only Rotterdam. The Admiralty would have liked to blockade Rotterdam at the outset of a war, but the Foreign Office argued successfully that it was in British interest to keep Holland neutral. That was so much the case that the British found themselves running convoys between Holland and England in 1916.

10. Fisher’s insistence that war plans were secret was tied up to some extent with both his assumption of operational (rather than strategic) control as First Sea Lord using W/T and also his feud with Admiral Charles Beresford. When Beresford became C-in-C Channel Fleet, Fisher provided him with war orders but not with war plans. Beresford protested that he needed something more in order to train his force; Fisher replied that he would find out what he needed to know when he needed it. British war planning documents for fleet commanders exist prior to 1906 and after 1912, but the only documents produced between those dates were Fisher’s elaborate printed version of alternatives. There is some question as to how seriously most of them should be taken, as they were partly a means of beating back Beresford and his supporters.

11. Austria-Hungary provides perhaps the most painful case in point. In 1913 the Austrians discovered that their chief war planner, Colonel Redl, had been a Russian spy for years. Even though they knew that the Russians had their war plan, they could not change it quickly enough to have any alternative available in 1914. Replanning railway movements was just too complex.

12. Churchill’s comments on proposed British army deployments (he wanted the army to deploy to Belgium so that it would be on the Germans’ flanks) survive as ADM 116/3474, ‘1911-Crisis: Miscellaneous Memoranda and Correspondence Between Cabinet Ministers’. These documents show, incidentally, that as of 1911 the British were aware that the Germans were likely to deploy through Belgium; General Wilson thought that would force the British to declare war.

13. Roy Jenkins, Churchill (London: Macmillan, 2001) writes that Asquith came to realise that Churchill was not well suited to the Home Office. The incidents were the ‘battle of Sidney Street’ after a serious outbreak at Tonypandy in the Rhondda, both being followed by over-enthusiastic reaction to a Merseyside strike. Jenkins wrote that many regarded Churchill as too trigger-happy after Sidney Street, but that Tonypandy unfairly caused intense resentment in the Labour Party, affecting events in 1940. When offered promotion in February 1910, Churchill had asked for either the Admiralty or the Home Office, in that order, and was given the Home Office at an unusually young age.

14. A J P Taylor, The Struggle for Mastery in Europe 1848–1918 (Oxford University Press, 1954), p 67, points out the particular hostility of the Junkers in eastern Prussia – the basis of the Prussian army officer corps and of much of its administration – to German nationalism, meaning to the idea that Germany was defined by a popular movement, rather than from above by the Kaiser.

15. V R Berghahn, Germany and the Approach of War (New York: St Martin’s Press, second edition, 1993), p 22. Bismarck relied on assumed powers during the ‘gap’ after the diet was dissolved, the idea being that he would seek retroactive approval for spending.

16. Berghahn, Germany and the Approach of War, p 27, cites a 1907 German General Staff study on ‘fighting in insurgent towns’ and studies by some local commanders of how they would deal with insurgency, including plans to arrest Reichstag deputies and others considered subversive. Prussia (and later the Empire) had an 1851 State of Siege Act under which local military commanders were empowered to take over in the event of internal disorder as well as war. They and their superiors were empowered to decide when such intervention was appropriate. If a state of siege were announced, all civil liberties would be suspended, censorship enforced and all of those considered subversive arrested.

17. For an account of this ‘Zabern affair’ see Jack Beatty, The Lost History of 1914: How the Great War Was Not Inevitable (London: Bloomsbury, 2012), pp 20–36. In this case the lost possibility was that, had the Reichstag pushed further, for example rejecting that year’s budget, the Kaiser and the General Staff might well have resorted to the often-discussed Staatstreich. In that case there would have been no point in triggering a war to win over German voters.

18. This point is repeatedly made by Berghahn in his Germany and the Approach of War in 1914. There was considerable sentiment in the army and its general staff for a preemptive war against France in 1905–6, on the ground that defeat in the Russo-Japanese War and the subsequent 1905 revolution had so weakened the Russians that they could not effectively support the French. To that end the 1905–6 crisis in Morocco was precipitated specifically to induce the French to declare war; but with British support the French managed to avoid this trap.

19. Fritz Fischer, War of Illusions: German Policies from 1911 to 1914 (New York: Norton, 1975), p 463, as seen by Dr Wilhelm Muelon, a member of the Krupp board. Fischer cites a number of German claims that the war was preventive, that the Entente planned to attack at some point in the next two years. That having been disproven, Fischer explores other possible German motives. However, if the point of prevention was internal German politics, the claims of preventive war make sense.

20. Even in 1870, the Germans almost found themselves fighting a drawn-out war when the Commune in Paris refused to comply with the surrender of the French field army. The Germans had to besiege Paris. Their task was simplified in that the Paris Commune was trying to promote a revolution against the French state which had surrendered, hence was also opposed by the post-surrender French army (with results which embittered post-1870 French politics).

21. The reader may object that in 1940, fighting a massive French army, the Germans did exactly that: they triumphed rapidly. Had the British not remained in the war, it would have been over in June 1940. The difference was that in 1914 the Germans could not move fast enough to envelop the French army, which was moving on interior lines. Nor did they have the command and control advantages which made it possible for them to outpace French command and control decisively in 1940.

22. ADM 116/3474, note by Director of Military Operations General Henry Wilson, who had presented the army’s war plan at the 1911 meeting.

23. Berghahn, Germany and the Approach of War, p 38, describes the domestic factors involved, assuming that they were responsible. He points out that the big industrialists were particularly receptive because in 1895–6 they were just recovering from a world depression which had begun in 1873 and they did not want to suffer similarly again. Tirpitz’ guaranteed future building programme was extremely attractive to them. Krupp founded and financed the German Navy League. Berghahn cites numerous comments by Tirpitz explaining how his projected fleet could defeat the Royal Navy and thus be a threat which could force the British to offer Germany ‘fair treatment’ (as opposed to high tariffs). However, these papers were for the Kaiser, who was disposed to dislike the British.‘

24. Patrick J Kelly, Tirpitz and the Imperial German Navy (Bloomington: Indiana State University, 2011), pp 108–10. The document was initially issued on 28 November 1895. Tirpitz was then asked officially to comment on it.

25. Kelly, Tirpitz and the Imperial German Navy, pp 131–2.

26. For this campaign, see Ruddock F Mackay, Fisher of Kilverstone (Oxford: Clarendon Press, 1973), pp 177–82.

27. Translated versions are in the Admiralty Monthly Intelligence Report, in PRO and in NHB, beginning with No. 12, 15 April 1920. The series begins in August 1914.

28. As described in a letter from a US Navy officer assigned to the Allied Control Commission, copy in OpNav Undersea Warfare Division records.

29. Translated extracts from ‘Why the German Revolutionary Movement Emanated from the Fleet’, in Monthly Intelligence Review No. 7, 15 November 1919, ADM 223/807.

30. After the war Admiral von Hollweg described the cruise as grossly irresponsible in a review of Admiral Scheer’s memoirs translated for Monthly Intelligence Reports in 1919.

31. In 1964, discussing whether to fight North Vietnam after a Viet Cong attack on some Americans in South Vietnam, a US official remarked that crises were like trolley cars; if you missed one, another would soon come along (this particular one was missed). It is difficult not to see 1914 in a similar light.

32. In 1870 Prussian Chancellor Bismarck manoeuvred French Emperor Napoleon III into declaring war on him, so that in theory he was the offended party, even though it is generally accepted that he was also the architect of that war.

33. In particular the French lacked heavy artillery, although in their ‘75’ they had an excellent rapid-fire medium gun.

34. The French Foreign Ministry was reading German diplomatic codes at this time and the British had sufficient contacts in Constantinople for their Foreign Office to be aware of the developing German intrigue with the Turks. Neither considered this information relevant to naval operations, so it was never communicated to the navies. A French researcher at their Higher Naval School discovered this lost opportunity in the 1920s. The lesson, which seems to have been learned both in the United Kingdom and in France, was that naval staffs should receive diplomatic information as well as purely naval data, such as ship positions. In the British case, the wartime naval code-breakers handled diplomatic as well as naval traffic and later code-breaking was handled on an all-service (including diplomatic) basis. I am grateful to Alexandre Sheldon-Dupleix, of the naval staff of the Service Historique de Défense, for this insight.

35. Sean McMeekin, The Russian Origins of the First World War (Cambridge: Harvard University Press [Belknap], 2011), pp 25–9.

36. Sean McMeekin, The Berlin-Baghdad Express: The Ottoman Empire and German’s Bid for World Power (Cambridge, Mass.: Harvard University Press, 2010)

37. The Turkish Caliph was, at least in theory, the spiritual chief of all Sunni Muslims. In 1914 the British Empire had more Muslim subjects than any other country.

38. The context was a proposal, mooted during a Cabinet inquiry into the army budget, that the army be reshaped as a small expeditionary force to be deployed by the navy. The army certainly became interested in expeditionary operations, but on a much larger scale.

39. Fisher’s friend and biographer Admiral R H Bacon described a crucial 18 November 1909 CID meeting. In the course of a crisis over Morocco, the French government had demanded that the British deploy 120,000 troops to the French border to deter the Germans. Fisher confirmed that he could provide the necessary transport. When pressed for further comment, he said that he had ‘nothing else to say that anyone present would want to hear’. When Prime Minister Asquith pressed him, he said that the Germans would stop at nothing to destroy that force and would do so, ‘Continental armies being what they are’. The army should be restricted to sudden descents on the coast, the recovery of Heligoland and the garrisoning of Antwerp. Fisher then pointed to the ten mile stretch on the Pomeranian coast near Berlin. Were the army to seize and entrench there, a million Germans would find occupation; ‘but to despatch British troops to the front in a Continental war would be an act of suicidal idiocy arising from the distorted view of war produced by Mr. Haldane’s speeches and childish arrangements for training Terriers [the Territorial Army]. After war broke out . . . the British Army should be administered as an annex to the Navy and that the present follies should be abandoned’. Asquith adjourned the meeting but presumably remembered Fisher’s comment. Admiral Sir R H Bacon, The Life of Lord Fisher of Kilverstone (London: Hodder and Stoughton, 1929), II, pp 182–3.

40. The British were aware of German thinking. In commenting for Winston Churchill on a note by Times military correspondent C A Repington, DNI Captain G A Ballard wrote on 20 April 1912 that ‘there is evidence to show that it is quite possible that it is the intention of the Germans to act at first on the defensive, in the hope that we shall expose our ships in close blockading operations which will afford them opportunities for torpedo attacks, until they can reduce us to something approaching equality. Their elaborate and expensive system of coast defence is a fairly clear indication of the ideas which find acceptance in responsible quarters, mistaken though they may be. And even their latest destroyers are of small displacement and radius of action as compared to our own and correspondingly less fitted for offensive operations’. Repington had correctly guessed British policy as to close or open blockade; ‘it is of the utmost importance not to let our general policy as regards blockade become generally known, or even known outside the War Staff’. Repington had asked the ‘assistance’ of Naval Opinion; in Ballard’s view it was appropriate that ‘a knowledge of Admiralty policy on the major points of the subject is confined to a very few officers within the Admiralty and to a small number of selected Admirals’. Churchill Companion Pt 3, 1534-1535 (Repington’s paper is not reproduced).

41. Field Marshal Lord Roberts was the strongest advocate of this view. Randolph S Churchill, Winston S. Churchill II, Companion Pt 3 1911-1914 (London: Heinemann, 1969) includes several of Roberts’ 1912 letters. In a 25 January 1912 letter (pp 1500–1) Roberts argued that the usual figure of a possible 70,000-man German attack was purely hypothetical, arising from his own statement that 70,000 was the least the Germans would use to invade the United Kingdom. He argued that the Germans would surely land many more men. He also argued that the projected six-division Expeditionary Force was ludicrously small in view of the size of Continental armies.

42. Thus a 12 August 1914 telegram from First Lord Winston Churchill to Jellicoe, warning him that the ‘silence and inertia’ of the High Seas Fleet (which really was silence and gross inertia) might indicate that a serious enterprise was afoot; Churchill enjoined Jellicoe to conduct cruiser sweeps to detect the Germans should they come out for a landing. A Temple Patterson (ed), The Jellicoe Papers Vol I (Navy Records Society, 1964), p 49.

43. Patterson (ed), Jellicoe Papers, I, pp 86–8, dated 12 November 1914. Churchill wrote that the shipping at Hamburg would suffice for a quarter of a million men, ‘and there is no doubt that their shipping has been specially fitted with all appliances necessary for the transport and speedy embarkation of an army’.

44. Comment in the Admiralty’s Monthly Intelligence Review when the translation of the German account of this operation was published.

45. This may have been the first British wartime use of signals intelligence. According to Krieg zur See, Nordsee 2, during the morning of 31 August call-signs were intercepted which indicated that the battleships Pommern and Brauenschweig were at sea, hence that evacuation had to be hurried. This turned out to be a false alarm.

46. According to Krieg zur See, Nordsee 2, on 16 September the Admiralstab wired C-in-C High Seas Fleet that the British were transporting a large force to Ostend and that it was of the utmost importance to the army to hinder this operation; a submarine should be ordered to attack ‘in spite of difficult conditions of navigation’. Commander Southern Force (RN) was aware of this threat and he planned a destroyer screen to operate with the armoured cruisers on patrol (Aboukir, Hogue and Cressy) to cover the transports. Initially that was impossible due to the short steep sea set by a gale, which it was thought would also make submarine operation impossible. The cruisers had recently moved position and there was reason to think that the enemy would be concentrating further north. Admiral Tyrwhitt soon took a destroyer force to sea. Three U-boats were deployed, of which U 9 (Lieutenant Commander Otto Weddigen) managed to get to her patrol area despite the weather. When the wind died down he found himself within close range of the cruisers, which he sank. The destroyers were still in port due to the gale just ended. This was not quite the example of mindless ignorance of the submarine threat which is often described; it was more an unlucky example of underestimation of the submarine’s ability to ride out weather which would keep a destroyer in port. According to the German account, U 9 did not have an easy time. Violent motion knocked out her gyro-compass, forcing her commander to fall back on navigating by soundings and then landfalls (the compass was restarted and recovered as the boat was able to steam with the sea). Even then the weather was bad: lying on the bottom in 13½ fathoms (81ft) she still felt the motion of the sea and had to surface and come into the wind, fighting the sea and exhausting the crew. She first saw ships when it was still dark, with Force 5 winds; she had to dive when they were within 1000 yds. She then proceeded at 8 fathom depth, apparently not using her periscope, until dawn. When she then surfaced, the cruisers were in sight (they were probably the ships seen the night before) and U 9 attacked them in sequence. She fired five torpedoes:. First she fired one bow tube at one of the cruisers, then (35 minutes later) both bow tubes (having reloaded No. 2) at a second, then both stern tubes (25 minutes later) at the third (torpedoes set to turn 180°). Having used up his torpedoes, Weddigen had none left for any transports he might find, so he turned for home. For the British the main impact of Weddigen’s attack was that slow heavy ships should no longer be employed in the North Sea and the Channel.

47. Krieg zur See, Nordsee 2 makes this point strongly: the Grand Fleet moved south in response to the threat to the transports, but no one at the top of the Imperial Navy seems to have understood what that could mean. The German history argues that this did not indicate a lack of will to fight under favourable conditions on the part of either the Fleet Commander (von Ingenohl) or the Admiralstab itself. It states that, under the influence of Chief of the Naval Cabinet Admiral von Müller Admiralstab chief Admiral von Pohl (who would replace von Ingenohl) adhered to the Chancellor’s view that the fleet had to be preserved as a bargaining chip at the end of the war.

48. This idea died when the German General Staff refused to dilute its planned attack on France sufficiently to provide the forces for Denmark.

49. According to Krieg zur See, Nordsee 1, the root idea of all German naval war planning before the war was an attack by the British, with Germany on the defensive. That caused difficulties, because as the 1914 crisis deepened the navy was told that Germany might be fighting only France and Russia, with the British attitude uncertain. The Admiralstab concluded that even then forces had to be concentrated in the West, its chief task the protection of overseas shipping, which was essential to restore economic life from the destruction of a war on land. This is less than credible. The history quotes chief of the Admiralstab Admiral von Pohl, writing on 25 July: if Great Britain remained neutral, in spite of important duties in the North Sea, Russia should be finished off first. That makes sense; the Admiralstab argued that it would be essential to keep the British out of the war and to do that anything that would irritate the British should be avoided (Admiralty translation in NHB, p 20). Only in the Baltic could an appreciable success be achieved and insufficient forces there might lead to failure. At this time the planned concentration in the North Sea left only a weak force of older ships in the Baltic. Other senior naval officers disagreed with von Pohl. In theory the completion of the Kiel Canal made it possible to transfer ships quickly between the Baltic and the North Sea, but no large battleships had yet made this passage (with a reduced coal load, the Kaiserin made the passage on 25 July). According to Krieg zur See, once the British had mobilised their fleet, the proposed concentration in the Baltic had to be cancelled in favour of the earlier planned concentration of German ships in the North Sea, as nothing less would have been prudent. The German fleet adopted its planned war organisation.

50. For example, in 1912 the British naval attaché reported in May 1912 the Danes, who seemed well informed about German thinking, said that in wartime the British fleet would be concentrated off Norderney. Report in Backhouse Papers, NHB.

51. Krieg zur See, Nordsee 1, p 75 (Admiralty translation, NHB).

52. Kelly, Tirpitz and the Imperial German Navy, p 289 mentions a pamphlet published in the autumn of 1907 by retired Vice Admiral Carl Galster advocating a cruiser and U-boat programme (for commerce warfare and Kleinkrieg) instead of the Tirpitz battleship programme. He was willing to accept the need for a battle fleet, but considered that against an opponent as strong as England only Kleinkrieg was possible. Only the Socialists were interested and at this time they were in a distinct minority.

53. The great gap was ocean surveillance: without it, no kleinkrieg force could easily find and attack enemy warships. One of the major surprises of the British 1913 Manoeuvres was that there were none of the expected night destroyer attacks, because the destroyers never found any targets at night. Even a narrow sea can be quite wide. Before 1914 no one seems to have realised just how difficult – and how vital – ocean surveillance could be.

54. Krieg zur See, Nordsee 1, p 89.

55. Krieg zur See, Nordsee 1, p 90. The author argues that, had the Admiralstab looked clearly at the implications of distant blockade, it could have guessed British policy: (i) attack Germany economically by suppressing foreign trade and communication; (ii) protect troop transports; (iii) bring the German fleet to action if circumstances are favourable. The reference to the Heligoland Bight obsession is the German official author’s. He assumes, incorrectly, that the British treated the possibility of invasion similarly. The author also points out that Tirpitz and his Navy Office had failed to build light cruisers and destroyers with sufficient range to break a distant blockade. A comment on the failure of the High Seas Fleet to take offensive action in the summer of 1914 (Nordsee 1, p 130) credits its destroyers with an endurance of two days and two nights; assuming a cruising speed of 15 knots and allowing 30 per cent of fuel for use in action, they could steam out only 250nm from their fleet base. The official history (p 92) quotes a January 1918 memo by Admiral von Ingenohl, High Seas Fleet commander in 1914: ‘before the war the whole training of our Fleet, our tactics, manoeuvres and to some extent even our shipbuilding policy and even certain constructional details (for instance, the small radius of action of a large number of our destroyers) were based on the hypothesis that a decisive action would take place in, or very close to, the German Bight, on the assumption that the British would organise a near blockade of the German Bight with their superior Fleet, in order to cut off our overseas supplies and at the same time force our Fleet to take up the strictest defensive and to accept action. It is true that in the years immediately preceding the war there arose now and again doubts and hesitations prompted by sundry expressions of opinion in articles written by British experts; it was asked whether the enemy would still adhere to this procedure or if he would perhaps content himself with the blocking of the Channel and the line Scotland-Norway. However, these doubts were always dissipated by the observation that a blockade of that kind would never be recognised by Neutral States, the USA in particular (italics in Krieg zur See translation)... even if the British did endeavour at first to get on without a blockade, they would soon find that they could neither cripple the activities of our Fleet, particularly of our minelayers, etc., nor cut off our overseas supplies without a near blockade of our North Sea coast. In our manoeuvres in May 1914, Vice Admiral Scheer, the Senior Officer of the side representing the enemy, wrote in his report that, in his opinion, British prestige would not permit of their foregoing the blockade or closer patrol of the German Bight’.

56. As quoted in Krieg zur See, Nordsee 1, from von Ingenohl’s memo of January 1918, previously cited (but quoted in a later part of the book). The official report attached less importance to submarines. In the peacetime manoeuvre, U-boats were forced to show their periscopes frequently in order both to manoeuvre and to attack and particularly to avoid accidents. In wartime U-boat commanders would feel constrained to use their periscopes much less often. Moreover, ships would be alert, so they would more often evade torpedoes. Ships in manoeuvres also suffered from limitations on the speed and course of ships and escorts could remain closer than in peacetime exercises. It would be difficult for a U-boat to hit a ship and her escorts zig-zagging at high speed and it would be extremely difficult for a submarine to penetrate a screen unobserved to get into position to attack the main body.

57. Krieg zur See, Nordsee 1 frequently refers to destroyer screens against submarines, but it is not clear what weapons were envisaged. The descriptions of the two August 1914 sorties include the use of guns against submarines which briefly appeared, to little avail. Krieg zur See added that in 1914 nothing was known of anti-submarine weapons: depth charges (the Germans adopted the Austrian one in 1915), anti-submarine mines, nets and means of detecting submerged submarines.

58. According to Krieg zur See, Nordsee 1, German war diaries revealed that the movements of heavy ships were strongly influenced by fear of submarines. The German history claims much the same effect on British movements. The first report of a British submarine in the Heligoland Bight came a few hours after war broke out. Light cruisers on patrol were immediately withdrawn, the destroyer defence was reinforced and the Senior Officer of Scouting Forces was ordered to have the German (Heligoland) Bight swept by the Mine-Seeking Divisions using their sweep gear. Nothing came of these sweeps, but British submarines were certainly seen further offshore, beginning with one sighted by a fishing boat about 100nm WNW of Heligoland on 5 August. The light cruiser Hela was torpedoed on 13 September 1914 in an area the fleet had used for an exercise the previous day.

59. Krieg zur See, Nordsee 1, p 79. Krieg zur See, Nordsee 1, p 78 emphasises the difference between the order to keep the fleet home while executing kleinkrieg and the Admiralstab’s insistence that kleinkrieg would be preparatory to fighting a fleet action on better terms. It criticises the staff for not laying sufficient emphasis on the basic principle of employing the fleet, particularly its heavy ships, more actively to even the odds. Given the political imperative to treat the battle fleet as a ‘fleet in being’, it was nearly impossible for any fleet commander to do otherwise. Krieg zur See generally ignores the fragmented character of the German command system, the Admiralstab being advisory only.

60. The German estimate of British intentions is given in Krieg zur See, Nordsee 1, pp 81–9.

61. Krieg zur See, Nordsee 1, p 123. The official history shows the claim the German General Staff wanted the transports spared so that the British troops could be dealt with on land is nonsense.

62. Krieg zur See, Nordsee I. Before the submarine operation failed, the fleet command thought that the transport of troops to France would occupy the British 2nd and 3rd Fleets and also some of the light forces of the 1st Fleet (Home Fleet), perhaps creating the desired opportunity to bring the Home Fleet to action on even terms. This idea was then abandoned because reports of enemy operations and intentions were clearly unreliable (Nordsee I, p 129).

63. Krieg zur See, Nordsee 1 argued that the scale of defence needlessly exhausted ships which should have been kept ready for a fleet sortie. On 5 August the defence comprised outer and inner picket lines, each held by one destroyer flotilla (twenty ships), radii being 35 and 25nm around Elbe Lightship I. Submarines operated between the two lines, at a radius of 29nm. One light cruiser was stationed behind each of the right and left flanks of the defence, with a third off Heligoland harbour. In addition, a cruiser and a flotilla were at instant readiness in the Weser and three flotillas were in the estuary. Inside the lines, the anti-submarine force was initially an additional destroyer flotilla, soon relieved by a minesweeping division (once the 3rd Minesweeping Division at Cuxhaven was ready. Until September, there was no mine defence, as it was thought that would hamper German forces in the expected decisive surface battle in this area. Given the strain picket duty imposed, the submarine line was almost immediately abandoned, most of the boats being released for offensive duty (some were retained in Heligoland, but were not to leave harbour until an enemy had been sighted).

64. Krieg zur See explains the smaller calibre of German cruiser guns on the basis of the fleet’s preference for larger numbers; between 1909 and 1912 the fleet demanded that a light cruiser be able to repel a simultaneous night attack by two destroyers on each side. Meanwhile there was pressure to replace the 4.1in (10.5cm) gun with the 5.9in (15cm). In the summer of 1911 that was deferred to the summer of 1912, when it was adopted. The Germans also expected that in most cases battlecruisers would be backing their light cruisers, so that their lighter main batteries would be balanced by the much heavier ones of the battlecruisers. Similar considerations applied to torpedo boats (destroyers). In a day battle the destroyers would be covered by battleship and cruiser fire as they approached their targets; their own gun calibre would not be as important as the number of torpedoes. The navy rejected the greater displacement required to mount heavier guns on the ground that it would reduce handiness; it also argued that a larger boat would not be as effective during a night attack. Krieg zur See offers no rationale for the poorer speed of the German destroyers, but does comment that Heligoland in effect demonstrated that the single-purpose German destroyer, optimized for night and massed torpedo attacks, might be unable to survive a fight with British destroyers.

65. German hits on British destroyers were ineffective. At least two German HE shells failed to explode after passing through the bows and funnel of a British destroyer. Krieg zur See was quick to add that the British had even more such cases. That was not too surprising, as shells were fuzed to explode after penetrating thicker plating.

66. According to Krieg zur See, the problem was an inability to concentrate fire on one of several approaching destroyers. It claims that German adoption of electrical bearing indicators (Richtungsanlagen) for cruisers was inspired by this battle. They permitted a gunnery officer to bring all available guns automatically onto one target. Note that although the German phrase was often translated as director control, it was something simpler.

67. As described in Krieg zur See, Nordsee 2 (the history to November 1914). Glory was already near Halifax. Now three of her sisters were sent: Albion to Gibraltar, Canopus to Cape Verde and Ocean to Queenstown to replace Illustrious, which became guardship at the new Grand Fleet anchorage, Loch Ewe. Comments in Krieg zur See suggest that the Germans did not understand the key significance of coal stocks, as the similar measure taken in the Falklands is derided. The defence of coal stocks had been a key anti-raider measure in British thinking as early as the 1880s. Obviously there were many other stocks of steaming coal, but most of them were British-owned, hence could be denied to the Germans in neutral ports. A small raider could seize colliers and coal at sea, but that was impossible for even a small fleet of larger ships.

68. ADM 137/2708, pp 301ff (PD. 065).

69. The Japanese were promised the German Pacific islands, which became the League of Nations Mandates so important in US inter-war planning.

Chapter 2. Resources

1. According to Lambert, Planning for Armageddon, the first Cabinet meeting on war finance was not held until 15 December 1914 and the first item to be considered was a Russian request for a £100 million loan (which exceeded the total British peacetime military budget and equalled one-seventh of the British national debt) for munitions and the plant to produce it. Among other requests, the Romanians wanted £12 million and threatened to go to the Germans if they did not get it.

2. Kitchener is quoted in Michael and Eleanor Brock (eds), H.H. Asquith Letters to Venetia Stanley (Oxford University Press, 1982), p 420.

3. This is Lambert’s view, in his Planning for Armageddon.

4. Nicholas Tracy, Attack on Maritime Trade (University of Toronto Press, 1991), p 129. He highlights the effect on Britain of the economic attack against Germany. The scholar he quotes is Kathleen Burk.

5. Lambert, Planning for Armageddon. He quotes a paper by Maynard Keynes which was circulated at the 28 January War Council which approved the Dardanelles operation. For Keynes, the only operation worth undertaking was to storm the Dardanelles. Lord Hankey, who was secretary to the War Council, described its discussions leading up to the Gallipoli operation in his memoir The Supreme Command 1914-18 (written before the Second World War and suppressed until then). Balfour pressed two issues, the need for wheat and the need to repair Russian finances. The Council favoured an attack on the Dardanelles because it would ‘cut the Turkish army in two; put Constantinople under our control; give us the advantage of having the Russian wheat; and enable Russia to resume exports … It would also open a passage to the Danube. It was difficult to imagine a more helpful operation’. Asquith’s letters to Venetia Stanley reveal that a Committee on Food Prices was formed on 20 January (he was chairman). According to his 22 January 1915 letter (Brock (eds), Asquith’s Letters to Venetia Stanley, p 390) ‘There is no doubt that we are at last beginning to feel the pinch of war in reference to food prices, mainly because all the German ships which used to carry food are captured or interned and the Admiralty has commandeered for transport etc over 1000 of our own. Further, the Australian crop has failed and the Russian (which is a very good one) is shut up, until we can get hold of Constantinople and open the Black Sea’.

6. In August 1914 the Grand Fleet alone had 1560 pieces of artillery, compared to 476 field guns (and no heavy siege artillery) in the BEF. Jon Tetsuro Sumida, ‘Forging the Trident: British Naval Industrial Logistics, 1914-1918’ in John A Lynn (ed), Feeding Mars: Essays on Logistics and Resource Mobilisation in Western Warfare from the Middle Ages to the Present (Boulder: Westview, 1993). Revised version courtesy of the author.

7. Operations Committee Minutes, 27 September 1917, ADM 116/3148. The Committee was initially chaired by First Lord Sir Eric Geddes and later by First Sea Lord Admiral Weymss. It seems to have been the executive committee for Admiralty materiel decisions in 1917–18. This volume effectively ends on 30 September 1918. However, ADM 137/834 Operations Committee Minutes and Memoranda, extends to the end of 1918.

8. DCNS pointed out that to provide two ‘K’ class flotillas (ten each), six ‘K’ class should be ordered, so the number of L 50s was cut to sixteen.

9. The proposal was to complete seven of the leaders under construction by the end of 1918, then one per quarter. That would retard six leaders. The rate of destroyer completions was to be reduced from 7.5 per month to four per month after 1 January 1919, provided the United States could complete and commission sixteen per month for eight months. US destroyers could take over definite and independent duties from British ships. At this time 116 destroyers were under construction and eighty would be slowed (3.5 per month). All work on three ‘M’ class submarines would be stopped at once. The ‘L’ (fifty-one) and ‘H’ (eleven) class submarines now building for completion by 1 January 1919 would be continued, the rest being slowed to two per month, with priority given the ‘L’ class. Work on ‘R’ and ‘K’ class submarines would continue as before. Notes indicated that the ‘M’ class ‘have no important operational value’. The rate of trawler completion could be reduced from 12½ to eight per month from 1 July 1919, provided the US could complete and commission 4½ per month (drifters, however, were not considered suited to US construction). Minesweeper completions would be reduced from six to three per month after 1 January 1919, provided the US could complete and commission three per month. At this time 116 were on order and fifty would be affected by retardation. An appended table compared the British destroyer force as of 31 December 1918 (total 404 ships) with the required establishment (498). That included 111 (123 required) for the Grand Fleet, fifty-seven (fifty-seven) for Dover and Harwich, twelve (twelve) minelayers, 105 (135) escorts, fifty-seven (sixty-seven) in hunting flotillas, twenty-one (fifty-seven) in local flotillas and none (six) at Halifax. Data from Appendix to 27 August 1918 Operations Committee Minutes, ADM 116/3148.

10. Requirements from Appendix to 27 August 1918 Operations Committee Minutes, ADM 137/834.

11. This was probably the ‘24 Class’ sloop.

12. On the basis of past experience, trawler casualties were not fewer than six per month (seventy-two for 1919) and eighty were obsolete or approaching obsolescence. The projected establishment as of 1 January 1919 was 1312, against a desired 1410, the increase of ninety-eight being needed to supplement currently inadequate patrol and escort forces. The required total of new trawlers was then 250. Perhaps as many as 100 would be released from minesweeping duties by the arrival of fast ‘Hunt’ class sweepers. Projected British construction in 1919 was ninety-six trawlers, so the United States was asked for fifty-four, which should if possible be on ‘fast gunboat i.e., Kil class. lines’. As with the destroyers, the US trawlers should operate in their own units. The 1918 list of requirements shows the expected numbers in hunting units as of 31 December 1918: sixty-four in Northern Patrol, sixty-four in Southern Patrol and thirty-nine in the Otranto Patrol, plus 951 in UK Home Waters, 166 in the Mediterranean and twenty-eight in other waters. Requested increases were eighty-two in Home Waters, fourteen in Mediterranean and two in other waters. The US minesweeper programme was a total of forty-five ships (three per month), for completion by 1 April 1920 or earlier. About two minesweepers were lost each month, for a total of thirty up to 1 April 1920. The estimated establishment as of 1 January 1919 was 152, the objective being 210; the British planned to match the rate of US construction. US sweepers were to take over particular Canadian and Mediterranean areas. The British accepted that they would have to be built on US lines, but hoped that they would burn coal rather than oil ‘with a view to safety when mined and in order to save oil tonnage’.

13. ADM 116/1349. This was part of a 17 January 1918 plea to keep enough men available for shipbuilding. Anson was justified on the ground that the British battlecruiser force was too weak, with seven proper battlecruisers (only three of which were equal to the German ships) and the two lightly-armoured Renowns, with Hood due to complete in 1919. The Admiralty claimed that the Germans were building fourteen light cruisers, compared to eleven British light cruisers of the 1918 programme; the eight proposed were all that could be built, as completing Anson would cut one. Destroyer and leader numbers were predicated on expected losses; the Germans were thought to be building thirty-six destroyers. Submarine numbers were required to maintain the current output. An appended table showed a current British destroyer strength of 371, with sixty-nine US destroyers and 193 German. At the end of 1918 there would be 465 British destroyers and 137 US destroyers compared to 211 German and at the end of 1919, 526 British, 144 US and 232 German destroyers. Estimated destroyer losses to the end of 1919 were thirty-seven British, ten US and twenty German. The British hoped that by 1919 the Germans would have lost 150 U-boats, leaving them with 279.

14. Data from the Admiralty’s September 1918 paper for the War Cabinet in ADM 137/834.

15. Jellicoe tried to limit capital ships to no more than four days for docking and urgent repairs, a remarkably short time compared with peacetime standards. However, frequent sustained steaming caused machinery problems which took much longer to repair. Four battleships were out of service for as long as three weeks each between August and December 1914 due to condenser problems. Boiler repairs to the two latest light cruisers took two to three months. As more ships joined the Grand Fleet, Jellicoe was able to permit three- to four-week refits. Overall, the wartime rate of refit and repair was ten times that in peacetime, seriously draining overall shipbuilding resources. Jon Tetsuro Sumida, ‘Sustaining Neptune: British Naval Operational Logistics, 1914-1918’, Journal of Military History (July 1993). Revised version courtesy of the author.

16. In mid-July 1914 the Royal Navy had 146,000 active personnel, but by mid-August, with the fleet mobilised, it had over 200,000. At the end of the war the Royal Navy had over 400,000 personnel. The Grand Fleet accounted for 52,000 men in August 1914 and nearly 76,000 at the end of the war. Sumida, ‘Sustaining Neptune’.

17. Patterson (ed), Jellicoe Papers I, p 15, quoted from his post-war autobiographical manuscript.

18. From another item in Patterson (ed), Jellicoe Papers, I, p 20, it appears that this visit was made in June 1910.

19. This was, however, a low-speed evolution, the battleship towing the destroyer and passing a hose along the tow cable. It was not the higher-speed side-by-side technique developed between wars by the US Navy.

20. The main pre-war sources were the United States and Russia. Under the deal struck with Persia (Iran), the Admiralty had access to the Persian oil fields and by 1917 they provided about 15 per cent of its supply. Sumida, ‘Supplying Neptune’.

21. Sumida, ‘Supplying Neptune’, based on an 18 July 1915 Admiralty document on requirements for Welsh coal.

22. Sumida, ‘Supplying Neptune’.

23. War Staff paper, 17 September 1917, enclosed in January 1918 ASW report to War Cabinet, ADM 116/1349.

24. ADM 116/1349. Presumably the idea was that the ships already existed. The destroyers or sloops should make at least 18 knots, with five or six days’ endurance at that speed, armed with not less than two 4in guns, four depth-charge throwers and a high-speed submarine sweep, with very good seakeeping qualities. The trawlers should make at least 11 knots, not less than 250 tons (for seakeeping), armed with one 4in gun, two depth charges and one deep sweep. Both types should burn coal. Of the sub-chasers, forty were wanted for the Mediterranean, twenty for the West Indies and twenty for the Canadian coast, all fitted for mine sweeping as well as ASW. The request seems to have been included in a Naval Staff paper dated 17 September 1917.

Chapter 3. Blockade, Trade Warfare and Economic Attack

1. A merchant anywhere in the world could accept a bill ‘on London’ in exchange for goods which would later be received by a buyer. The bank holding the bill accepted various risks, such as that the cargo would not arrive at its destination. Thus the credit system was intimately bound up with the shipping insurance industry – which was also cantered in London, at Lloyd’s. As with other systems of (in effect) paper money, the whole system depended on the confidence of all concerned. Lambert, Planning for Armageddon, p 111, quoting a 25 March 1908 RUSI paper by the leading British statistical economist of the day, Sir Robert Giffen KGB FRS. His research was inspired by observation of the effect of the US financial panic of 1907, which caused the collapse of US internal and external trade. Giffen’s immediate argument was that an outbreak of war would cause a panic and that the British government had to have reserves in place to overcome it.

2. Tracy, Attack on Maritime Trade, pp 93–4 points out that the Union ships caught only a small percentage of blockade-runners; in 1865 only 16.7 per cent of steam blockade runners but 71.1 per cent of sail blockade-runners. Profits from cotton smuggled out of the Confederacy bought 600,000 stands of arms. The Union forces connived at smuggling to New York, partly because the city’s economy depended heavily on its role as the main trans-shipment centre for cotton. Economic cost to the blockader figured prominently in the history of the First World War blockade.

3. The text of the Declaration of London shows just how slippery this definition was. ‘The question whether a blockade is effective is a matter of fact ... A blockade, to be binding, must be effective – that is to say, it must be maintained by a force sufficient really to prevent access to the enemy coastline’. Nowhere does the Declaration explain how leaky a blockade had to be to be declared ineffective. Text in Nicholas Tracy (ed), Sea Power and the Control of Trade: Belligerent Rights from the Russian War to the Beira Patrol, 1854- 1970 (London: Navy Records Society, 2005), pp 136–7.

4. Lambert, Planning for Armageddon, p 29.

5. Lambert, Planning for Armageddon, based on 1904 CID papers.

6. Lambert, Planning for Armageddon. Note that Bacon placed the conference in 1909 and claimed that Fisher used it to press his Baltic strategy. The Baltic strategy does not figure at all in Lambert’s account

7. Asquith laid out his principles in a 2 August 1914 letter to Venetia Stanley: (1) we are under no obligation to give military help to France or Russia; (2) despatch of the expeditionary force is out of the question and would serve no object; (3) we mustn’t forget the ties created by our long-standing and intimate friendship with France; (4) it is against British interests that France should be wiped out as a Great Power; (5) we cannot allow Germany to use the Channel as a hostile base; (6) we have obligations to Belgium to prevent her being utilised and absorbed by Germany. Brock (eds), Asquith Letters to Venetia Stanley, p 146. Of the principles, (4) was the classic British insistence on a balance of power on the Continent and (5) and (6) were to prevent invasion. His opposition to sending the BEF to France was partly due to the strong opposition felt by a majority in the Cabinet. Many Liberals refused even to consider what should be done with the BEF. They were shocked by the German invasion of Belgium. Asquith commented in a 4 August letter that the German invasion simplified matters for him.

8. Lambert, Planning for Armageddon, p 129. He takes the qualifier ‘by naval force alone’ to indicate that a more complete economic attack would be needed to achieve quick results. That would involve unprecedented controls on service industries, particularly banking.

9. Tracy (ed), Sea Power and the Control of Trade, pp 121ff.

10. From a précis of Hankey’s remarks by CID Secretary Sir Charles Ottley, 17 February 1911, in Tracy (ed), Sea Power and the Control of Trade, pp 154–5.

11. According to Ottley, Hankey proposed putting off the announcement until the Royal Navy had defeated the German fleet and thus had achieved such preponderance that neutrals would not feel they could object. Ottley considered Hankey’s views unrealistic. It would be a violent reversal of existing British policy. He also doubted that it would make much difference in wartime whether or not the British Government ratified the Declaration of London.

12. Distant blockade was tricky because blockade referred to specific ports. Under the Declaration of London a blockader could not bar access to neutral ports or coasts. A ship which had broken the blockade was subject to capture as long as a blockader pursued her. Announcing that ships in a wide swath of the North Sea were part of a blockade broadened that provision considerably.

13. The issue was raised in a 3 July 1911 open letter to Asquith, signed by 138 admirals; Tracy (ed), Sea Power and the Control of Trade, p 123. They argued that the navy clearly did not have enough cruisers to protect trade. To some extent the letter can also be seen as a reaction to Admiral Fisher’s dramatic change in trade protection strategy away from occupation of focal areas and towards interception of raiders.

14. September 1913 letter to Prime Minister Asquith, enclosing a 21 August 1913 memo on trade protection. Churchill Companion Pt 3, pp 1770–1.

15. Matthew S Seligmann, The Royal Navy and the German Threat 1901–1914: Admiralty Plans to Protect British Trade in a War Against Germany (Oxford University Press, 2012), p 137.

16. Admiralty Technical History (TH 13), Defensive Armament of Merchant Ships (Technical History Section, August 1919), RN Historical Branch. In January 1914 White Star strengthened the poop deck of the liner Brittanic to receive a single 6in gun and Cunard fitted some ships trading with the United States. The ships in the programme all had 4.7in guns.

17. As recounted in the first volume of the Krieg zur See series on cruiser warfare. German shipping companies and the German Home Office objected to placing ammunition on board in peacetime. Before the war the German navy considered a speed of at least 20 knots necessary for a converted liner and 17 knots a bare minimum for a ship beginning her operations outside home waters (the liner had to be faster to break out past the expected British close blockade). In 1916, however, the Germans adopted the alternative of using slower disguised cruisers not readily recognisable as warships. (e.g. Moewe and Wolf)

18. The pre-war list, by line, was White Star (to Australasia): Cufic, Runic, Tropic, Zealandic, Medi, Afric and Suevic; Royal Mail (to River Plate): Aragon, Amazon, Deseado, Desna, Asturias, Demerara, Darro, Drina and Alcantara; FHA Line (to River Plate): Rosorina, La Correntina, El Uraguayo, La Negra and El Paraguayo; G. Thompson & Co. (to Australasia): Themistocles, Demosthenes and Euripedes; Wilson Line (to USA): Francisco, Idaho and Colorado; NZS Co. (to Australasia): Rotorua and Hororata; Federal Steam Navigation (to Australasia): Wiltshire and Shropshire; Shaw, Savill, Albion (to Australasia): Tainui and Pakeba; and Turnbull, Martin, Co. (to Australasia): Argyllshire. The Admiralty History does not list the additional ships envisaged in August 1914. The extra ten were: Allan Line: Grampian and Scotian; Canadian Pacific Railway: Montreal, Manitoba and Montezuma; White Star: Arabic, Adriatic and Baltic; and International Navigation Co.: Haverford and Merion. The three disarmed ships were those of the Wilson Line.

19. The Declaration was signed in November 1909. It should have come to a vote in 1909, but that year there was a constitutional crisis when the House of Lords rejected Chancellor Lloyd George’s ‘People’s Budget’. The crisis was resolved the following year.

20. Lambert, Planning for Armageddon, p 204.

21. A C Bell, A History of the Blockade of Germany and of the countries associated with her in the Great War: Austria-Hungary, Bulgaria and Turkey 1914-1918 (CID Historical Section, 1937 for official use only, published by HMSO, 1961). German provisions to resist economic attack are summarised on pp 192ff. The Germans called economic attack the third front of a war.

22. Bell, History of the Blockade of Germany, p 194, sees the Schlieffen Plan not so much as an attempt to win in a short war, as an attempt to preclude a dreaded long war. Schlieffen’s predecessor von Moltke the Elder had warned that Germany might face a seven- or even a thirty-year war.

23. Germany was slowly splitting into two almost separate economic units. The manufacturing areas in the west were fed via Holland and the Rhine. The more agricultural areas in the east sent their surplus to Russia and industrial Bohemia (now the Czech Republic) rather than to the west.

24. According to Bell, History of the Blockade of Germany, work on such studies was suspended during the Hague and London conferences, the Declaration of London being seen as a satisfactory guarantee against economic attack. Even after the British failed to ratify the Declaration, the Germans took it as effective international law, protecting them, although clearly Tirpitz disagreed. The results of new studies (one from the General Staff, one from a Dr Fröhlich) were circulated within the German government in November 1911. Bell associates the new studies with panic over the Agadir crisis that year. It appears from Bell’s account that the Germans did not take into account possible attack through the credit market.

25. Bell, History of the Blockade of Germany, p 196. The commission rejected the public demand for economic mobilisation directed by an economic general staff.

26. Bell, History of the Blockade of Germany, pp 196–8.

27. The initial application for reflagging, which was casually accepted, came from Chile. It later emerged that some of the ships, which had retained their German crews, replenished Graf Spee’s cruiser squadron when it appeared on the Chilean coast.

28. Tracy (ed), Sea Power and the Control of Trade, p 126.

29. This announcement is in Tracy (ed), Sea Power and the Control of Trade, p 177.

30. Bell, History of the Blockade of Germany, p 567.

31. I am indebted to Dr Stephen Prince, head of NHB, for this example.

32. Bell, History of the Blockade of Germany, pp 568–70. Maximum prices for staples were first set in December 1915 and in nearly all cases they rose substantially within a year. Between December 1915 and December 1916, for example, the maximum allowable price of a dozen eggs rose by 52 per cent to 357 per cent above the July 1914 price. The equivalent rise for lard was 315 per cent, for pork 249 per cent and for beef 170 per cent.

33. Conclusion drawn by Bell, History of the Blockade of Germany, p 674, based on German statements.

34. Bell, History of the Blockade of Germany, p 686. The exception was quality clothing.

35. Tracy, Attack on Maritime Trade, p 129.

36. As quoted in Tracy, Attack on Maritime Trade, p 144.

37. Bell, History of the Blockade of Germany, p 203, based partly on the German official history.

38. Correspondence quoted by Bell, History of the Blockade of Germany, pp 206–7.

39. Tirpitz also argued that, since Fisher had just replaced Battenberg as First Sea Lord, the Grand Fleet would probably soon mount an attack, so U-boats assigned to the High Seas Fleet should not be detached at this time for the trade warfare campaign. In December 1914 he wanted to wait until the Flanders U-boats were available in the autumn of 1915.

40. Bell, History of the Blockade of Germany, p 211.

41. In January the highest authorities outside the government, plus the professor of political science, the professor of law and five other dignitaries of the University of Berlin signed a paper advocating the submarine offensive; it went to the Chancellor as well as to von Pohl and von Ingenohl. They thought that the current economic problems would soon be overcome, but only temporarily. Recovery could only be made permanent by breaking the British blockade. To do this they advocated a co-ordinated attack by airships against London and by submarines against trade.

42. Bell, History of the Blockade of Germany, p 212, based on Zimmermann’s post-war recollection of a meeting at which Bethmann-Hollweg and, among others, General von Falkenhayn, were present. No minutes were taken. When Admiral Bachmann relieved von Pohl as chief of the Admiralstab the day after the meeting, he was astounded that the order for the submarine campaign had already been prepared, though the Kaiser had not yet signed it. Bell states that von Pohl made sure his assurance would not be questioned by other high naval officers and he did not take Tirpitz into his confidence.

43. According to Bell, History of the Blockade of Germany, pp 599–600, this was because in September, under the prize rules, the Flanders U-boat commanders managed to sink 82,000 tons of shipping. Their commander thought that on this basis he could persuade Scheer that a campaign embodying the prize rules would be worthwhile. The success of the Flanders flotilla led to the 6 October order that submarines in home waters were to concentrate on trade warfare (but under the prize rules). Scheer accepted the order as a preliminary to unrestricted submarine warfare. According to Bell, had the results of this short campaign been presented honestly, the Germans would have realised that they could cause the British considerable harm without risking American entry into the war. No comparison between prize-rule and unrestricted submarine warfare was ever presented to either the Chancellor or to Hindenburg and Ludendorff. In September Ludendorff told the Admiralstab that he favoured unrestricted submarine warfare and, according to Bell, that he thought it a great pity that the civil authorities had been given any say in the matter. Surely this was a purely military matter.

44. Krieg zur See, Nordsee 6 is explicit about the relationship between the situation in the East and German willingness to risk bringing further powers into the war. Bell, History of the Blockade of Germany, p 599, regards this reasoning as somewhat absurd, since it was the most distant of dangers. He sees the August meeting as the point at which it was decided definitely that unrestricted submarine warfare would be resumed. Apparently the decision was to await the defeat of Romania.

45. Lance E Davis and Stanley L Engerman, Naval Blockades in Peace and War: An Economic History Since 1750 (Cambridge University Press, 2006), pp 174–6 analyse this claim, which they trace to Prof Dr Hermann Levy, who specialised in British agriculture. He wrote that British grain stocks amounted to only 6½ to 17 weeks, so that the British relied on continuous supply. It seemed to follow that if imports were cut off when the stock was lowest, there would be an immediate panic in England, followed by a shortage so bad that the war could not go on. In the summer of 1916 Levy reported that overseas harvests had been poor, meaning that there was no surplus of North American wheat. It followed that the British would rely on more distant supplies, e.g. from South America and India, hence that each British freighter would carry much less wheat per month (because it would be making much longer voyages). That would further stretch the limited pool of British shipping. Most of Levy’s data turned out to be wrong and even some German politicians pointed out some of the gaps in his reasoning.

46. Davis and Engerman, Naval Blockades in Peace and War, p 176.

47. Ludendorff’s claim that the Admiralstab paper did not influence him and Scheer’s refusal to predict results within some fixed period, make it difficult to credit the explanation that the Germans expected to win in six months, before any US troops could reach Europe. Bell, History of the Blockade of Germany, suggests something far simpler. He sees Ludendorff’s published memoir as clear evidence of a very narrow military outlook, focussed on Europe alone. In that case the 15 December 1916 French counter-attack at Verdun really could have convinced him that the Germans were in deep trouble on the Western Front and that only some radical step could save them.

48. Krieg zur See, Nordsee 6, pp 230–1.

49. Davis and Engerman, Naval Blockades in Peace and War, cite a July 1916 Output of Beer Restriction Bill which saved 150,000 tons of imports annually; the creation of a Royal Commission on Wheat Supplies in October 1916; compulsory changes in the way flour was made to increase supplies (creating ‘war bread’) by Order in Council of 20 November 1916 and the creation of a Food Controller to buy food abroad and control its distribution in the United Kingdom. A Food Production Department was set up in the Board of Agriculture. It increased domestic wheat production by 60 per cent between 1916 and 1918. Measures were also taken to use merchant ships more efficiently; a Ministry of Shipping was formed in December 1916 to control all British merchant ships. In effect it added tonnage to replace what was being sunk. The Ministry was also responsible, initially, for new construction (but that role was taken over by the Admiralty). Later an Allied Maritime Transport Council (organised in December 1917) co-ordinated all Allied-flag shipping.

50. Tracy, Attack on Maritime Trade, p 139.

Chapter 4. Expectations versus Reality

1. Pre-war issues of Jane’s Fighting Ships generally included silhouettes of the major fast liners alongside warships in each national section. The Royal Navy became interested in using armed liners as auxiliary warships as early as the 1870s and it subsidised such liners as Lusitania and Acquitania. In the process it learned just how vulnerable such ships were. When war came, the Royal Navy mobilised a considerable force of armed merchant cruisers, whose superior seakeeping made them useful in enforcing the Northern Patrol element of the blockade of Germany. The Germans arranged to have minor warships on foreign stations and obsolete warships employed as station ships available to arm their own large merchant ships, but the results were very limited. Only six ships were taken over. The newest was Cap Trafalgar, a 23,640-ton liner which had been launched in March 1914. At Buenos Aires on the outbreak of war, she was armed by the obsolete gunboat Eber, which steamed from German Southwest Africa and then interned herself in Bahia. Cap Trafalgar was commissioned on 31 August and on 14 September she was intercepted and sunk by the British armed merchant cruiser Carmania, a considerably older (and much better-armed) former Cunard liner. Vineta (ex Cap Polonio) was a new liner launched in March 1914. She was never commissioned, apparently because she was not considered fast enough. The much older near-sisters Kaiser Wilhelm der Grosse and Kronprinz Wilhelm were also taken over. They were the fastest of the pre-war German liners converted into cruisers, at 22.5 knots. Kaiser Wilhelm der Grosse was in Germany when war broke out and was commissioned on 2 August 1914. She operated in the northern and mid-Atlantic. She scuttled herself when caught by HMS Highflyer while coaling off Africa on26 August 1916. By that time she had captured all of three ships (10,683 GRT). Her near-sister was in New York on the outbreak of war; she was armed at sea by the light cruiser Karlsruhe, but received no ammunition for her heaviest guns. Nevertheless, she managed to capture eleven ships (33,423 GRT). She operated with Graf Spee’s squadron for a time. She was forced to put into the United States, where she was interned, on 26 April 1915 and was later taken over as a US troopship. Two slower ships in the Far East were taken over and armed: Prinz Eitel Friedrich and Cormoran. On mobilisation the former was sent to Kaioutchao where she was armed with guns from the obsolete gunboats Luchs and Tiger. She operated for a time with Graf Spee and was the most successful of the lot (fifteen ships, 60,522 GRT). Once she ran out of coal, she was forced to intern herself in the United States, 26 April 1915. Cormoran was a prize taken by the cruiser Emden on 4 August 1914 and then armed at Tsingtao with the guns and personnel of the old gunboat Cormoran. She was interned at Guam on 13 December 1914 after running out of coal and provisions.

2. For details, see the author’s British Cruisers of the Victorian Era (Barnsley: Seaforth Publishing, 2012), much of which is based on the Trade Protection file held by NHB.

3. Budget figures from Jon Tetsuro Sumida, In Defence of Naval Supremacy: Finance, Technology and British Naval Policy 1889-1914 (London: Unwin Hyman, 1989), Tables 1 and 3. The Cressy class, the first British armoured cruisers, cost 36 per cent more than their protected cruiser predecessors of similar size; the next (Drake) class cost 32 per cent more again. The French were clearly well aware of the effect of building armoured cruisers; their 1897–8 programme sacrificed battleships to add armoured cruisers. Figures from Sumida, In Defence of Naval Supremacy, pp 19–20. A graph on p 21 shows that whereas first-class cruisers built in 1889–96 cost in toto about a third as much as the battleships of that period, first-class cruisers (mostly armoured) of 1897–1904 cost nearly as much as the battleships, despite the drastic attempt to cut cost by building inferior ‘County’ class armoured cruisers. Sumida points out that the big cruisers also demanded much more manpower, a problem Admiral Fisher’s fleet reforms were intended to address. It was also necessary to add barracks and docks, particularly as the cruisers were much longer than battleships of similar displacement.

4. Taylor, Struggle for Mastery in Europe, pp 376–8. Taylor sees the proposal as directed against Russia in the context of Russian seizure of Port Arthur, which revealed British inability to back the Chinese. Port Arthur gave the Russians a cruiser base outside the Sea of Japan, with a wide (hence difficult to block) exit to the Pacific. Imanuel Geiss, German Foreign Policy 1871-1914 (London: Routledge & Keegan Paul, 1976), p 85, notes that in November 1899 (after the attempt at alliance had died) German Foreign Secretary von Bulow was relieved that ‘nobody in Britain really knows how anti-British German public opinion is’. On 29 March 1898 British Colonial Secretary Austen Chamberlain, acting for the Prime Minister Lord Salisbury, who was ill, told the German ambassador that the British wanted to revise their policy of non-alignment and that they wanted above all alliance with Germany, because frictions with Germany were only secondary. According to Geiss, in the British Cabinet only Chamberlain and Chancellor of the Exchequer Balfour believed in the move towards Germany. They were exactly the ones who realised what the cruiser problem was doing to British finances and to her Empire. According to Geiss, the Germans feared that the British wanted them to fight the Russians. According to Geiss, the British had already approached Japan and the United States.

5. Initial very secret (presumably in the Merchant Navy Code) advice was for British shipping to steer course parallel to and from 30 to 150 miles away from their regular tracks, trying to fill their bunkers with enough fuel so that they did not have to coal on passage. They should try to pass through areas where traffic was most congested (i.e., focal areas) at night and to use neutral territorial waters when possible. Ships were to pass this message visually to any British ships they encountered. ADM 137/987, ‘Policy: Various’, 1914, p 337, repeated in modified form on p 386.

6. The Backhouse Papers (NHB) include a brief proposal by Captain Frederic Dreyer, dated 4 May 1914, to use plotting boards both to track the approach to the enemy fleet and to support divisional tactics. Document courtesy of Dr Jon Tetsuro Sumida. ADM 116/2090 is an early post-war compilation of descriptions of different approaches to plotting. ADM 1/8662/109 describes early post-war standardised action plotting arrangements for battleships and cruisers. According to the Admiralty Handbook of Plotting (CB 3039, August 1936, in NARA II ONI collection RG 38), the earlier practice of maintaining a chart of the situation on a navigational chart became unsatisfactory with the advent of radio and fast ships, so during the First World War two separate charts were maintained, one navigational and the other showing the tactical situation, for the fleet commander. Neither was satisfactory and plotting really became practical only with the advent of automatic devices such as the mechanical plotting table (ARL Table). They were being tested by 1930.

7. See the author’s Network-Centric Warfare: How Navies Learned to Fight Smarter in Three World Wars (Annapolis: Naval Institute Press, 2008).

8. Printed rules, as used by the Royal Naval College, in the Backhouse Papers, NHB. One Hercules could neutralise another in 22 minutes at 7000 yds, but in 50.7 minutes at 10,000 – figures somewhat more optimistic for gunfire than those Jellicoe used in notes for his 1912 manoeuvres. Unfortunately the more detailed figures Jellicoe used (printed as a Home Fleet memorandum) have not come to light; the PZ instructions in the preserved pre-Grand Fleet Home Fleet Orders do not include the ‘knock-out’ tables. Jellicoe’s estimate is in ADM 1/8268, the report of the January 1912 tactical (“P.Z.”) and strategic exercises by the Second Division of the Home Fleet, the only pre-war exercises for which such a package has survived. Objects included experiments with divisional tactics, communications and the use of fast divisions by an inferior fleet.

9. An undated version of Jellicoe’s war orders (probably 1912) is in Patterson (ed), Jellicoe Papers I, pp 23–5. They describe divisional tactics and the offensive use of destroyers. These orders are probably modified versions of Callaghan’s, which have not survived. Despite later claims that there were no written fleet battle orders before 1913, it seems likely that fleet commanders had been issuing them in memorandum form for many years; for example there are references to Fisher’s fleet memoranda. Without earlier orders, it is impossible to say how Jellicoe’s differed from those of the past.

10. The October 1913 instructions are given in ‘Naval Tactical Notes Vol I’, OU 6183, ADM 186/80, issued 1929.

11. Callaghan’s memo HF.0184 of 18 March 1914 in Backhouse Papers, NHB, argues that the Germans will not use theirs as independent hunters because they were unlikely to find anything (as in the 1913 manoeuvres). His diagram showed two flotillas, one either flank of the fleet in cruising formation (columns) so that they could deploy at the ends of the line. Callaghan expected 40 per cent hits in a ‘browning’ attack against ships in close order, say 2½ cables apart because if a ship was 200 yds long, the total space was 500 yds, so two-fifths of space was ships.

12. Churchill’s printed Minutes include one to Second Sea Lord and DNO dated 26 August 1913, asking for a demonstration ‘in the near future’ of ‘a squadron of at least 4 ships firing torpedoes at a similar squadron at the longer ranges. I am told that the torpedo practice of battleships is very far below the standard which obtains in submarines; and, further, that the danger to both sides of reciprocal exchanges of torpedoes when the lines are within effective range is greatly exaggerated. I wish to test this firstly by experience . . . the least possible warning should be given to the ships which are to fire. Each ship should fire several torpedoes; and the conditions of the work should be arranged so as to correspond as closely as possible to those which would arise in a fleet action where the lines had closed to within effective torpedo range’ (Churchill Minutes I, p 259, NHB). Churchill’s Minutes were apparently collected and printed as reference material when he began writing The World Crisis (alternatively, they may have been printed for the Dardanelles Committee).

13. Admiralty Gunnery Manual (23 December 1915 edition, superseding the 1911 edition) Vol III, pp 17–18. The section seems to have been slipped in without revising earlier sections which in effect warned against some of the new techniques.

14. That is not surprising. Because the new medium-range technique went nowhere (at least for major units), it did not figure in memories of fleet development. It is unlikely that anyone writing after Jutland wanted to say that the Royal Navy, which found itself fighting at longer ranges, had developed an effective way to fire at shorter ones.

15. Captain Henry G. Thursfield, ‘Development of Tactics in the Grand Fleet: Three Lectures’, delivered Feb. 2, 3, 7, 1922, in Thursfield papers, THU 107, National Maritime Museum, Greenwich.

16. Jellicoe’s August 1914 battle orders are in Patterson (ed), Jellicoe Papers I, pp 52–62, with the ranges given on p 59.

17. Patterson (ed), Jellicoe Papers I, p 63.

18. A paper on this topic is in the Backhouse Papers, NHB. The August 1914 Grand Fleet Battle Orders state that the case of fleets on opposite courses is to be described later.

19. Lieutenant R A R Plunkett (later Admiral Lord Drax), Modern Naval Tactics (Admiralty, 1910), NHB. The reference to regular turns to throw off enemy gunnery is from p 52.

20. NID report 973, dated December 1914 (ADM 137/4799).

21. May’s compilation of exercise data, issued on 19 September 1911, is in NHB. ADM 1/8120 is a parallel discussion of destroyer integration into the fleet, with comments by various subsidiary fleet commanders.

22. Comments are from Admiral Jellicoe’s ‘War Orders and Dispositions . . . prepared when in command of 2^nd Division, Home Fleet,’ but they seem to be his version of Callaghan’s orders. Item 18, Patterson (ed), Jellicoe Papers I. Much the same points were made during a discussion of the 1912–13 destroyers (‘L’ class) in the autumn of 1912.

23. NARA RG 38 ONI series E-2-d Register 09/572 of 1909–14.

24. May’s report is in ADM 1/8051. His account of the tactical implications (not in the PRO report), printed for fleet use, is in the NHB Library (copy courtesy of Dr Jon Tetsuro Sumida).

25. Conduct of a Fleet in Action: C-in-C’s Instructions – HF.04 Memo dated 14 March 1914. Fleet Memos 1914 from Backhouse papers in NHB.

26. Based on C-in-C High Seas Fleet, ‘Very Secret Tactical Order No. 1: Hints for Battle’, dated 29 September 1914, in NHB PG Box 634, SL 3078. The same document is in NID 979, ADM 186/17 (January 1915) and PRO also has a copy of the January 1914 version published for the Royal Navy in October 1914. The NHB copy is in typescript, but the wording is identical to that of the printed 1915 version.

27. This manoeuvre was described in paragraph 103 of the German fleet signal book (on the first page of the manoeuvring section). This is in the set of German tactical orders in ADM 186/17, the draught Manoeuvring Orders being dated 3 February 1914. A draft set of Signals in Action had not yet been issued. The signals were separate from the German Tactical Orders, which in this collection were ID. 979, O.X.O., dated January 1915. They included Tactical Order No. 1, Hints for Battle. This document included the note that the 4th Battle Squadron (pre-dreadnoughts) should take up a position on the disengaged beam of the battle line so that it could intervene or assist on its own initiative. The battlecruisers would be designated the ‘Fast Division’. Once the fleets were engaged, its duty would be to fire torpedoes at the head of the enemy line while engaging the advanced fast fighting forces of the enemy (i.e., British battlecruisers at the head of the British line) with gunfire. The best position would be about 4 points before the beam from the head of the enemy line.

28. In 1914 the 1st Battle Squadron was the Nassau and Helgoland class dreadnoughts (with reciprocating engines); the 2nd Battle Squadron was the Deutschland class pre-dreadnoughts, the last before the dreadnoughts); the 3rd was the turbine dreadnoughts of the Kaiser and König classes. The fleet flagship Friedrich der Grosse was separate, in contrast with British practice which made it part of a battle squadron. Tactical organisation of the High Seas Fleet in 1914, ADM 137/4715. By 1918, the 3rd Battle Squadron had been split into a five-ship squadron (Bayern and König class) and a five-ship 4th Battle Squadron (Kaiser class including Friedrich der Grosse, Baden now being fleet flagship). No pre-dreadnoughts were any longer associated with the fleet. Revised organisation from notes in German Navy Tactical Orders 1920, NHB.

29. ADM 137/2018, Grand Fleet Orders and Memoranda Aug–Dec 1914, p 116.

30. Patterson (ed), Jellicoe Papers I, p 75.

31. ADM 137/2018, pp 225–6, notes by Paymaster Hugh Miller on interrogation of a German lieutenant, distributed 3 December 1914 by Jellicoe.

32. Patterson (ed), Jellicoe Papers II, p 49, revision of Grand Fleet Battle Orders September 1916.

33. Tactical Orders for the High Seas Fleet War Edition, NHB Box PG 634. The numbered Tactical Orders, with their dates, were: (1) Method of screening the Fleet when under way of 9 April 1918; (2) Destroyer submarine screen of 10 February 1918; (3) Recognition Signals during Night Action of 19 October 1918; (4) Enemy Recognition Signals of 25 June 1915; (5) Conduct against Submarines of 25 June 1915; (6) Attack signals for TBDs in day action of 11 June 1915; (7) Depth adjustment of mines of 17 September 1914; (8) W/T Duties of Scouting Groups of 2 November 1917; (9) Rules for conduct of Fleet Flagship of 14 July 1914; (10) Report on Distribution of Fire of 19 July 1914; (11) Aircraft Division of 23 July 1918 (actually notes on the capabilities and characteristics of various fleet aircraft, including reconnaissance endurance 300nm, fighting reconnaissance endurance 240–300nm, long-range reconnaissance endurance 550nm. and ‘Giants’ endurance 750nm; all large reconnaissance and Giants had W/T with a range of about 150nm, using a hanging aerial 80 m long and the Giants could transmit on the Fleet Wave range over 450nm); (12) General Statement on naval airships of 5 July 1918 (L 50 and high-altitude L70 types); (13) Reconnaissance and screening by Airships of 13 September 1918; (14) Records during War of 6 October 1918; (15) Signals when attacking submarines with depth charges of 21 November 1918 (as written, but clearly wrong); (16) Employment of 2nd Battle Squadron of 10 May 1916; (17) Aerial escort service of 1 April 1918; (18) Conduct of fleet during night destroyer attacks of 24 May 1916; (19) Rapid mine searching gear (no date); (20) Reinforcement of 1st Scouting Group (Battlecruisers) by Battleships of 13 August 1916; (21) Cruising station of flotilla cruisers of 12 September 1916; (22) 3rd and 4th Battle Squadrons of 8 November 1916; (23) Conduct in submarine areas of 10 February 1918; (24) Approach formation of 1 March 1917. All of these were very brief items, typically one or two pages compared to the fourteen pages of ‘Hints for Battle’. Originally Tactical Order No. 15 (2 August 1914) was Orders for conduct after an action ending shortly before nightfall, which discussed whether battleships should be held back during the night to keep them out of the way of hostile destroyer attacks. These orders were printed up by the Admiralty as CB 1548 (April 1920), German Navy Tactical Orders, but the typescript includes notes on Jutland. NHB.

34. German and British night recognition signals are described in the 1920 collection of German tactical orders (CB 1548, NHB). The British data are dated 25 June 1915.

35. This section benefitted greatly from the material on the Royal Navy Radio Museum (HMS Collingwood) website, which includes the texts of the W/T Appendices to the annual Vernon reports from 1900 through 1913. Unfortunately it does not reproduce the secret W/T Appendices for 1914–18. The site also includes other important Royal Navy radio documents. It seems to be the closest approach to a history of Royal Navy radio taking technology into account.

36. Policy set in 1905 was for all ships above destroyer size to be fitted with W/T, with destroyers and lesser craft under consideration. The following year thirty-nine destroyers (‘River’ and ocean-going classes) were ordered fitted.

37. The Type system, superseding an earlier Mark system, was introduced in 1910, the first sets being Type 1 (Mk II: high power ship set for cruisers and battleships), Type 2 (Mk I*: 14 kW set for surface ships larger than destroyers, low-power shore stations), Type 3 (1 kW short-range set, initially for 5nm range; by 1913 redesignated the battleship auxiliary set), Type 4 (1 kW destroyer set), Type 5 (portable), Type 6 (harbour defence set), Type 7 (high power installation for Horsea/Cleethorpes) and Type 8 (Malta high power set). The first submarine installation was Type ‘X’ on B 5, a modified Type 4 destroyer set. Sets introduced in 1913 were Type 9 (cruiser auxiliary set) and Type 10 (submarine set, modified Type X; twenty-six RN submarines and two Australian ‘AE’ class fitted before the outbreak of war). On the eve of war a submarine set (Type 11) and an aircraft set (Type 12) were introduced. Wartime sets were: Type 13 (‘after-action’ set used if main sets were destroyed), Type 14 (3 kW Poulsen set, originally in Europa, later in submarines; Europa later had a prototype 14 kW set), Type 15 (3 kW Poulsen destroyer set, supplemented by Type 4), Type 16 (3 kW auxiliary capital ship set), Type 18 (25 kW arc set, 1916 for shore stations, capital ships and cruisers), Type 19 (Poulsen set at Horsea, operational 1914), Type 20 (Bermuda and Jamaica stations, 1915, spark sets), Type 21 (strategic CW stations, guaranteed range 1000nm at 1300m). There were three major wartime tube radios: Type 31 (coordination of fire by two or more ships using tubes to multiply frequencies), Type 32 (auxiliary to Type 14 submarine set) and Type 33 (auxiliary to Type 9 cruiser set). The first (post-war) generation of long-range tube sets was Type 36 (10 kW for large ships, tested 1921), Type 37 (destroyers) and Type 38 (submarines). There were separate designations for receivers.

38. Vernon Annual Report for 1906, It included three major fleet exercises carried out when the three fleets met off Lagos in February. Exercise I represented a division of each fleet watching a common enemy (HMS Exmouth), 40 miles away. Every half hour Exmouth made a signal and each watcher simultaneously tried to communicate with her own fleet. The watchers in turn received orders through a line of cruisers. Exercise II tested interference (jamming), an inner division closing with Exmouth, the main bodies being 40 miles away. Exmouth and battleships in company with her jammed. In Exercise III, Exmouth tried to evade and jam watchers with other ships at eight-mile intervals to transmit information. In all, thirty-nine signals were sent, of which the Channel Fleet correctly received eleven, the Mediterranean Fleet thirteen and the Atlantic Fleet only on. A further strategic exercise (IV) in which the Channel Fleet tried to bring one of the others to battle before they could unite. To do that it used a new frequency (F tune, 3000ft). One important lesson was that it was possible to use the new frequency while completely tuning out the A and B tunes used by the two other fleets. The Channel Fleet was therefore able to jam both A and B tunes. The wash-up conference (1 March 1906) concluded that ships should communicate on only a single frequency rather than the combination of two previously used. The conference also recommended adoption of a roof aerial instead of the existing fourfold type. Using a roof (distributed horizontal wire array) made it possible to stay the gaff from which antennas were slung.

39. The 1907 Vernon report contains the report of a W/T Conference held at Vernon on 31 October 1907. It cautioned that special cipher should be limited to messages absolutely to be kept secret, all other signals being made in Service code (a means of limiting the length of messages, but not secure). The means of automatically enciphering and deciphering was to be ‘on the typewriter principle’. The exercise that year showed that the Service code could be used very rapidly by anyone familiar with it, but that a ciphered message often was not attended to until after it became useless. It was recommended that a special war signal book be prepared, although it would become useless if it fell into enemy hands. Therefore the Supplementary Signal Code should not be issued to destroyers (which might well be lost near enemy shores), which should simply encipher their messages. A crypto machine called the Teletyper was tested in 1912, operating by transposing letters (the cover name was chosen so that manufacturers of parts of the machine would not know its purpose). It consisted of two standard typewriters side by side, with a common roller controlled by the left-hand typewriter. Depressing a key on the left hand typewriter sent an electrical signal to a ‘magic box’ which was electrically linked to the right-hand keyboard, so that depressing a letter on the left printed that letter and also the appropriate transposed letter on the right. A keycard in the ‘magic box’ set the connections. One sheet of paper provided eight different codes, obtained by turning the card 90° and turning it over. Twelve were ordered for initial trials. It was rejected on the ground that a transposition code was easily broken. Vernon proposed an internal means of changing the transposition after every letter had been transposed, but it was abandoned as unreliable. The encryption machine was apparently intended to solve a problem demonstrated by 1911 Home Fleet tests: a hundred-word message took thirty-five minutes to code and twenty-five to decode. Transmission (ten minutes) was half as fast as for an uncoded message, presumably because the receiver could not use the natural redundancy of language to overcome transmission errors. The process was far too slow for emergency tactical purposes. Experiments in 1914 with a coding machine (probably the Teletypewriter) failed. Admiral Jellicoe defined a series of fixed four-letter signals for wireless transmission. The 1911 and later Home Fleet W/T Orders include ‘Organisation of a Coding Branch’ on board ships as distinct from signallers, with two coders to stand each watch. ADM 116/1665. Instructions for the use of cipher, code and plain language were included in Grand Fleet W/T Memoranda dated 1 December 1916 (ADM 116/1664).

40. ADM 186/682, Wireless Instructions 1915 (CB.0121). The group system seems first to have been laid out in 1912; it can be found in the printed 1913 Home Fleet W/T instructions in ADM 116/1665.

41. These were the frequencies adopted about 1903: Tune A, Tune B and the two alternatives for Tune C.

42. It was said to produce a damped wave, because the burst of energy into the antenna (aerial, in British parlance) was brief and soon died out. The shorter a burst of energy, the wider the range of frequencies it represents.

43. All of this is from a modern perspective, not the way the inventors of early wireless systems wrote. It avoids numerous technicalities such as the difficult problem of feeding wireless energy from transmitter to antenna. Initially the aerial was connected directly to the spark gap (the Royal Navy called that the ‘plain’ system). The result was a highly-damped wave (a ‘whipcrack’ in American parlance). The wavelength was the natural wavelength of the aerial. About 1903 an alternative ‘oscillator method’ appeared (it became standard in the Royal Navy about 1907). Instead of feeding a spark directly into the antenna, the transmitter was linked to the antenna by induction (a receiver was linked to an antenna in a similar way). Because the transmitter was an oscillator, the wave was not as sharply damped as that of the ‘plain’ system (i.e., was not as broadband), although it was still a short burst of power. Induction coupling had a limited frequency range, so it provided a degree of tuning. The rival Marconi and Slaby-Arco systems differed in the technique they used for coupling, but both were roughly equivalent. In a typical system shown in the Vernon report for 1904 (p 22, in this case the US deForest system), a motor-generator supplied power through a transformer to a circuit containing the sending key. This circuit was linked by an inductance to a circuit containing a condenser, which fed into the spark gap proper. The spark gap was linked to an oscillator, a wire helix. That helix was linked inductively to the antenna. The oscillator converted the broad-band sparks into slightly narrower-band signals. Later in 1906 Vernon was able to receive signals from HMS Furious despite jamming due to HMS Culver six miles away, due to the ‘high note’ (modulation) inserted into the sparks produced by HMS Furious. Such high notes also made signals more recognisable despite atmospherics.

44. The frequency is more or less pure because any interruption or imposed signal, which is necessary to transmit information, mixes in other frequencies in small amounts.

45. That would be a heterodyne receiver, a type common in radios, which produces its own continuous-wave signal and ‘beats’ it against the received signal. Hence early references to Poulsen sets include references to special receivers. As of 1914, Poulsen receivers had been installed on board the battleship Marlborough and the cruiser Defence. The latter received signals at a range of 1100nm. ASWE History, ADM 220/2448.

46. ADM 1/8403/430 dated 17 November 1914. Director of Operations Division Admiral Oliver ordered Europa fitted with a Poulsen set during a machinery refit. She was the second long-range W/T ship, much wanted to improve Atlantic communications. Vindictive was the first such linking ship, described as ‘a wireless asset of great importance’, albeit limited because no other ship had similar equipment. She could communicate with Horsea at up to 1200nm range and Defence (which had a Poulsen receiver) could receive from her at a similar distance. On 12 November it was proposed that Europa and Argonaut or Amphitrite be similarly fitted (it is not clear which ship if any was fitted). A diagram showed that with a Poulsen set (on a cruiser) at Fayal or off Cape St. Vincent, Horsea and Gibraltar could link to each other and to ships half way to America. To the south, a link ship could connect with Ascension and Cape Horn. At this time Poulsen stations were being built at Jamaica and Ascension, but the cruisers would be ready first.

47. The Poulsen transmitter created an electric arc (in a gas) oscillating (Poulsen called it ‘singing’) between two carbon poles. It could not be keyed on and off, but superimposing another signal slightly shifted its frequency. The oscillator fed the aerial through a tuned circuit, the resonant frequency of which was determined by factors such as the length of the coil in it. That made it possible to signal, by sending the signal through one tuned circuit or another. When the Morse key was not pressed, the signal passed through one circuit; when it was pressed, through another. In one example, the difference was whether the signal went through one coil or two (1700 vs 1600m 176.5 vs 187.5 kHz.). The alternatives were called ‘back’ (‘back wave’) and ‘front’ (‘front wave’), receivers being tuned to the ‘front wave’. A receiver tuned to the back wave would hear an incoherent version of Morse Code. Within a few years the back wave was eliminated altogether. In effect the Poulsen was the first application of Frequency Shift Keying, the frequency indicating the information. The curator of the RN Radio Museum web site points out that the Poulsen was the only example of a true continuous wave (CW) transmitter, even though the term CW was later applied to all dot-and-dash transmitters, as an alternative to voice. The arc was conceived to support voice (telephony), but in 1909 the Royal Navy decided to abandon that and stick with Morse to avoid mistakes in recognising spoken letter and words with similar sounds.

48. The most important type of vacuum tube (triode) consists of a grid between a hot cathode and a cold anode (collector). Electrons boil off the cathode and travel to the anode, but the field created by the current in the grid controls how strong the current between cathode and anode is. Thus the relatively powerful current between cathode and anode can be made to reflect changes in the current in the grid. The output of one tube could go into another, or even back into the grid. In this way a vacuum-tube circuit could be made to resonate at a desired frequency, creating a continuous-wave transmitter. Because the grid current controls the flow of the main current in the tube, the British term a vacuum tube a valve. The great problem in tubes is the heat they necessarily generate (radar transmitter tubes are typically water-cooled).

49. The 1908 tube was used in a tuner; it lit up when exposed to a high-frequency discharge. It could also be used to indicate when an aerial was not radiating. In 1909 Marconi was using tubes for reception (sparks for transmission) at its long-range station at Clifden, Ireland, described in that year’s Vernon report.

50. The post-war set of Technical Histories (TH) does not include a W/T history, but does include (as TH 41) an account of the development of (a) The Model MN Eight-Valve Amplifier; (b) a CW Receiver; and (c) Tube experiments conducted in 1915–19.

51. Vice Admiral Sir Arthur Hezlet, Electronics and Sea Power (New York: Stein and Day, 1975), p 107. Tubes were also used in a herodyne circuit, but not for transmission, because they were considered too fragile. Hezlet may have referred to a 1916 project for a tube amplifier considered ‘essential for the furtherance of several W/T developments by HMS Vernon, which were in an experimental stage’. There was already a Model L tube receiver, which was needed, it appears, for reception by submarines at periscope depth. The British were trying a French three-tube amplifier for submarine reception. Most of TH 41 is a monograph describing the theory and practice of vacuum tubes.

52. There was no equivalent to the simpler more or less vertical whips used by the US Navy (but not the Royal Navy) during the Second World War, because wavelengths were far too long. Whips operated at HF (high frequency: roughly 1–30 MHz, meaning a wavelength of 10 to 300m), the operators accepting low efficiency in return for clearing the sky arcs of anti-aircraft guns. Marconi’s first antenna was a vertical wire – a whip.

53. Coherers were initially used to ink a moving tape rather than to feed headphones. By 1904 they had been replaced in British service by magnetic detectors. At that time the US Navy was using an electrolytic detector. In 1904 typical British W/T performance was ten to fifteen words/min (wpm), but de Forest (who provided US Navy sets) achieved thirty in trials. The Royal Navy adopted crystal detectors in 1909. They became popular after the First World War in home radios, the user moving a wire (the ‘cat’s whisker’) over the surface of the crystal to find the signal. This was the first solid-state device to be used in electronics, although no one then understood how it worked.

54. The Home Fleet W/T instructions included an article on Fleet Manoeuvring by W/T in its 1913 edition. Ships would use the vocabulary in the Fleet Signal Book, but with letters indicated by radio rather than by flags. A special ‘flag sign’ would precede each letter representing an alphabetical flag, to distinguish it from a letter indicating a procedure. Pendants would be preceded by a pendant sign. The technique seems to have been experimental, since an appended note pasted to the page requires all ships to keep rough records of all signals made and to be able to state course and order of the fleet at any instant. ADM 116/1665 (Article XVI). Grand Fleet W/T orders (ADM 116/1663) contain much more elaborate instructions for manoeuvring by W/T dated 1915, superseding those dated 24 December 1914 (not included).

55. ADM 137/2027, Post-Jutland Committees Pt 1, pp 373ff.

56. By the end of the war, the MBTW/T direction-finding set was standard. It equipped HMAS Australia and HMS Barham, Courageous, Lion, Queen Elizabeth, Princess Royal, Campania, Alsatian (AMC) and Orvieto (AMC). In the AMCs the direction-finder was installed for rendezvous purposes. A table of performance data showed HMAS Australia achieving 2° accuracy (but not reliably), much depending on the operator. Only comparative bearings could be relied upon. To avoid interference, the D/F system had to work on a frequency well separated from that of the ship’s main antenna. A difference of 200m was needed near S wave. The ship had two sets, one for Q and U waves and one for P to T waves. Her antenna was 85ft high, with a base of 40ft, placed unsymmetrically between the foremost and second funnels, very close to the main W/T feeder. The peak of the array was supported by the ship’s main aerial (actually, by the horizontal part of her aerial array). Lion reported an effective range of 300 miles, using a symmetrical array between her second and after funnels, 25ft high, with a 50ft base. Queen Elizabeth had a symmetrical array between her funnels, offering roughly 5° accuracy (range was not given). Princess Royal (which claimed 400 to 500nm range by day on Q wave and 5° accuracy with a skilled operator) had a symmetrical array over Q turret, about 30ft high, with a 45ft base. Errors were blamed on her lack of a gyro compass (presumably in her W/T office). Courageous (which claimed 3° accuracy out to 500nm) had a symmetrical array abaft her funnel, 66ft high, with a 48ft base. Barham (which claimed a range as great as 1000nm) had a symmetrical array 75ft high, with a 66ft base, suspended from the triatic between her fore topmast and her main top, close to her funnel. All of these installations used fixed vertical wire antennas and a rotating receiving coil, the operator deciding when he could hear the signal. The Grand Fleet also tested a French rotating-frame set intended for use ashore on trucks and in 1919 a more suitable rotating loop antenna was being made. Annual Report of Signal School 1917 (issued 1919), CB 1523 (ADM 186/753), pp 86–7.

57. Telefunken (Gesellschaft für drahtlose Telegraphie) was a monopoly formed by the merger of Braun-Siemens and Arco-Slaby. In 1905 it claimed to have fitted 450 stations, including eighty shipboard ones and about a dozen shore stations for the German navy. According to the 1905 Vernon annual report, Telefunken claimed that it had installed as many, if not more, stations than all others combined. The US Navy was using Telefunken systems extensively and they were erecting one in New Orleans with a 360-mile range to communicate with twenty-three of their other stations up the coast.

58. Patrick Beesly, Room 40: British Naval Intelligence 1914-18 (London: Hamish Hamilton, 1982), pp 30–1 comments that the Germans fell into the trap of using their W/T excessively because their transmitters were so good, better even than they realised (meaning that the British were intercepting their signals at extraordinary ranges, even when the Germans turned down the power to cut range). Using 800m signals, their big land station at Nauen could reach the Mediterranean, the Adriatic, the United States, Southwest Africa and even China. U-boat transmission (400m) extended over several hundred miles rather than, as supposed pre-war, fifty or sixty miles. Doubtless some of the British successes at long range were due to the adoption of vacuum-tube amplifiers developed from 1916 on.

59. In 1916 the British obtained and translated the German 1913 W/T regulations, which were still in force (July 1916: CB 0233, in NHB compilation, German Naval Warfare: Translations of Instructions Regarding Tactics, Torpedo Firing, WIT, etc 1914-1918). In contrast to contemporary British W/T instructions, it shows no emphasis on W/T silence or on intercepting enemy communications (it does discuss jamming enemy communications).

60. ADM 186/793, CB 1516(D), Reports on Interned German Vessels: Pt IV, Wireless Telegraphy Installations (June 1919). There is no indication of a separate report on radio telephone installations.

61. For example, Article XXVIII of Grand Fleet W/T instructions, issued on 21 September 1911, called for signals from all foreign ships and shore stations to be intercepted unless ship was told off by a W/T guard ship (i.e., unless other ships were already doing so). ADM 116/1665, Grand Fleet Signal Orders (up to 1917). A special log was to be kept; ‘much useful information with regard to foreign methods of W/T signalling can be obtained by careful and systematic interception’. An Admiralty letter had recently been issued explaining how analysis was to be organised. Pre-war Vernon annual W/T reports include comments on intercepts as measures of foreign capability and progress.

62. M (merchant ship method) was for a warship to simulate a merchant ship, among other things by running her alternators dead slow so as to create unmusical notes similar to pure spark transmissions; the NC (numerical call) method would avoid the use of known call signs; the F method permitted ships to answer while obscuring any connection with the relevant message; and the I (Indirect) method would be employed by ships at bases and by shore stations when transmitting to ships when the ships were not answering calls or messages. Ships were not to re-transmit by radio, but if necessary to pass the message to its intended recipient using visual signalling. Instructions issued 27 December 1915 in ADM 116/1663.

63. Beesly, Room 40, p 5.

64. Krieg zur See, Nordsee 3, p 126 points out that although the British might have obtained the keys to German codes from the stranded Magdeburg, it was also possible that they might simply have broken the codes on the basis of intercepts and analysis, ‘as the German coding system was by no means inviolable according to present 1922. ideas’. The Germans later broke British codes in this way. Beesly, Room 40, pp 22–3 describes the SKM (which was sometimes called the SB). It provided a three-letter code for each phrase, plus equivalents to letter which might have to be spelled out. Because obsolete phrases had never been edited out, by 1914 it had grown to a massive 34,304 groups. Because it was so massive, it could not easily be replaced and it remained in force as late as May 1917. Remarkably, the SKM remained in force for more than two years after the inquiry into the loss of the light cruiser established that it might well have been recovered by the Russians. Without alternate groups for each phrase, it offered opportunities for Room 40 to solve the keys used to super-encipher the text. Orders to change key were made by wireless and often not all ships had the new key, so that the same message was sometimes sent in both old and new keys.

65. According to Beesly, Room 40, p 27, the VB was the best of the German codes, intended for attaches and some embassies, consulates and agents abroad, as well as warships overseas (all by cable) and, with special super-encipherment, for flag officers. It was a 100,000-group code using five-digit groups with a particularly good potential for super-encipherment. Although not very significant in the war at sea, it was extremely useful because it enabled Room 40 to read German diplomatic traffic, particularly that between Berlin and Madrid and Berlin and Washington.

66. Hope’s collection of notes survives as ADM 137/4685 and ADM 137/4686. After the First World War Room 40 (ID 25) produced its own history of wartime German naval operations under the title ‘A Contribution to The History of German Naval Warfare 1914-1918 in Three Volumes’. Marked ‘Most Secret’, it is HW 7/1, 2 and 3.

67. Patterson (ed), Jellicoe Papers I, p 181, in Document 158, a letter from Jellicoe to the Secretary of the Admiralty, 14 August 1915.

68. Beesly, Room 40, p 145.

69. Beesly, Room 40, p 274.

70. Beesly, Room 40, p 277 describes various tips that Room 40 should have been able to assemble into a viable prediction. One reason for the 17 October debacle was a serious delay in passing on intelligence which should have alerted Admiral Beatty to what was happening. He speculates that the episode reinforced Lloyd George’s determination to get rid of both Admiral Jellicoe and Admiral Oliver, who as Chief of Operations was responsible for passing messages to the fleet. Beesly does not give details of Room 40’s information about the December destroyer raid.

71. Beesly, Room 40, pp 283–9. By this time the Germans were finally trying to take reasonable security measures, changing keys every seven to ten days (Room 40 now needed two or three days to crack them). They were also using low-powered transmitters, which were more difficult to intercept fully and accurately. For this operation, Scheer imposed not only full wireless silence, he issued all orders, even to minesweepers, in writing. Necessary wireless orders were passed by transmitting single code phrases (‘catch-words’). There were hints, but they were ambiguous. The worst failure was that the submarine J 6, on picket duty in the Heligoland Bight, mistook the High Seas Fleet for a British formation.

72. Beesly, Room 40, p 32. The operation is described in detail by Hilmar-Detlaf Brückner, ‘Germany’s First Cryptanalysis on the Western Front: Decrypting British and French Naval Ciphers in World War I’, in Crytologia (June 2005). According to Brückner, Roubaix had two Heavy W/T Stations, a Fortress W/T Station at Lille and a Motor Car W/T station, of which the expected ranges of the heavy stations were 125 to 250km and of the fixed fortress station up to 1000km. I am indebted to Dr Josef Straczek for this article. Beesly’s comments on German code-breaking are taken largely from a 1934 official paper by Kkpt Gustav Kleikamp, Der Einfluss der Funkaufklarung auf die Seekriegsfuhring in der Nordsee 1914-1918 (‘The Influence of Radio Intelligence on Naval Command in the North Sea 1914-1918’), which was presumably originally written for classified official use. Kleikamp was particularly disgusted that it took three years for the Germans to recognise their radio vulnerability.

73. This message (with its source) is mentioned in Krieg zur See, Nordsee 4, p 10.

74. Beesly, Room 40, p 32 also mentions centres at Libau for the Baltic and at Pola for the Mediterranean.

75. Krieg zur See, Nordsee 5, p 20. References in the 1915 volume suggest strongly that no messages, other than those sent in plain language, were then being read.

76. Brückner speculates that much of Roubaix’ success could be traced to the rapid expansion of the Royal Navy and particularly its auxiliary services, whose officers cannot have received much training in coding and therefore had to be given simple ones.

77. According to Brückner, the volume of signals dwarfed that handled by Room 40. Between August and November 1917 Roubaix handled 29,189 messages, whereas Beesly records that through the whole war Room 40 dealt with about 20,000 German messages (nearly all of which it broke). The success rate achieved by Roubaix is not recorded and most of the messages it handled were low-level.

78. Beesly, Room 40, p 33, quoting Kleikamp.

79. The purchase of the first German naval airship was correctly reported in 1912, with rumoured purchases of two more and press reports that airship sheds were to be built at Kiel, Hamburg, Wilhelmshaven and Emden. Submission by Captain Hugh Watson, naval attaché, 21 October 1912: Note on German manoeuvres 1912, in Backhouse Papers, NHB.

80. The Royal Naval Air Service had previously raided Friedrichshafen from Belfort in France (21 November 1914), nearly destroying L 7 as she neared completion. Douglas H Robinson, The Zeppelin in Combat: A History of the German Naval Airship Division 1912-1918 (London: Foulis, 1962), p 43. A British naval pilot flying from Antwerp the day before it fell destroyed the army’s Z IX in her shed at Dusseldorf on 8 October 1914.

81. The steamer Santa Elena completed conversion in mid-December 1917; she carried four seaplanes. The light cruiser Stuttgart was to have been the second aircraft mothership: Krieg zur See, Nordsee 7, p 174. Santa Elena does not figure in Gröner’s compendium of German warships, but the Stuttgart conversion does: it entailed adding a hangar with a flat deck on top and clearing the quarterdeck to take a seaplane which could be handled by a crane. The ship was rebuilt between February and May 1918 (she re-entered service on 16 May 1918). Gröner lists the incomplete Italian liner Ausonia as the first German ship earmarked as a carrier. Unlike the seaplane carriers, she had a flying-off deck forward and a landing deck aft. Conversion was stopped by the Armistice. The Ausonia design seems to have been developed from a 1917 proposal to rebuild the armoured cruiser Roon with a flat deck aft. According to R D Layman, Naval Aviation in the First World War: Its Impact and Influence (London: Chatham Publishing, 1996), p 41, as Baltic commander Prince Henry (Heinrich) of Prussia (the Kaiser’s brother) ordered the first embarkation of an aircraft on a German cruiser, Friedrich Carl, to reconnoitre the Russian coast. The prince was an early aviator. Layman credits him with the first German aerial torpedo operations and also with the first German aerial minelaying operation. The German July 1918 notes on air operations credited battlecruisers and sperrbrechers with two aircraft each and light cruisers one. In each case aircraft were slung out and retrieved after landing on the sea.

82. Based on an account of German wartime aeronautics in Marine Rundschau (March 1928 issue) as translated by ONI: RG 38 ONI series A-1-g Register 15461. Note that despite its title this article and others on German naval aeronautics were limited to airships. The Imperial Navy had an Airship Division. These data were supplemented by RG 38 ONI ‘German Naval Dirigibles’ (1921) A-1-p Register 9602, based partly on notes provided by the British Control Commission enforcing the Versailles Treaty. The British had good information on German airship construction because they intercepted the W/T messages sent when airships left their sheds for initial trial flights.

83. According to Layman, Naval Aviation, initially Admiral Tirpitz resisted investment; he later justified his resistance on the grounds that, as a seaman, he was too aware of the vagaries of weather, which would limit airship reconnaissance. Prince Henry of Prussia, the Kaiser’s brother, became an aviator and managed to reverse Tirpitz. According to Robinson, The Zeppelin in Combat, pp 20–1, Tirpitz changed his mind in 1911 in response to pressure from the Kaiser and from public opinion, but more in response to a threat by the Zeppelin company that without an order it would sell a ship to the British. The contract for L 1 was placed on 24 April 1912. Tirpitz demanded a larger ship than the standard type Zeppelin wanted to sell (25,000m³ vs 20,000m³). L 1 was completed on 25 September 1912 and first flew on 7 October 1912 (22,500m³, 518ft 2in long and 48ft 6in in diameter). On 13 October she took off for a flight lasting more than thirty hours. On 18 January 1913 the Kaiser approved a five-year programme envisaging two squadrons of Zeppelins (each of four operational and one spare airships). The first was L 2, a modified type with increased capacity but with engines too close to the gas bags. L 3 was delivered on 11 May 1914.

84. Based on Zeppelin war diaries, Robinson, Zeppelin in Combat, p 40, states that on the outbreak of war the naval airships were subordinated to Admiral Hipper, who commanded the First Scouting Group (battlecruisers) of the High Seas Fleet. Initially L 3 was assigned to patrol the inner destroyer line in the Heligoland Bight, making two cruises during the first week of the war. Her first flight under full war conditions was an 11 August reconnaissance of the Dutch coast out to a distance of 170 miles in response to reports of British minelayers. On 17 August she was assigned to reconnoitre the Skaggerak in response to a report that the British were operating there. She was also sent out during the battle of Heligoland Bight, but only glimpsed the retiring British force (and was fired on by German destroyers). On 19 October L 5 reconnoitred to within 60 miles of Great Yarmouth without seeing British ships, encouraging von Ingenohl to execute the Yarmouth raid. As an example of the limitations of Zeppelin reconnaissance in bad weather, L 4 passed close to the Harwich Force on 25 October when it attempted an abortive carrier strike. There was apparently no Zeppelin support for the Scarborough Raid in December 1914, but L 5 and L 6 did take off in an attempt to find the Harwich Force as it executed the Cuxhaven Raid on 25 December. L 6 tried but failed to hit the carrier Empress with a single bomb.

85. Airship size was measured in cubic meters (m³) of hydrogen volume. The L 3 class measured 22,500m³ (payload 8700kg). The successor L 9 of 1915 measured 25,000m³; the next class (L 10–L 19) measured 32,000m³ (payload 15,600 kg). They were followed by L 20–L 25 (35800m³, payload 17,800kg), L 30–L 42 (no L 26 to L 29: 55,000m³, payload 30,000 kg), L 43–L 52 (55,800m³, payload 39,000kg), L 53–65 (56,000m³, payload 40,000 kg, commissioned late 1917–early 1918; L 71 was in service on 10 August 1918, L 72 only after the Armistice). L 57 and L 59 were a special high-lift type (68,500m³, payload 52,000kg) with a maximum altitude of 6850m compared to 6000–6500m for the L 53 class, 5500m for the L 43 class and 4000m for the L 30 class (the maximum for the L 3 class was 2500m). L 70 was a new type embodying wartime lessons, capable of climbing to 7000m. L 59 made an extraordinary round trip to East Africa. She was a 55,000m³ ship lengthened 30m to accommodate cargo (her sister L 57 burned while on trials). She was recalled by W/T because East Africa had already been evacuated, so she flew back to Jamboli in Bulgaria after a 96-hour, 7000km flight. The final Zeppelins were L 70 (62,200m³, payload 47,000kg) and L 71 and L 72 (62,200m³, payload 49,000kg). There were apparently no L 66–69. Schütte-Lanz airships were SL 3 and SL 4 (32,400m³, payload 13,400kg) of 1914–15; SL 6, SL 8 and SL 9 (35,000m³, payload 20,000kg); and SL 20 and SL 22 (56,000m³, payload 35,500kg). There were also three Parseval (PL) semi-rigids used only in the Baltic and one M-ship (M-4, 19,500m³, 7000kg). Data from the Marine Rundschau article, which was presumably based on German records. According to Robinson, Zeppelin in Combat, p 35, the jump in size to the L 10 class was due to the fact that at the outbreak of war Zeppelin was building an enlarged airliner (LZ 26) for DELAG, the Zeppelin airline. L 9 was of this type. L 10 was an enlarged version offered by Zeppelin soon after the outbreak of war. Schütte-Lanz was already building a very large airship, which it sold to the army in May 1914; it received a navy contract for three airships (SL 3, SL 4 and SL 6, presumably part of the navy’s mobilisation programme). They were disliked for their fragile wood structure. At the end of the war the firm switched to metal, but too late to complete any airships before the Armistice.

86. Krieg zur See, Nordsee 4, p 108, arguing that in 1915 the German fleet was closer to equality with the British than its commander realised.

87. This PL 6 was used pre-war for advertising by the Stollwerck Chocolate Company. A second Parseval, PL 19, was one of four building under a British Admiralty contract. The only other navy Parseval was PL 25, ordered by the army. None was considered satisfactory. Robinson, Zeppelin in Combat, pp 36–7.

88. Initial Admiralstab and Navy Office proposals preceded those for the U-boat campaign. They were opposed on the ground that few airships were in service and all were essential for reconnaissance. Thus proposals made in September and October were postponed. When the idea was revived in December, it fell afoul of Chancellor Bethmann-Hollweg’s view that London should be spared. Admiralstab chief von Pohl (soon to command the High Seas Fleet) considered London and the military establishments further down the Thames the chief objectives in the campaign of terror which would soon include the U-boats. He also argued that the British would surely repeat their Christmas Day carrier air raid on the Zeppelins in their sheds. With its long nights, winter was the best time for an attack. A German deputation pressing the government to open the U-boat campaign linked it with a Zeppelin campaign. On 9 January the Kaiser approved an attack, but he barred any attack on London itself; bombs were to be dropped only on dockyards, arsenals and other military establishments. No one seems to have realised that bombs could not possibly be aimed that precisely. After delays due to bad weather, the first raid was carried out by two Zeppelins (one had to turn back) on the night of 19/20 January 1915. They attacked Great Yarmouth, a target justified by the presence of coast defences. L 3 dropped six 110lb HE bombs and seven incendiary bombs from 4900ft. L 4 missed the target and attacked a town from which anti-aircraft guns had fired at her; she dropped seven 110lb bombs. The Royal Navy saw Zeppelins as an extension of the coastal threat from German warships, hence as its responsibility. It armed warships in the approaches to London with anti-aircraft guns and it also placed guns on board submarines which it hoped could suddenly surface and, in effect, ambush Zeppelins.

89. Robinson, Zeppelin in Combat, p 81. The Zeppelin was soon forced up and out of visual touch by British light cruiser fire. The observation that only four British battlecruisers could be seen encouraged the Germans to think they had sunk HMS Tiger.

90. Tactical Order No. 12 in the collection of wartime German tactical orders, NHB.

91. Stephen W Roskill, Documents Relating to the Naval Air Service 1908 - 1918 (Navy Records Society, 1969), p 7.

92. Roskill, Naval Air Service, pp 45–1.

93. In January 1912, in addition to aircraft orders, Royal Navy Chief of Staff Admiral Troubridge called for the purchase of two non-rigid Zodiac (or Astra) airships, one from the manufacturer and one from Vickers (under license), for training. He also thought Mayfly should be reassembled for ground training and as a target simulating a Zeppelin. The CID suggested that the Royal Aircraft Factory build a non-rigid ‘Gamma’ type airship, of the type the army already had, for the navy; it was suited to training and to coastal scouting, but not to the ‘larger naval strategical requirements’, meaning long-range fleet scouting. In this sense the ‘Gamma’ was the forebear of SS 1.A ‘Willows’ airship which happened to be on the market should also be purchased (Roskill, Naval Air Service, p 53). By this time the navy was also buying a French Astra airship (Astra Torres). It later negotiated for rights to the German Parseval non-rigid (one of which flew from Germany to Sheerness and back).

94. Roskill, Naval Air Service, pp 87–107. Of the Parsevals, one had a capacity of 500,000ft³ and three of 300,000ft³.

95. According to the British Airship Trust web-site, the Vickers design (HMA 9) was based on technology revealed when German Army Zeppelin Z IV landed accidentally in France on 3 April 1913. Design work began on 9 April and the order was placed on 10 June. The order was cancelled on 12 March 1915 on the grounds that the war would probably end before the airship could be delivered, but it was reinstated in June and HMA 9 left her shed on 16 November 1916, the first British rigid airship to fly. Of the many missing numbers, a 1913 chart reproduced in Roskill, Naval Air Service shows all non-rigids except for HMA 9: HMA 2 was the Willows non-rigid, HMA 3 was the Astra-Torres, HMA 4–HMA 7 were Parsevals, HMA 9 was the big Vickers rigid, HMA 10–HMA 11 were the Armstrong Forlanis, HMA 12 was the projected third Armstrong Forlani, HMA 16 was the second Astra and HMA 17–HMA 20 were Royal Aircraft Factory non-rigids. The Forlanis were semi-rigids.

96. Drawings of a stretched version of HMA 9 were approved in October 1915 and three rigid airships (Nos. 23–25: ‘23 Class’) ordered from Beardmore, Armstrong, Shorts and Vickers (which was delayed by work to complete HMA 9; HMA 23 was completed 23 August 1917). By December plans called for sixteen airships of this type. Five were ordered in January 1916 (Nos. 26–30), but only HMA 26 (designated R (for Rigid) 26) was completed. These airships had a gun platform on top for a 2pdr and two Lewis guns, from then on a standard feature of British rigid airships. The other four were reordered to a modified 23X design, but R 28 and R 30 were later cancelled to concentrate on a new R 33 class based on technology revealed when Zeppelin L 33 was brought down relatively undamaged in North Essex. Five of these modified Zeppelins were ordered (R 33–R 37). In August 1917 the Admiralty proposed extending the order to a total of sixteen (to R 48) and cancelling R 28, work on which had not yet begun. R 31 and R 32 were derived from Schütte-Lanz technology, which paralleled Zeppelin’s. R 31 flew in July 1918 and was commissioned on 6 November, the last British airship of the First World War. As of the Armistice, R 32 to R 40 were to be completed, followed by one each year to keep the builders alive. The programme was soon heavily cut. Some of these data are from Roskill, Naval Air Service, p 531, quoting the Admiralty submission to the War Cabinet, 30 August 1917.

97. Roskill, Naval Air Service, p 319 reproduces a paper indicating that Flight Commander H A Williamson RNAS, who was invalided home from the Dardanelles to the supply side of the Naval Air Department, stated in 1915 that his own experience proved to him that the future of naval aviation lay with wheeled aircraft capable of taking off from and landing onto ships. To that end he conceived a primitive form of arrester gear and produced a crude model of a carrier with a flush flight deck and an island. He showed it to Captain Sueter, who in turn showed it to Naval Constructor J H Narbeth. An alternative put forward by Lieutenant Commander Gerard Holmes RNVR (also in DNC Department) had a completely flush deck, smoke being directed under it. This design was adopted for the carrier Argus. Williamson’s contribution was recognised post-war by the Royal Commission on Awards to Inventors.

98. Roskill, Naval Air Service, p 327, dated 26 February 1916. In the list, raiding targets ashore (including ships in harbour) came fifth and attacking enemy ships at sea came eighth and last. Sueter pointed out that only four carriers were large enough: Campania, Ben-my-Chree, Ark Royal and Vindex, but Ark Royal was too slow, Ben-my-Chree had no hangar forward (hence could launch only one aircraft over her bows before stopping to move aircraft around) and Vindex was too small, hence could fly off only two scouts before stopping. Sueter pointed out that it would be best to arrange a ship so that an aircraft could fly off and return aboard, citing the recent experiments.

99. Roskill, Naval Air Service, p 460.

100. Roskill, Naval Air Service, p 470.

101. ADM 137/1938, p 137, preface to a description of present policy on aircraft in ships, September 1917.