IT IS OFTEN SUGGESTED that most naval officers of the pre-1914 generation were obsessed with the new technology and hence neglected to think about how it would be used. Much the same was later said of US nuclear submarine officers. It certainly took considerable effort to master the technical aspects of new ships and weapons. However, there was lively interest in the tactical implications of the new technology. If anything, there seems to have been too much appreciation of where the technology might lead and too little of its current limitations. In Britain Admiral Fisher seems to have been the leading example of the problem. He could see the tactical implications of such new technologies as torpedoes and long-range wireless, but he grossly down-played current limitations. He seems to have had little or no interest in promoting exercises which might have highlighted steps needed to achieve what he expected the new technology to deliver. That is particularly obvious in hindsight in his approach to naval strategy using ocean surveillance via radio intelligence.
Lessons of the Russo-Japanese War
The most recent naval war, the Russo-Japanese War, was largely a fleet-on-fleet fight. Neither side had so many battleships that it could afford many losses and neither had any hope of adding them in wartime (the Russians tried but failed to buy several ships). Merchant shipping and blockade hardly figured in the war, but naval supremacy certainly did. At the outset, the Japanese considered the large Russian Pacific Fleet based in Port Arthur the single great obstacle to moving an invasion force across the Sea of Japan to Manchuria, the prize they sought. The Russians also had Baltic and Black Sea Fleets, which they could deploy (albeit laboriously) to reinforce the Port Arthur force. They could move troops, again somewhat laboriously, from European Russia to the Far East via the Trans-Siberian Railway.
In February 1904 Japanese destroyers successfully penetrated Port Arthur in an attempt to solve the problem at the outset. This attack was not very successful; torpedoes did not match their advertising. The idea of a surprise attack made a considerable impression on other navies and the Port Arthur attack may have inspired the attack on Pearl Harbor a quarter-century later. An initial fleet-on-fleet battle was also indecisive. Afterwards a Japanese mine sank the Russian Pacific Fleet flagship, killing its charismatic leader Admiral Makarov. His successor was a lot less aggressive, perhaps fearing that the Japanese had closed the mouth of Port Arthur with mines. The Japanese thus gained sufficient sea control to land their army. It seized enough ground to emplace heavy mortars on the heights above Port Arthur. They sank most of the ships inside the base during a lengthy siege. The Russians sought to lift the siege by bringing their more powerful Baltic Fleet halfway around the world into the Sea of Japan. By the time it arrived, Port Arthur had fallen. The Japanese destroyed the Baltic Fleet at the battle of Tsushima, which helped shape expectations for fleet battle in the next war.
The role of heavy guns was changing dramatically as the war was fought. Until very recently they had fired very slowly, the faster medium-calibre guns being considered decisive (the heavy guns were reserved to deal with the heaviest armour). Advances in turret design also made the heaviest guns far more usable. Experience at Tsushima in effect ratified a growing opinion that the heaviest guns were the weapons of the future. An increasing rate of fire was crucial, because it was assumed that the effect of shellfire would be cumulative. To disable or sink an opponent, a ship had to pour in a considerable volume of fire. That might mean fire by a numerous battery of large-calibre guns on a sustained basis. This perception was an important basis for the dreadnought revolution in capital ship design.
By way of contrast, it was understood that an underwater weapon – a torpedo or mine – could disable or sink a ship with a single hit. However, the chance of a hit by a slow-moving torpedo was considerably less than that with a fast-firing gun. In 1904 torpedo range was beginning to increase dramatically with the introduction of heaters, a simple form of internal combustion. At least in the Royal Navy, an important reason for extending heavy gun range was to keep British battleships out of torpedo range of enemy battleships. At least in theory, the art of successful naval battle was to combine guns and underwater weapons using new tactics.
Although torpedoes were relatively ineffective during the war, mines were quite the opposite. Their success led both the Royal Navy and the US Navy to convert cruisers to specialised minelayers (the Germans already had them). The Japanese revealed to the British that at Tsushima one of their destroyers sank a Russian battleship by laying a pair of connected floating mines in her path. By 1909 the floating-mine attack was a standard Japanese tactic. Destroyers passing 1000 to 2000m ahead of the enemy, on the same course, would drop mines, then steam away at maximum speed. Although analysis showed that such mines were not particularly effective, the idea made a considerable impact on Director of Naval Ordnance John Jellicoe. It almost certainly explains Jellicoe’s repeated wartime statements that the Germans planned to lay mines in the path of his fleet. Many German destroyers were fitted for minelaying, but of a more conventional sort. The floating-mine report appeared at about the same time the Royal Navy became interested in re-integrating destroyers into its battle fleet. On 24 April 1913 First Sea Lord ordered that all destroyers of ‘River’ and later classes should carry four mines each on the upper deck (each with 120lbs of TNT), the mines normally to be drifting but suitable for mooring. Little came of this idea, because the mines were never developed. However, ‘L’ class destroyers carried mine rails until 1915.
For naval thinkers in 1914, the great question was how to use guns and torpedoes in action. It was assumed that even a single underwater hit could be fatal – as when a single German mine sank the new dreadnought HMS Audacious in the Irish Sea on 27 October 1914. In fact it turned out that such ships could often survive multiple underwater hits. (Abrahams via Dr David Stevens, SPC-A)
The Japanese experience at Tsushima did not address other important questions. The most difficult was how to find the enemy at the outset. In the past, it was assumed that at the outset one fleet would blockade the other in its port. When the enemy tried to get to sea, the blockading fleet would attack and destroy it. During the late nineteenth century, that became less and less attractive because a blockading fleet made a good torpedo target. Initially that meant that the attacking fleet could not penetrate the enemy’s harbour, because it might be infested with small torpedo craft. Soon the torpedo boats grew to the point where they could operate well offshore and the area within sight of the enemy port became too hot to occupy. It might still be possible to station fast cruisers within sight of the enemy port. By 1904 early forms of radio had made it possible for such ships to provide an offshore fleet with sufficient warning. However, the trend was not good and the advent of seagoing submarines would make traditional forms of blockade altogether impossible.
At Tsushima the situation was simplified. The Russian fleet had to pass through a narrow strait (which gave its name to the battle). Japanese cruisers in the strait gave Admiral Togo sufficient indication of the approach path of the Russians for him to deploy appropriately. Once guns began to fire and coal-burning ships were steaming at high speed, no Admiral on his bridge could easily see (or communicate with) most of his fleet. Increasing gun range expanded the battle space, making it even more difficult for the Admiral even to visualise what was happening. No one said as much in 1904, but increasing numbers of ships also made it nearly impossible for an Admiral to maintain a mental picture of what was happening around him.
At Tsushima, Togo split his fleet into two squadrons because he doubted that a single officer could control more than eight ships. Later the choice was between such ‘divisional tactics’ and the tactics of a single concentrated fleet. Tsushima demonstrated both the potential of divisional tactics and their dangers. In theory one squadron was to support the other, but in fact the force split up. Togo was fortunate that one squadron did not fire on the other (it may have helped that the two fleets were very differently painted). The split offered the Russians valuable opportunities, which fortunately for Togo they wasted. In more modern terms, the issue was how to maintain both control and situational awareness on the part of the fleet commander.
Seapower and Seaborne Trade
In the popular mind, seapower in 1914 meant capital ships, or perhaps large cruisers standing in for them. Naval supremacy would be decided by a fight between the two opposing fleets. In fact the maritime prize was free use of the sea coupled with the ability to deny free use to the enemy. The capital ship fight was expected to determine who could use lesser warships to block enemy trade and to protect friendly trade. For example, it took large numbers of ships to stop numerous merchant ships. They could not possibly survive in the face of enemy capital ships. As a covering force, the Grand Fleet had to be able to fight and win a battle against the concentrated German High Seas Fleet, because if the High Seas Fleet broke out it could wipe out British shipping and also the British ships enforcing the blockade against Germany. Naval warfare was always about who could and who could not freely use the sea.
Despite much post-war talk to the contrary, the demands of trade protection shaped British naval policy – and even foreign policy – for decades prior to the war. The great charge against the Admiralty is that by the First World War it had lightly abandoned the successful trade protection policy of the past, convoy. Convoy Acts forced merchant ship owners to submit to Royal Navy orders and to join convoys with escorts. It was argued after the war that convoy had been abandoned because as a defensive strategy it was inferior to a more offensive form of protection, hunting down U-boats. The reality was considerably more complex.
Prior to the unrestricted U-boat campaign, the threat was cruisers, which combined high performance with endurance and firepower. With the advent of large fast liners, there was considerable pre-war speculation that they might represent a new kind of threat, but in practice armed liners were ineffective.1 In 1915 the Germans commissioned a new kind of auxiliary cruiser, a merchant ship operating in disguise (it was called an auxiliary cruiser [‘Hilfskreuzer’]). Similar ships were used during the next war. In neither case did they contribute seriously to commerce destruction.
Convoy died by the 1870s because with the advent of steam the Royal Navy had fewer and fewer ships which could match modern merchant ship performance (endurance at speed).2 Late nineteenth-century British war plans envisaged attacking French convoys and Russian outposts, at which commerce-raiding cruisers would be based. Troopships sent out for this purpose would be convoyed, but there was no hope of providing enough fast long-range cruisers to convoy most merchant ships. The only option was somehow to hunt raiding cruisers down. In December 1874 First Naval Lord Admiral Milne wrote an analysis of trade protection which shaped future policy. He argued that a commerce raider would be drawn into the areas where the great sea routes were concentrated – what were later called focal areas. Raiders would find themselves drawn into the focal areas, where they would meet British cruisers which would destroy them. Milne identified eighteen such areas. Even a focal area defence required numerous cruisers, which the British built.
It happened that the British enjoyed an important advantage. By far the best steaming coal in the world came from Wales and was controlled by British companies. Protecting overseas stores of this coal, which were mainly in British-controlled harbours, would go a long way towards immobilising raiders. In this sense the direct defence of British colonial harbours was a means of trade protection.
The focal area strategy was never made public, because implicit in it was acceptance of heavy early losses in exchange for the destruction of the enemy raiders. The problem would be solved in the first months of the war. This may seem bizarre. In fact in 1960 the US Navy’s position on trade protection was that it would be compelled to accept losses of about 100 merchant ships per month for the first three months of a naval war against the Soviet submarine force, during which time the Soviet force would be destroyed. The US strategy involved some convoying, but it was assumed that convoy would be ineffective against modern submarines. The emphasis was on blocking choke points and hunting using long-range detection. This was very much reminiscent of pre-First World War thinking about trade protection.
Pre-dreadnoughts were much more vulnerable to underwater damage. HMS King Edward VII sinks after being mined off Cape Wrath on 6 January 1916. In much the same way, the German pre-dreadnought Pommern was lost to a single torpedo hit the night after Jutland, but German dreadnoughts survived multiple torpedoes. (SPC-A)
Unfortunately an attacker did not have to match the Royal Navy’s numbers; it could build a few cruisers which could overmatch the focal-point ships. In the 1890s new lightweight steels made it possible to build fast armoured cruisers. Focal-point cruisers had to deal with whatever came their way. Armoured cruisers had other roles, too, such as fleet scouts and screens, but probably trade protection was the most punishing financially for the Royal Navy. By 1904 the British policy of maintaining a sufficient edge over the next two naval powers (France and Russia) meant maintaining a 2 to 1 advantage in armoured cruisers, presumably meaning roughly equal numbers as fleet scouts and as trade protection ships. To fill the focal areas the Royal Navy needed about as many armoured cruisers as it had battleships. Unfortunately a big armoured cruiser cost about as much as a battleship.
The Royal Navy was the single largest item in the British budget and the budget exploded with the rise of armoured cruisers. In 1896–7, before the British began building such ships, the British national budget was £101.5 million, of which £23.8 million (23.4 per cent) went to the Royal Navy. In 1904–5 it was £142 million, of which £41 million (28.8 per cent) went to the Royal Navy.3 The 1904–5 figure was worse than it looked, because the total budget was still swollen by increased army expenses related to the Boer War. The British sought an ally for the first time in their peacetime history at just this time, surely to balance off part of the Franco-Russian cruiser threat to British trade, the lifeblood of the Empire. The British first approached the Germans (as a counter to France), but found their terms unacceptable.4 The alternative was Japan, which could neutralise the Russians in the Far East. The Russian Far East threat was not so much to important British possessions (mainly Australia and New Zealand at this time), but to Far Eastern, including Indian, trade using large long-range cruisers.
Even with the Japanese alliance, the British faced a large French cruiser force, which could be based outside European waters the Admiralty might hope to control. The financial crisis was not quite as bad, but it still loomed. ‘Jacky’ Fisher was brought to the Admiralty in 1904 to solve it. He realised that if there could never be enough cruisers to fill out the focal areas, then enemy raiders would have to be hunted down. However, if the Admiralty kept track of British trade it would also be keeping track of losses to enemy cruisers. On that basis it could direct cruisers to hunt down the raiders. The new W/T technology would make direction more efficient. Fisher reorganised the Trade Division formed in 1901 as a trade tracking centre, the forerunner of a related ocean surveillance operation. These initiatives built on Fisher’s experience as Mediterranean C-in-C, using code-breaking to predict the movements of the French and Russian fleets he faced.
As long as the Germans relied on cruisers (including armed merchant cruisers), the policy Fisher had developed worked. It took enormous effort to hunt down some of the cruisers, but they did not last long. The Germans later enjoyed some success with disguised armed merchant cruisers like SMS Moewe, because even when spotted they could not necessarily be identified as warships by the cruisers hunting them. As an immediate defence while the German cruisers were hunted down, the naval war college recommended dispersal, as raiders relied on known trade routes to find their quarry. Dispersal might be the best protection. When war broke out in 1914, among the first Admiralty instructions to merchant ship owners was to shift away from the established routes.5
The great surprise of the First World War was that a particular kind of raider, the submarine, could pass right through any barrier created by surface forces. For U-boats, British supremacy on the surface of the sea was irrelevant – unless, as at Zeebrugge, it translated into an ability to destroy the submarines in their base. At the outset submarines were not counted as viable commerce raiders because they could not be used effectively without violating internationally-understood rules of war.
It proved possible to protect a ship against underwater damage by bulging, building protective structures outside her hull. Four Edgar class cruisers were the first British warships to be blistered (they were the only cruisers so modified). This is the bulge built onto HMS Endymion. The bowsprit was part of a bow protection device against mines, a predecessor of the paravane. The four modified Edgars went to the Dardanelles, where they were considerably more survivable than the pre-dreadnoughts previously sent there. None of the pre-dreadnoughts were bulged and the shipyard effort involved was too great to allow installation on board existing Grand Fleet capital ships. However, in May 1915 permission was given to bulge the incomplete Ramillies. Her sister-ships Revenge and Resolution were later fitted (in, respectively, October 1917 – February 1918 and late 1917 – May 1918). Surprisingly, the bulges in these ships did not cost speed. The Renowns and the ‘large light cruisers’ were all built with ‘internal’ bulges which did not cost speed.
Tactics: Gun and Torpedo
In the 1890s tactics often meant fleet manoeuvring in elaborate patterns (a well-known photograph of the time showed the British Mediterranean Fleet battleships executing the ‘gridiron’, in which lines of ships passed through other lines). By 1914 tactics meant how to manoeuvre a fleet so as to destroy or evade another fleet. The large numbers of the past had given way to smaller numbers of individually much more powerful ships.
Gun and torpedo ranges both expanded enormously during the pre-war decade. Fleets spread out. As a result, the battle space expanded to the point where an Admiral on his bridge could no longer easily see or understand what was happening. Even if he could have seen all of his ships under good conditions, in combat his vision was badly compromised by the combination of funnel smoke (inevitable for coal-burning warships) and gunsmoke. That combination was first really experienced at Tsushima in 1904.
In 1914 only the Royal Navy had the slightest idea of how to solve the problem. Commanding the Grand Fleet, Admiral Jellicoe adopted a suggestion by his fleet gunnery officer Captain Frederic Dreyer that he use a plot to visualise the positions of his own and enemy ships. In effect it was a small-scale equivalent to the strategic plot being developed at the Admiralty, first to visualise trade patterns (and find raiders) and then to control a fleet operating in the North Sea.6 Plotting had an unsuspected consequence. It was impossible to reproduce the flagship’s plot precisely, even if all the ships in the fleet were maintaining their own plots. Only the fleet commander, looking at his plot, could draw the implications needed for command. The plot did make it possible for the commander to assign, for example, a fast division where it might be needed, but it did not provide for him to devolve command to divisional commanders – who would not have sufficient plots of their own. The plot, moreover, made the C-in-C hostage to the accuracy of the information provided by his ships, which acted as his sensors.
Successful plotting involved not simply the location (at some particular time) of an enemy force, but its course and speed, so that it could be projected ahead and appropriate action taken. At Jutland, as Beatty’s Battlecruiser Fleet fell back on the supporting Grand Fleet, Jelicoe’s signalman asked for the enemy’s course and speed. Beatty, who seems not to have understood how important that might be, could signal only the direction to the enemy, which did not really help. Jellicoe managed to learn enough, probably from the smoke of the German fleet on the horizon, to estimate their course and speed and his simple plot made it possible for him to deploy his fleet on the appropriate course. The achievement was not only that he crossed the enemy’s T, but also that he avoided the entirely possible head-on approach which would have been so dangerous, at least in his eyes.
Jellicoe certainly realised that plotting required that his ships, particularly his scouts, report regularly and that they give the positions of the enemy ships they spotted. His wartime orders emphasise the need for including their own positions when reporting. However, plotting as a tactical information system appears not to have been tested before Jutland. Jellicoe seems not to have understood how much testing tactical plotting demanded. Perhaps the great surprise was that without precise navigation plotting could not be effective. As early as 1906 a British cruiser admiral pointed out that using W/T for scouting (i.e., abandoning chains of ships repeating back to the flagship) carried risks because reported positions might well be several miles off. That might not matter too much, because in a coal-burning era all ships smoked (though the Royal Navy’s Welsh steaming coal smoked less than most), hence could be seen beyond the horizon. However, ultimately what mattered was a series of reports on the same enemy ships by different scouts. The plot made it possible to combine such data to deduce the course and speed of the enemy. The results made sense only if enemy positions were indicated consistently. Otherwise the plotter would deduce an altogether wrong course and speed. Unfortunately the point of the plot was to predict where the enemy was going, which depended on his course and speed. The most famous aspect of Jellicoe’s plot at Jutland was its use to estimate the threat posed by enemy torpedoes. Several officers were detailed specifically to plot torpedo tracks and Jellicoe kept their plot within sight. Later it was suggested that errors in plotting made him overestimate the threat.
Failure always teaches more than success and Jellicoe learned a great deal from the failure of the attempt to intercept the German fleet after their Scarborough Raid in November 1914. The British light cruisers – the fleet’s scouts – made contact but failed to report it properly. Cruiser commander Commodore Goodenough disengaged due to a poorly-drafted signal sent by Admiral Beatty, who wanted to detach one of his cruisers to screen him against German destroyers he might encounter. Scarborough was the first success of British signals intelligence and Jellicoe, Beatty and Churchill thought it might well be the single golden opportunity to destroy the German fleet. Jellicoe and Beatty fastened on the cruiser failure, because had the cruisers held contact the battleships could have followed up. In one way the problem was the peacetime assumption that the force commander knew all and had to be obeyed. In another it was poor understanding of the full role of a scout, the most important part of which was full reporting of the enemy’s position, course, speed and composition. Commanding the British battlecruisers, Admiral Beatty was all for dismissing cruiser commander Commodore Goodenough in favour of one of his battlecruiser captains. Fortunately Goodenough survived to render excellent scouting service at Jutland. His failure at Scarborough explains numerous injunctions in Jellicoe’s fleet orders demanding reliable reporting by scouts.
Before Jutland, Admiral Jellicoe did his best to ensure that his scouting captains would report enemy positions and courses, so that he could form a full tactical picture. All Grand Fleet warships were required to maintain plots, on the basis of which they could (and should) report. However, the system was apparently not tested during fleet exercises, so Jellicoe was unaware of the dramatic effects serious navigational errors could have. Given bad data, the plot on his flagship could easily show ships doing impossible things – appearing to make 60 knots, say, or no more than 3 knots. Jellicoe may have suffered particularly badly from slapdash navigation on the part of ships of the Battlecruiser Fleet. After Jutland ships were required to maintain plots in terms of distance and bearing from the flagship, rather than in absolute terms of position. The use of relative positions largely solved the problem, except for ships, such as distant scouts, reporting from beyond the horizon.
Contemporary documents reveal no discussion of plotting or its profound implications and historians of Jutland have mentioned it only in passing. However, the Grand Fleet analyses of fleet exercises in 1918 always included a section on Navigation and Plotting, evaluating ships’ performance in this vital subject. For all the navies allied to the British, tactical plotting must have been a revelation. All of them adopted plotting post-war and it proved key to the night battle tactics used successfully by the British and the Japanese during the Second World War.7 The Germans had no equivalent. A neglected aspect of the battle of Jutland was that German commander Admiral Scheer continuously felt confused as he tried to extricate his fleet from the British. He had, it appears, very little sense of where ships were.
British wargame rules indicate what heavy guns were expected to do. British officers were familiar with them because they were used to evaluate success in frequent tactical (PZ) exercises. As of 1912 British exercise rules (quoted in instructions Admiral Jellicoe issued for tactical experiments) assumed that at 10,000 yds it would take 75 minutes for a modern battleship (like HMS Hercules) to disable another such ship. At 15,000 yds, it would take 300 minutes – five hours, an unimaginably long time. Yet 10,000 yds was the sort of range sought to keep battleships out of torpedo range of each other. It is no surprise that the Royal Navy became interested in concentrating the fire of several ships. Admiral Jellicoe credited torpedoes with 10,000 yds range at 30 knots – a distance they would cover in ten minutes. It must have seemed impossible that any battleship, or even any pair of battleships, could deal with her opposite in line that quickly. These figures are enough to explain why the Royal Navy adopted heavier guns. Official wargame rules issued in July 1913 took 12,000 yds as the maximum range for 12in and 13.5in guns.8 According to these rules it would take about 20 minutes for one King George V to neutralise another at 7000 yds and about 26.5 minutes at 10,000 yds. These rules reflected what British naval officers thought would happen in battle. The rules imply that even at 7000 yds it would take a long time to destroy an enemy battleship. The great wartime surprise, particularly at Jutland, was that gunfire did not have to be cumulative to be effective. The destruction of three British battlecruisers by what appeared to be single hits shocked not only the British but also the Germans. Later it turned out that horribly flawed British magazine practices had made these hits as effective as they were.
During the period after 1904, the Royal Navy sought longer and longer gun ranges so that its battleships could fight outside the range of torpedoes fired by enemy battleships. Ultimately North Sea weather limited gun range most of the time, but torpedoes kept improving. Trials in 1908 showed that a lengthened 21in torpedo could reach 10,000 yds at 30 knots and Jellicoe assumed this performance in tactical trials he conducted in 1912. The 21in Mk II which equipped British battleships in 1914 had a range of 10,000 yds at 28 knots. The German G7 was credited with 10,000m at 27 knots.
It proved difficult to keep increasing gunnery range whatever the weather. British fleet performance in 1911 was so disappointing that a special gunnery conference was called. In 1912 Battle Practice range was decreased from 9000 to 8000 yds. No explanation was given, but shorter ranges would have been consistent with the belief that North Sea conditions would make longer ones irrelevant. Partly as a result of reducing range, the eight best ships made better than 30 per cent hits in 1912, whereas in 1909 the fleet average had been about 20 per cent. Apparently existing gunnery techniques had reached their limit. It would take considerable investment in new technology to maintain the hitting rate while much raising Battle Practice range – which measured the fleet’s capability. The fleet would usually fight at or inside typical Battle Practice ranges. That was not to deny that, under good visibility, it might open fire at greater ranges.
One of the great wartime surprises was that the battles were fought at long ranges because fleets were unlikely to get into contact except when visibility was excellent. The failure to engage after the Scarborough Raid is a case in point: contact was lost partly due to poor weather (and partly, it should be added, due to a gross signalling blunder on Admiral Beatty’s part). That ships could and would engage at great ranges particularly surprised the Germans, who had justified a policy of using smaller-calibre higher-velocity guns on the basis of expected short North Sea ranges.
The 1911 Admiralty conference did not mention it, but by that time the Royal Navy was developing a new kind of analogue computer fire control under contract to Anthony Hungerford Pollen. Pollen’s Argo Clock modelled the firing situation to predict the range and bearing to which shells should be fired. Although Pollen had a monopoly contract, the Admiralty was also supporting an analogous system under development by its rangefinder firm, Barr & Stroud. Gunnery officer Frederic Dreyer was promoting a much simpler device which used a geometric approximation to predict range, without any modelling. In 1912 plans called for comparative trials of the Argo Clock and the Dreyer Table, but they were never carried out. Instead, the simpler and far less expensive Dreyer Table equipped most British ships during the First World War. It was a variation on the device most navies used to predict range: a ‘clock’ fed with the measured or estimated rate at which the range was changing. If the rate itself did not change very rapidly, using a constant rate was not too bad an approximation. The more sophisticated analogue computer was more flexible and it could allow for manoeuvres by the firing ship. During the war the US Navy adopted the modelling technique in the form of the Ford Rangekeeper, which seems to have been derived from the Argo Clock. The Germans used a simpler range clock fed with an estimated range rate.
Jellicoe seems to have opted for longer range despite reduced effectiveness. As commander of the Second Division of the Home Fleet in 1910–12, he issued war orders: if weather permitted he would open fire at 15,000 yds and develop maximum fire at 12,000 to 13,000 yds. He expressly cautioned against going inside 7000 yds, for fear of torpedoes. Apart from his 13.5in superdreadnoughts, Jellicoe’s fleet could not achieve much very rapidly at 10,000 yds.9
In October 1913 Home Fleet commander Admiral Sir George A Callaghan issued war orders envisaging opening fire at 15,000 yds (if weather permitted), closing to a ‘decisive range’ of 8000 to 10,000 yds where superiority of fire might be established.10 Ships might press home attacks at shorter ranges. No one had yet tried really long-range firing. Callaghan called for experiments to determine maximum effective range. The only ones he was able to run before the war were at about 14,000 yds. This figure is interesting because explicit war orders seem to have assumed that the fleet could fight at much greater ranges – as in fact it did during the war.
Both Callaghan and his successor Jellicoe planned to open fire at a range beyond that it which many shells would hit in order to establish ‘fire supremacy’. Having done that, they would close in to fire at a more practicable range. It would take a storm of shells to have the desired effect at that range. In order to allow for the wasted long-range shots, ships had to carry much more ammunition. To be able to fire steadily at long range, in 1913 Admiral Callaghan ordered ammunition added well beyond the usual eighty rounds per gun. In half an hour a gun might well fire sixty rounds and enough had to be left for the short-range action. Once war began, this likelihood that shells would be expended before ships closed to effective range became worrisome: surviving orders issued in 1914–15 explicitly warn against such waste. It must have been difficult both to urge gunners to open fire and not to waste. Yet there was no way to expand magazines.
Overloading magazines made for congestion at the bottoms of ammunition hoists. Extra cordite would crowd the spaces at the bottoms of the powder hoists. Supply to the guns would be anything but fast enough. Comments by German survivors of the Falklands battle must have encouraged attempts to fire faster. They said that slow British fire had made their gunnery easier. If a German ship was smothered in splashes, her gunlayers might well fail to hit altogether. Thus rapid fire was both an offensive and a defensive measure. Experience at Dogger Bank reinforced this view. Orders issued early in 1915 emphasised the need for rapid fire. Admiral Beatty sanctioned attempts to overcome congestion by drastically relaxing safety measures, with horrific effects at Jutland.
All of this left open the question of what to do if visibility precluded long-range firing. In that case the battleships would be within torpedo range and they would have only a few minutes of firing time until enemy torpedoes reached them. A battle line in close order was a perfect target for long-range torpedoes, because the ships in it filled so much of the line. These were ‘browning’ tactics. On land long-range guns fired, not at individuals, but ‘into the brown’, meaning into the mass of enemy troops. An enemy battle line could be imagined as an extended target. In order to concentrate its fire, its ships had to steam fairly close together – much closer than modern naval officers would consider safe. Perhaps 30 per cent of the total length of the battle line consisted of ships, which suggested that 30 per cent of ‘browning’ shots by long-range torpedoes would hit. Unlike shells, individual torpedo hits were expected to sink ships. ‘Browning’ shots figured in British destroyer instructions drafted in 1913.11
In The World Crisis Winston Churchill remembered that in November 1912 Second Sea Lord Admiral Lord Louis Battenberg warned of the threat of ‘browning’ shots from the German battle line. About a year later the danger had apparently receded.12 In Vol III of his history (1927) Churchill summarised what he remembered of pre-war expectations: first the British would smash the German fleet, then they would evade any torpedoes the Germans had launched. Evasion meant following a standardised signal, a blue pennant which ordered each ship to turn away from the enemy line. Ships’ fire-control solutions would inevitably be ruined by so radical a turn. Whatever damage they would do would have to be done during the ten minutes or less between coming into torpedo range and turning away.
Churchill’s account probably reflects what he was told when serving as First Lord before the war. Smashing means neutralisation. Even at 10,000 yd range, it would take a German torpedo only about ten minutes to reach the British line. Churchill associated the tactics he described with a half-hour engagement. The half hour could well mean a battle opening at much longer range, including an approach phase lasting as much as twenty minutes. During that time the British would try to achieve the ‘fire superiority’ Jellicoe sought. However, they would not begin to do devastating damage until the Germans came much closer. To do that they had to fire far more rapidly and more effectively than in the past, which suggests that they expected to use some new technique. A section hurriedly inserted in the wartime edition of the official Gunnery Manual describes a new means of rapid-fire control for medium ranges.13 It was not discussed after the war, as war experience emphasised long-range fire.14 An exercise Jellicoe conducted in October 1916 shows that by that time he was well aware of the battleship-to-battleship torpedo problem and that he did not consider it solved. The reason was the advent of really long-range torpedoes (15,000 yds or beyond).
When war broke out, the British lacked any proven capability to control guns at long range. That placed Jellicoe in a very different position when he took over the Grand Fleet. The fleet had been trained to fight at 8000 yds or less, well within torpedo range. Effective rapid fire at medium ranges was a future rather than a current proposition. In his initial Grand Fleet Battle Orders Jellicoe told the fleet that it would fight at long range. The British had long (correctly) believed that the Germans wanted to fight at much shorter ranges, where their heavy secondary batteries would be effective. Short range would also favour their favourite manoeuvre of passing torpedo craft through their line to attack the enemy during a gun engagement, not least to break up the enemy line. After the war it seemed that Jellicoe had concluded that since the Germans wanted to fight a close action, it would be advantageous to fight at longer-range, to ‘develop and practise the game of long bowls’. It also seemed that Jellicoe had decided to concentrate completely on gunnery.15 Jellicoe’s avoidance of torpedo range made sense. His pre-war experience as DNO and then Third and Second Sea Lords made him painfully aware of the failure of British attempts to develop underwater protection. He believed, again correctly, that the Germans had done much better. His only superiority over the Germans was in long-range heavy gunnery. He was also aware that gunnery, particularly at long range, could not give decisive results in the time available.
Jellicoe had no particular reason to imagine that ships and materiel which had struggled to hit targets at 9000 yds could suddenly hit at 50 per cent greater ranges. His August 1914 draft tactical orders envisaged deploying at 16,000 yds, but opening range was dropped from Callaghan’s 15,000 yds to 9000–12,000 yds (which he described as long range).16 Jellicoe expected to benefit heavily from British gunnery superiority at such ranges. These orders suggest that Callaghan’s longer-range test firings were less than successful and, moreover, that Jellicoe expected decisive range to be something like 6000 to 8000 yds (no decisive range was cited in the draft orders). A few weeks later Jellicoe issued a radically different Addendum No. 1 to Grand Fleet Battle Orders envisaging opening with 13.5in guns at 15,000 yds and with 12in guns at 13,000 and at even greater ranges should the enemy fire first. Ships would shift to rapid fire at about 10,000 yds. The orders were extraordinary. Almost nothing had been done to practice such firing.
The blister could be badly torn up without endangering the ship. This is damage to HMS Endymion.
Torpedoes armed many kinds of ships, from capital ships down. Pre-war tacticians had to take into account both torpedo fire from an enemy battle line and possible torpedo fire from enemy destroyers and light cruisers. This is the torpedo room of HMAS Australia. (Josef Straczek)
Jellicoe did not intend to open fire beyond 18,000 yds unless the enemy did so, or the conditions required it – as in a chase, which occurred at Dogger Bank. Ships must always be ready to fight at extreme ranges. Jellicoe’s April 1915 gunnery orders show that to do that he relied on the pre-war technique of bracketing to find the range, which meant not firing the next salvo until the first had splashed and been spotted. Thus at 18,000 yds a 13.5in ship could fire salvos at 50-second intervals (at 12,000 yds, 40 seconds). Once the range had been found, ships would shift to ‘rapid salvos’, the next salvo being fired as soon as guns were ready. Any spotting corrections would be applied to a later salvo. Heavy guns could fire at least two such salvos each minute. Jellicoe’s April 1915 instructions envisaged accelerating to a salvo every 20 seconds once spotting could be discounted (as in rangefinder control). The maximum range envisaged was at least twice pre-war range. Except for directors, nothing had been added to improve performance. Jellicoe later wrote about how hard he had worked the fleet to improve its long-range performance. He was fortunate in having the Moray Firth in which to fire (Beatty’s battlecruisers had no such practice area). In 1915 they fired at least twice at 16,000 to 17,000 yds.
Jellicoe’s most significant change to Callaghan’s battle orders was to assign the two leading pairs of British battleships to concentrate on the two leading German ships (if the fleets were on opposite courses, the two rear pairs would attack the German van). The idea that concentrated fire could break up the German formation became more important after Jutland and was an important post-war theme in British naval tactics. It backfired to some extent at Dogger Bank, when the captain of HMS Tiger interpreted it to mean that he should fire on the leading German ship instead of on the ship opposite him. Post-Jutland British fire-control developers accepted that hits would be few and involved new types of salvo firing and spotting designed to make the most of whatever opportunities arose. Pre-war concepts involving rate measurement were abandoned.
After the war many British naval officers expressed their disappointment in Jellicoe’s performance at Jutland: why had he failed to come to grips with the Germans? Surely he could have achieved more at shorter ranges. In fact it did not much matter what ranges Jellicoe wanted to adopt. He fired as soon as he could, as the Germans approached and his fleet did not have all that long to fire before the Germans tried to disengage. Jellicoe could have chosen the range only if he had much faster ships.
Soon after war broke out, Jellicoe wrote in his battle orders that he sought action on parallel courses because he thought it would give the most decisive results and because it would avoid the danger of German minelaying: the Germans would not run their own fleet (alongside the British) into a minefield they had laid.17 This was striking because the object of deployment as developed as early as 1901 (and as practised by Jellicoe at Jutland) was to lay the fleet athwart the path of the enemy fleet, crossing his ‘T’. Jellicoe’s preference for parallel courses may have been tied to his fire-control capability. The closer the enemy course paralleled his, the lower (and less variable) the range rate and the simpler the fire-control problem. Jellicoe almost certainly associated a low range rate with a high hitting rate. One of the nightmares considered by his officers just before the war was a German reversal of course, which would greatly increase the range rate.18 Not only that but a fleet on the opposite course would enjoy greater effective torpedo range, because the opposing fleet would be running towards its torpedoes as they ran through the water.
The Germans apparently expected to fight at much shorter ranges. It was generally understood that the Germans considered their gunnery most effective at medium range, about 6000 yds. Evidence included the fact that they had retained medium-calibre (5.9in) guns on board their dreadnoughts, at a cost in weight (for armour or for main-battery guns), on the theory that the smaller faster-firing guns could contribute usefully in a fleet action. As if to confirm British guesses, in his 1915 tactical orders (which the British obtained) German fleet commander von Ingenohl announced that he expected to fight at 6600 to 8800 yds (presumably a translation of 6000 to 8000m).
The Germans had to get to their preferred battle range while the British shelled them from greater ranges. They adopted a tactic of evasive action (zig-zagging) while closing the range, on the theory that it would defeat any British attempt at measuring rates. German comments on British tactics strongly suggest that they were aware of British reliance on plotting to measure rates Because their rate estimates did not involve sustained plotting, the Germans could still exercise effective fire control while manoeuvring Zig-zagging frustrated British fire control, particularly at Jutland and after that battle the British changed their fire-control practices to deal with it. Ironically, a British officer proposing tactics pre-war had suggested exactly this technique, pointing out that fire control would still be possible from a ship moderately zig-zagging, whereas her manoeuvres would frustrate conventional fire control used against her.19
Torpedoes could turn a small ship into a giant-killer. Before the war, that meant destroyers and submarines. During the war, the advent of powerful aircraft engines made it possible to build very fast torpedo boats. The first were built in Austria-Hungary, but the British CMBs (Coastal Motor Boats) saw much more action. CMB 3, a 40-footer of the initial batch, runs trials. Her hard-chine hull form was based on that of pre-war racing boats. (Dr David Stevens, SPC-A)
Until about 1911 the Germans seem to have thought that by adopting high range rates as they approached the British, they could avoid almost all damage. Then they realised that might not be sufficient protection. If they could fire on the way in, while the range was changing, they might make British fire control ineffective. Observing the Germans, in 1914 the Russians noted that the Germans emphasised radical changes of course and high range rates in their exercises. After Jutland, the Germans told the Austrian Naval Attaché that in their exercises they had always worked with big and rapid alterations of range and exercise firing while turning. They thought the British relied too much on range clocks and hence on measured range rates, which required that they maintain a steady course. In the autumn of 1914 British naval intelligence published the secret German report of gunnery practice for 1912–13.20 The Germans had recently begun practicing long-range firing under ‘difficult conditions’, at ranges of 11,000 yds and above. The longest range for any of the capital ships was 15,000 yds. For example, the recent battleship SMS Kaiser had fired at 14,000 yds down to 13,500 yds at a closing rate of 43 yds/min. Typical results for heavy ships were 9.2 per cent hits at 12,000 yds.
It is not clear why, without a much higher speed than the British, the Germans thought they could ever close the range, whatever damage the British could or could not do. During the inter-war period, the Royal Navy considered medium range (about 15,000 yds) best and had to solve the Germans’ problem, of how to survive while closing an enemy which could hit at much greater range. Among other things, it hoped to use mass destroyer torpedo attacks to help force an enemy into position. The one problem it could not solve was that its battle line was significantly slower than that of its most likely enemy, Japan (this particular problem was unsuspected, as the British did not realise that the Japanese had greatly increased the speeds of their battleships during reconstruction).
Once war broke out, the Germans concentrated on the tactics they had developed to minimise their chances of being hit; their mind-set was fundamentally defensive. They employed the tactics they had designed to get them into fighting range without being destroyed on the way, then stopped closing well before getting there. Thus battle ranges were almost always much longer than the Royal Navy had practiced before the war. The Germans had never expected to achieve much en route to decisive range. The destruction they wrought at Jutland was a surprise as much to them as to the British.
The torpedo was radically different from the gun. Each shot was much more lethal, but at long range hitting was far less likely, because the torpedo took so long to reach its target. The navies of the First World War dealt with two dramatic changes in the torpedo. One was the gyro, which ensured that a long-range torpedo would run more or less straight. The gyro could also be set to turn the torpedo away onto a pre-set angled course. Navies differed as to how valuable this feature was, because it required more sophisticated fire control and also because it might make for reduced reliability. The other was the ‘heater’, a form of internal combustion which dramatically increased torpedo range. By 1914 ‘heaters’ had boosted maximum range from perhaps 1000 yds (1900) to about 10,000 yds or more. Even well before 1910 navies were taking the torpedo seriously as a complement to or even an alternative to guns. The US and Russian Navies and possibly others, seriously considered capital ships whose main batteries would be large numbers of torpedoes. Some British ordnance experts suggested that now the torpedo might be used alongside the gun in a fleet action. It was, however, generally accepted that even a gyro torpedo had limited accuracy at full range.
The torpedo turned a destroyer or even a motor boat into a giant-killer – as the Italians demonstrated when they sank the Austrian battleship Szent Istvan in 1918. During the run-up to 1914 admirals had to take into account torpedoes on board both battleships and destroyers (designated torpedo boats in the German navy). The British Mediterranean Fleet integrated torpedo-firing destroyers into its formations beginning about 1900, but as First Sea Lord Admiral Fisher (who had originated the Mediterranean tactics) argued that destroyers should be used independently to dominate the narrow North Sea. The British imagined, wrongly, that the Germans planned to use their destroyers similarly.
Pre-1914 navies treated destroyers not too differently from the way later navies treated aircraft. Discussing fleet torpedo boat tactics, a US Admiral later wrote that of course any battleship which spotted an approaching torpedo boat (or destroyer) would open fire. Waiting for a positive identification would be too dangerous. The surface torpedo firing zone grew rapidly in the years between 1904 and 1914, making the identification problem less and less tractable.
The situation changed after Fisher left the Admiralty and reports that the Germans were integrating their destroyers into the Grand Fleet gained currency. The Germans practiced a showy manoeuvre in which destroyers on the unengaged side of their battle line passed between their battleships to attack the enemy battle line during the gun action between the two lines. Initially the British reaction was derisory. It would be far more efficient to attach destroyers to the van or the rear of the fleet, to disrupt the enemy line by concentrated attacks on his van or rear. Destroyers venturing into the space between the two fleets would make few hits and would probably be wiped out. However, the German attacks could disrupt the British formation, ruining its gunnery: the German torpedo tactic could be read as a tactical gunnery countermeasure. Until this point, the British saw destroyers as a threat entirely separate from a day gun battle, so their capital ships had their anti-destroyer guns atop their turrets. They switched to protected positions and new capital ships, beginning with the Iron Duke class battleships and their battlecruiser equivalent, HMS Tiger, had more powerful anti-destroyer guns.
To see what the Royal Navy should do, in 1909–10 Home Fleet commander Admiral William H May conducted exercises.21 As in the past, he was interested mainly in using destroyers to finish off enemy ships crippled during a gun action, as Admiral Togo had planned to use his own destroyers at Tsushima. Only if weather was misty (visibility less than 8000 yds) could they get close enough to an undamaged and unengaged enemy fleet to attack it before the gun action. However, May did take account of the new long-range torpedoes, which could be used in ‘browning’ shots against the enemy line rather than against individual ships. In his exercises, destroyers suddenly emerging from the mist made ‘browning’ shots at 3000 yds.
An exercise seems to have demonstrated the value of ‘browning’ rather than close-range aimed attacks. White’s destroyers emerged from the mist and attacked as Black’s fleet began to form battle line out of cruising columns. During that manoeuvre the enemy fleet could not turn away to evade ‘browning’ shots by the attackers. He would have to rely either on counter-attack by his own destroyers or on his battleships’ fire – in which case the battleships might be unable to engage the opposing British battleships. At the least, dreadnoughts would have to assign one turret to anti-destroyer fire, reducing their effective broadsides.
The destroyers could have made a successful ‘browning’ attack, but in order to hit particular ships they came too close to Black’s ships. Black would have been unable to form his battle line unless he had light cruisers to beat off this attack. There would be no time to call the cruisers forward if they were not already present on the wings of the battle fleet. May concluded that in misty weather the destroyers should be well up on both wings of the battle fleet.
If a fleet had already formed, its battleships could turn away together – but that might well lead to great confusion. Once the fleets were engaged, the enemy battleships would be concentrating their efforts on the British battle line. Small ships might not even be too visible in a haze of gun and funnel smoke. Running at maximum speed, destroyers might well survive until they fired. Again, May envisaged ‘browning’ rather than short-range aimed shots. To avoid being sunk (wasting their torpedoes), destroyers should be able to fire as soon as they came under enemy fire.
Once the battle had begun, destroyers approaching the enemy fleet had to pass through a danger zone extending from 1000 yds in front of the British line to 2000 yds beyond it, but they would be fairly safe from attacks by enemy light cruisers. A run from the end of the British line would place destroyers in the danger zone for longer (nine minutes in the worst case), but it would hamper the British line far less than the pass-through.
CMB 4 is preserved at the Imperial War Museum at Duxford to remember the great successes CMBs achieved in the Baltic during the British intervention against the Bolsheviks in 1919. The single 18in torpedo was stowed in the trough, head first. A piston, which has not survived, pushed it into the water tail first to launch it. The boat swerved out of the way of the accelerating torpedo. This arrangement made it possible to aim the torpedo (by aiming the boat). Although the cockpit appears to be nearly in the bow, it was actually well astern (abaft the step of the hull) and there was a machine gun position forward. The boat had a single aircraft-type engine driving one propeller. In February 1919 the British Secret Service asked Lieutenant Augustus Agar RN to run agents into St Petersburg. In June, Agar made the small harbour of Terrioki, three miles from the Russo-Finnish border, his base. It was 35 miles from St Petersburg. His two 40-footers were towed from Abo, Finland to Terrioki by the destroyer HMS Voyager. As Agar was setting up his base, the fortress of Krasnaya Gorka rebelled against the Bolsheviks and Agar decided to attack the Bolshevik cruiser Oleg. His CMB 4 sank the cruiser on the night of 17/18 June after having to stop for twenty minutes to replace the torpedo-launching cartridge. This was much the kind of attack the Harwich Force planned against the German fleet from 1916 on. Between 10 June and 26 August 1919, Agar made nine trips to St Petersburg to insert or pick up agents. Of his two boats, CMB 7 was seriously damaged on the last trip and had to be scuttled. Agar’s success led Baltic commander Rear Admiral Sir Walter Cowan to ask for more CMBs to attack the two battleships and one submarine depot ship in St. Petersburg. He received a small flotilla of 55-footers. On the night of 17/18 August Agar in CMB 4 led six 55-footers through the forts to the entrance to Kronstadt. They were preceded by an air attack intended to distract the defenders. Two of the boats were lost to enemy fire and a third to a collision, but the others destroyed the enemy fleet, sinking the battleships Petropavlovsk (hit by three torpedoes from CMB 31 and CMB 88) and Andrei Pervosvenni (one torpedo from CMB 88). Petropavlovsk was later raised and returned to service, but Andrei Pervosvenni saw no further service. In a further operation in September, CMB 4 and two 55-foot CMBs laid mines in the channel leading to the naval base of Kronstadt. Agar received the VC and the DSO. (Dr Raymond Cheung)
Most CMBs were 55-footers, as depicted by this model. Only the 40-footers could be transported on the davits of a light cruiser. A 55-footer could deliver two 18in torpedoes (or, in theory, one 21in) and it had a substantial anti-aircraft battery of two twin Lewis guns. This model also shows four depth charges. Note the twin rudders and twin propellers and the hard-chine hull form. Not visible is the step under the steering position. (Dr Raymond Cheung)
In wartime, given their limited endurance, destroyers could operate with the fleet if they returned to base every two or three days or if action were imminent. The limit would be strain on their crews. Destroyers could be replenished at sea, particularly with oil in ordinary weather.
The great problem was identification: because destroyers were potentially so deadly, battleship doctrine was usually to fire at any that were not positively identified as friendly. The only effective means of identification was to confine them to a no-fire sector. In cruising formation that might be the rear of the fleet. With action imminent, a squadron of destroyers could be stationed on either flank of the battleships (beam or quarter), with orders not to hamper the main fleet when it changed from cruising to battle formation. The best position would depend on whether the commander wanted an early or late attack, based on weather. To reserve destroyers for attacks once the action was fully developed, destroyers should occupy unexposed positions, for example on the unengaged side of the battle line, 2000 yds off. May envisaged them passing through gaps between divisions (the Germans passed between ships) or they might attack from the head or tail of the British line. In the latter case a well-placed cruiser at one end of the enemy line could pour fire into the destroyers as they approached.
Given the new heaters, battleship torpedo range now roughly equalled effective gun range. May wondered whether it would be worthwhile to risk destroyers in a day action when the battleships themselves offered more torpedo firepower. He concluded that the role of destroyers working with a fleet was first to attack with torpedoes and only then to frustrate enemy destroyer attacks. Cruisers and scouts were better anti-destroyer weapons in a fleet context. Given long torpedo range, destroyers should never fire towards their own fleet. Once separated, they should not close the British fleet, as the battleships would fire at them. They should return to their bases after attacking. Destroyers should never stay with the battleships after dark, as the battleships would fire indiscriminately.
It followed that destroyers operating with the fleet needed heavier torpedo armament, at the least twin tubes rather than single tubes plus single stowed torpedoes. It could even be argued that guns should be traded for more torpedoes. May’s successor Admiral Callaghan argued that torpedo attack supporting the fleet was the single most important destroyer mission, hence that torpedo armament should be emphasised over guns. Like May, he considered light cruisers his best defence against enemy torpedo attack. The Board initially swatted him down. In line with earlier thinking, it held that the primary British destroyer role was to kill enemy destroyers in the North Sea while operating independently off German destroyer bases. The German fleet might take some short-endurance destroyers to sea with it, but they would have to return to port quite soon for reliefs. British destroyers off the German coast would pick them off as they came and went, much as, four decades later, NATO envisaged submarines in choke points picking off Soviet submarines as they returned to their bases for fuel and torpedoes.
Callaghan was unconvinced. In the 1913 manoeuvres destroyers were unable to find each other at night; the entire concept of hunting in the North Sea was disproved. Thus Callaghan’s 1913 fleet orders envisaged deploying his fleet destroyers on the unengaged side of his line, half ahead on the beam and other half astern on the quarter. Whichever way the enemy line approached, half would be in position to attack. For Callaghan, light cruisers were the antidote to enemy torpedo attack. Destroyers should be used against the enemy’s battle line.22 New destroyers should be armed more heavily with torpedoes, their designs placing less emphasis on guns. This position was controversial for a time, but the destroyers planned for the 1914–15 programme would have had a heavier torpedo armament and lighter guns (because of the outbreak of war, the previous year’s more conventional class was repeated).
It was generally agreed that at sea light cruisers were the best antidotes to destroyers. The original torpedo-boat destroyers had been stationed off French ports in hopes that they could sight emerging enemy torpedo boats and run them down, all the while firing at them. Although they were poor gun platforms, they were the only warships which could stay long enough with a fast target to disable it. By way of contrast, any ship opposing an enemy torpedo attack would have only a brief chance to do so. A light cruiser was a far better gun platform than a lively destroyer. That is why Jellicoe’s predecessor Admiral Callaghan preferred to emphasise the offensive role of his destroyers. Jellicoe understood why, but he never felt that he had enough light cruisers. He therefore used his destroyers, which were very much second-best as defensive assets, to make up the numbers and his tactics did not involve mass destroyer attacks against the German battle line.
Battleships could also fire ‘browning’ shots, but the British emphasised attacks by other craft within a battle fleet. The same gunnery personnel were used either for gun fire control or for torpedo control on board capital ships. During his tenure with the Grand Fleet, Jellicoe pressed for increased torpedo range, mainly in hopes of giving his battleships a better ‘browning shot’ weapon. In the spring of 1916 the British provided battleships with a 15,000 yd setting. Until the end of 1915 the settings were 10,000 yds at 28 knots and 4000 at 44 knots. Readjustment responded to reports of longer German torpedo ranges (at Jutland, however, British torpedoes considerably outranged German ones).
Pre-1914 British knowledge of evolving German tactics was limited at best; the British tended to mirror-image. As the British developed the ‘browning’ concept, they assumed the Germans would do the same. The German tactical publications which fell into British hands after the outbreak of war showed, among other things, that the Germans regarded torpedoes as far too valuable to waste in ‘browning’ shots. Presumably that was because pre-war German naval economics, shaped by the Navy Laws, had precluded providing them in great quantities.
In practice torpedoes were much less accurate and reliable than they seemed to be in pre-war exercises. The theory of ‘browning’ shots was still valid, but it took many more torpedoes to execute them. ‘Browning-shot’ tactics were never used, although they were still destroyer attack doctrine in both the Royal Navy and the US Navy at the end of the war. The British never provided enough torpedoes on board their destroyers and cruisers to make effective browning shots. A major wartime surprise was that at long ranges torpedoes were so ineffective.
Torpedoes had other unsuspected limitations. Before the war it was assumed, at least by the Royal Navy, that submarines would close to 500 yds to fire, so as to be sure of hitting. In pre-war exercises, torpedoes were set to run somewhat deep so as pass under their targets. That hid the reality that a torpedo launched by a submarine submerged to periscope depth had to climb to reach its set depth and the climb took some time, hence distance. To hit a shallow-draught ship such as a destroyer, a submarine had to back off to something more like 1000 yds. The faster the target, the better the chance that she would miss at that range. Much the same might be said of a British submarine lying in wait to torpedo a surfaced submarine passing at fairly high speed.
Capital ships turned out to be tougher than expected. A pre-dreadnought might well sink after a single torpedo hit. Dreadnoughts and their successors were more difficult to sink. The Royal Navy lost a single such ship, HMS Audacious, to an underwater hit and she took a long time to sink – so long that it seems likely that better damage control and some minor improvements would have saved her. HMS Marlborough survived a torpedo at Jutland, although it can be argued that she was lucky in where it struck. German ships proved remarkably resistant to underwater damage. Multiple torpedoes eventually did sink the battlecruiser Lützow, but she had been pounded so badly that they were only part of the problem. During the war, DNC discovered a way of protecting ships against torpedo hits by blistering, but no Grand Fleet capital ship could be taken out of service long enough for that and the speed penalty due to blistering would have been unacceptable in a battlecruiser. The only Grand Fleet capital ships properly protected against torpedo hits were the two Renowns.
As the British built more and more battleships, a key question was how to use them en masse. By the 1890s they expected to fight in line ahead, so they devoted much attention to the problem of deploying from cruising formation (columns) into the line best oriented to the enemy – preferably athwart his course (crossing his ‘T’). When the President of the US Naval War College visited his British equivalent in 1909, he observed that this was the main problem addressed in its tactical game.23 The normal fleet cruising formation was three columns, the C-in-C heading the centre column. Cruisers were normally thrown out well ahead and on the flanks and the rear. A key factor was visibility range, which in 1909 at the War College was taken as about four miles (8000 yds) by day under average North Sea and Channel conditions, about ten miles in West Indies by day and two miles at night.
Fleet manoeuvre required extensive signalling and by 1900 the Royal Navy had a very elaborate signal book. In effect it was the vocabulary an Admiral could use to manoeuvre his fleet. The role of the Admiral’s flag lieutenant was to translate his commands into the language of the signal book. There was internal debate as to whether even the best signalmen could help an Admiral control a fleet in combat. In the 1880s Admiral Tryon argued that the fleet could not effectively fight because its signalling system would fail when stressed. He pressed for a radically simpler system which amounted to ‘follow the leader’ manoeuvres. The Admiral would simply indicate whether or not the fleet was to conform to his movements. It did not help that Tryon died when his flagship HMS Victoria was rammed while the Mediterranean Fleet manoeuvred to anchor in 1893. Contrary to his teaching, Tryon was attempting a particularly elaborate, if elegant, manoeuvre and he seems to have miscalculated the distance between columns of ships. Jellicoe, at that time a Commander, narrowly avoided drowning as Victoria sank. Presumably he was being given a particularly graphic lesson in the dangers of complex manoeuvring.
The British came to understand that the simpler the formation, the better the chance that it would work – in more modern terms, ‘keep it simple, stupid’. The simplest and most supple formation, it turned out, was line ahead, because maintaining it required little signalling. Although the Signal Book remained as elaborate as ever, line ahead was very much the kind of tactic Tryon envisaged. Line ahead had a long history under sail, but with the advent of steam navies tried many alternatives, including line abreast and line of bearing. Under sail, line ahead made for simplicity, but given short gun range it was impossible to concentrate fire on any one enemy ship. Each side in a line-ahead battle would deliver roughly the same weight of fire. Nelson broke the enemy line at Trafalgar specifically so that he could concentrate fire on particular enemy ships, breaking up their formation. He accepted the risk that they might be cut off and overwhelmed by an alert enemy. Was there a battleship-age equivalent to Nelson’s tactic?
Having tried alternatives, by the 1890s the Royal Navy returned to line-ahead close-order tactics, a position not shared by some other major navies. The British found that line-ahead tactics much simplified station-keeping and the line was less liable to be thrown into disorder. In any other formation, course could not safely be altered without any signal – at a time when signalling was likely to be slow and uncertain, if not impossible. Perhaps as importantly, a line-ahead formation seemed to solve the problem of distinguishing friend from foe. Anyone in line was a friend, anyone outside a foe. Problems would arise only if the line kinked (as during the night action off Guadalcanal in 1942), so that some captains thought they were watching an enemy breaking through. That was unlikely in daylight.
In 1914 the heavy gun was the dominant weapon, but it was assumed that its effect would be cumulative. Gunfire would pound ships but only rarely sink them. The British liked Lyddite shells because it appeared that the fumes they generated would incapacitate crews. No one imagined sudden explosions like those which sank three British battlecruisers at Jutland. Slow destruction, as of Graf Spee’s big cruisers at Coronel, or Blücher at Dogger Bank, conformed much more closely to expectation. Thus fleet tactics envisaged night attacks by destroyers after a day gunnery action, largely to finish off cripples. This fleet battle practice target gives some idea of the area over which shells were expected to hit (shells falling outside were not counted as hits). At about 8000 yds the Royal Navy expected better than 50 per cent hits and even then it did not expect to knock out modern battleships very quickly. This target was photographed on 29 September 1912. (Dr David Stevens, SPC-A).
This was logic, not fetishistic devotion to the traditions of the age of sail. When he conducted tactical experiments in 1910–11, Admiral May sought an alternative to line ahead, which he considered basically defensive. In his view a fleet in rigid line-ahead formation could not readily counter-attack. The alternative divisional organisation was inherently flexible and, it seemed, relatively easy to command. May saw it as inherently offensive. Also, it was far easier to exploit ships’ speeds when they were organised in divisions (no more than eight ships, preferably four). From a theoretical point of view (Admiral May’s phrase) divisional tactics were extremely attractive.24
The rub was what Togo had found at Tsushima: individual divisions had to be kept under sufficient control that they reinforced each other but did not accidentally fire at each other. The fleet C-in-C had to know a lot more about what was happening in the battle in order to do that. Devolving control to the commanders of the divisions was risky in a world filled with smoke. Jellicoe’s experience of tactical experiments in 1912 seems to have convinced him that it was too easy for the divisions to tangle and even to engage each other instead of the enemy. These exercises revealed another problem as well. The British believed that a modern battleship could survive under fire for a considerable time. In the 1912 manoeuvres enemy ships were not exposed to fire for long enough. The fleet needed some way of laying down sustained fire for much longer. To do that it had to remain in formation, shelling an enemy fleet which would presumably also remain in formation.
Jellicoe, and probably other British officers, came to see rigid control as insurance against a dangerous melee punctuated by torpedo hits. When he came to write his Grand Fleet Battle Orders, Jellicoe abandoned his earlier acceptance of decentralised command once battle began. He sought to maintain control throughout. Divisional tactics might be the best way to concentrate fire on parts of the enemy’s fleet, hence to achieve results in a reasonable length of time. Jellicoe later said that he would have adopted divisional tactics if he could have, but that he did not have enough time to train his fleet in them.
Divisional tactics were not quite the same as accepting that once battle began a C-in-C might no longer be able to exert full control and much would have to be done by those commanding the fleet’s divisions. That begged a question. Who would control fleet resources such as massed destroyers? Who would direct them so that they attacked only the enemy’s fleet and not a hapless division of one’s own fleet?
Deploying across the enemy’s path was the best possibility. The worst, which seemed entirely possible, was that the British and German fleets would approach head-on, both fleets being in line ahead. In that case the battle might well open with a salvo of German torpedoes fired towards the British fleet. Because the British would be running towards the torpedoes, their effective range would be increased considerably. A quick run past the Germans on parallel courses would never provide enough time for gunfire to be effective. Jellicoe believed that it also presented dangers. He wanted to get onto a parallel course steaming in the same direction, because only then would he have enough time to pour enough shells into the German ships. Having run past the Germans, he could turn behind them and pass through their wake. However, Jellicoe was aware of the Japanese use of floating mines and he had sponsored its British equivalent. What if the Germans had the same idea? The wake of the German fleet was the last place he wanted to be. What sort of manoeuvre would bring his battle line onto the desired parallel course, in the same direction, without exposing it to floating mines?
Jellicoe’s predecessor Admiral Callaghan, who championed offensive destroyer operations in a fleet context, wrote in his March 1914 battle orders that the tactics of a fleet consisting of three or more battle squadrons plus battlecruisers and many other ships were fundamentally different from those of a fleet of one or two battle squadrons.25 Probably a single officer could command the smaller fleet for much longer than a larger one. Surely the larger the fleet, the greater the need to decentralise. Callaghan thought that he could control approach and deployment, but once firing began he would have to devolve to squadron commanders, subject to general instructions. He hoped to be able to deploy quickly enough to establish superiority of fire. The objective would normally be the enemy’s rear, a departure from the earlier idea of deploying across the enemy’s ‘T’. Callaghan cautioned against ships being drawn off by an enemy fast division, which could break up his line. Presumably he had the German battlecruisers in mind.
Overall, the British tactical problem was simple to state but difficult to solve. If the British fleet were far more numerous than its German enemy, how could its apparently crushing numerical superiority be translated into tactical superiority? How could targets in the enemy battle line be designated so that all were covered and how could multiple ships concentrate their fire on the enemy? The more numerous the fleet, the longer the line. A very long line steaming more or less parallel to a shorter enemy line would overlap it, the ships at the end too far from the enemy to fire. British experiments in longer-range fire may have been intended specifically to solve the tactical problem presented by the ships at the ends of the British line. Breaking up the fleet into mutually supporting units (‘divisional tactics’) could solve the geometric problem, but not the tactical one.
The Germans also planned to fight in line ahead, but with important differences from British practice.26 Draft tactical orders the British obtained in 1914 included the ‘Gefechtswendung’ (Action [Battle] turn) of the whole battle line through 16 points (180°). It was the first item described under the category ‘turning together in action’, alternatives being available if it was impracticable (as stated in the book).27 The parallel Tactical Order No. 1 (Hints for Battle) included first the injunction that the enemy might turn together to escape; ‘it would be unfavourable for us to pursue him, as we should then run into the area commanded by the enemy’s torpedoes, without being able ourselves to make use of them’. The Germans would have to use the same manoeuvre to extricate themselves, even if that took them out of range. The manoeuvre might have to be repeated until the German fleet was once more in range. ‘It would be advantageous for us, if we could make the enemy follow us, so as to get him into the area commanded by our torpedoes’ (italics in the original). Using this manoeuvre, a German fleet being pursued by the British could suddenly turn on its pursuer at a high speed of approach. The resulting high range rates would make gunnery difficult for both fleets, but the Germans could also fire a salvo of torpedoes towards the approaching German fleet. Effective torpedo range would be considerably increased, because the British would be running into the German torpedoes.
A British captain raised exactly this problem in May 1914, before the British had the German manoeuvring orders. It later turned out that the British Naval War College had become interested in exactly this kind of manoeuvre, as a way of subjecting the British fleet to mines and U-boats. At the end of the war the British were still trying to find a tactical solution. A British fleet pursuing the fleeing Germans would find itself running towards German torpedoes, whereas the Germans would be opening the range British torpedoes would have to cover.
The 1914 German orders emphasised the lack of modern light cruisers for scouting. In cruising formation, the light cruisers, perhaps supplemented by destroyers, would form an arc of a circle about 25 miles ahead of the main body, a distance chosen so that the enemy could not get around the scouting line but also far enough away that the enemy outside the scouting line could not see the main German fleet. The German battlecruisers of the 1st Scouting Group would be concentrated between the main fleet and the scouts, about 10 miles back, so that they could back up the scouts. The orders allowed for the possibility that the position and direction of the enemy’s scouting line could be determined by aerial reconnaissance, in which case cruisers could be concentrated to penetrate it.
Once the enemy had been found, the battlecruisers would fall back to a position at the head of the German line formed out of the cruising columns of battleships. Its two tasks, presumably in order of importance, were to fire torpedoes at the head of the enemy line and to deal with the enemy’s fast division (his battlecruisers) at the head of his line. When the fleet formed up, the battlecruisers would constitute its fast division operating semi-independently. To avoid being surrounded, it was to manoeuvre at high speed; an action between the fast divisions on both sides might develop. It was accepted that it would be very difficult for the fast division to reach the assigned position. It needed the earliest possible information as to the direction in which the main fleet was to deploy. In another departure from British practice, the 4th Battle Squadron (Wittelsbach class pre-dreadnoughts) would form about 5000 to 6000 m on the disengaged beam of the battle line (consisting of the other three squadrons), available to reinforce it as needed.28 This was a sort of divisional tactics.
If there were no doubt as to the direction in which the action would develop, all flotillas would try to steam ahead as far as possible. If they could do so without obviously manoeuvring to attack, the foremost destroyers should attack with long-range torpedoes. The main destroyer role during a day battle would be massed attack on the van and line of the enemy. Where the British were interested in attacks which would spoil enemy deployment, the Germans wanted their destroyers to wait until the two battle lines had formed and were punishing each other. Where the British feared that the Germans would attack specifically to ruin their fire control, the German instructions emphasised the way in which fire control requirements would prevent an enemy from manoeuvring to avoid torpedoes. If the enemy refused to maintain formation in the face of torpedo attack, at the least the destroyers would be reducing the weight of fire on their battleships. Perhaps the key phrase was that ‘every officer in command of a destroyer must remember that his torpedoes must put at least one of the enemy’s ships out of action . . . the probability of a hit is the only thing to justify the expenditure of a torpedo. A destroyer which has fired all its torpedoes without obtaining a hit has not fulfilled its purpose in the action’. Torpedoes should not be fired at excessive range, a safety margin of at least 25 to 30 per cent of range being allowed. The instructions cautioned that ‘great care and exact knowledge about the positions of our own ships’ (my italics) was a prerequisite for angled shots.
There was no reason to open fire beyond 12,000m (13,200 yds) unless the enemy forced that. Nor was there any reason to concentrate fire, as ‘the advantage of engaging all the ships of the enemy is out of all proportion to the loss of effective fire through its dispersion’. Given the ‘destructive effect of underwater hits at very great ranges with [HE] shell . . . these should be used as soon as the range is found, as well as AP’. As in the Royal Navy, there was a severe injunction against wasting ammunition, the implication being that if the enemy kept the range wide, the medium-calibre guns might not be used at all.
As the British watched the Germans and as they obtained German documents, they began to feel some insight into German thinking. In October 1914 Admiral Jellicoe circulated a German report of gunnery practice by the 1st Battle Squadron of the High Seas Fleet. He particularly noted that the Germans intended to develop a high rate of fire when they found the range, that they were achieving a very small spread with their heavy guns and finally that, as he thought, they expected to fight at limited range – since the secondary battery (5.9in guns) took part, it seemed that they would fight within 8000 yds. In at least two cases visibility was low. He was relieved that there was no sign of a director system.29
Dreadnought battleships were conceived by the Royal Navy as a way of using guns effectively at greater ranges, initially outside the torpedo ranges of battleships. This is the forward superstructure of HMS Colossus as built. The external sign of increasing gunnery sophistication is the big spotting top. When Colossus was being designed, fleet commander Vice Admiral Sir Francis Bridgman argued that masts themselves were a danger. It would be better to use an armoured position near the conning tower. DNO Captain Reginald Bacon replied that aloft control would be particularly important during the approach phase of an action. Once the action became general rangefinding and formal control would no longer be very important. It followed that a single aloft position would be vital, but that there was not much point in providing a second such position aft, where it would be smoked out. Since they did not need long radio receiving range, battleships did not need a mainmast to support high antennas (battlecruisers did). HMS Colossus therefore revived the single-mast arrangement of HMS Dreadnought, which had been discarded. A single mast also had the virtue that it made determination of the ship’s course by an enemy more difficult. The single mast was placed abaft the forefunnel so that its vertical member cold support the heavy boat crane (ships with two masts used the vertical member of the mainmast). Unfortunately the spotting top, which clearly suffered from smoke interference, became more rather than less important as range rapidly increased. The Orion and Lion classes were designed with similar masts before Colossus went to sea and demonstrated the smoke problem; the Lions had to be rebuilt at considerable cost. (Author’s collection)
On 30 October Jellicoe sent the Admiralty his understanding of how he would have to fight, based on what he had learned so far about the Germans.30 He felt that the Germans had shown that they planned to rely heavily on submarines, mines and torpedoes, presumably based on German submarine and mining operations to date. Surely they would use the same weapons in a fleet action, ‘especially since they possess an actual superiority over us in these particular directions’. It seemed to him that the Germans could not be sure of having all of their submarines and minelayers with them unless they could choose where to fight, probably in the southern North Sea where they could also have air support. He therefore planned to fight in the northern North Sea, which incidentally was closer to his own base, giving his wounded ships a better chance of getting home. That was also where he could be sure of concentrating his own short-legged destroyers with his cruisers. Given the need to concentrate torpedo craft, the Germans would not come out except at a moment of their choice, when all these craft were ready. He, on the other hand, always had to be ready and his cruisers were often using up their coal. Yet he needed a large cruiser force to scout and to screen his battle force, ‘so that the latter may be manoeuvred into any desired position behind the cruiser screen’. That also favoured the northern North Sea.
Jellicoe was sure the Germans would use their submarines with their battle fleet. The cruisers might lead them into position so that the High Seas Fleet could lead him over them. The basic safeguard was to manoeuvre the Grand Fleet at high speed so as to upset essentially fixed U-boat dispositions. Given their limited underwater endurance and speed, the submarines could not follow unless they surfaced; after an interval of high-speed manoeuvring, Jellicoe felt that he could safely close the High Seas Fleet. The situation was grim: German underwater weapons might disable half his fleet before he could open fire ‘if a false move is made’. He might have to refuse to follow a turn-away by the High Seas Fleet and he was anxious that the Admiralty understand the situation and not see a turn-away on his part as refusal to fight, which would rightly be condemned.
Information from a German lieutenant captured at Heligoland seemed to confirm Jellicoe’s belief that the Germans would mass their torpedo craft in the closest possible formation to ‘brown’ the British. If they went to sea,
they would take with them the largest possible number of torpedo craft, which would attack from between their ships in threes or as many as possible in close formation, pushing their attacks to within two or three thousand yards. They are quite sure that our torpedo craft bunched at the head of the line will not stop them, nor will the fire of our ships, as they say it takes many hits to sink a torpedo boat destroyer (which seems proved). Apparently they place so much importance on this that they will not now risk their new boats and officers and men are instructed to push their attack home or die. They believe that their big destroyer attack in a fleet action will put half our ships at least hors de combat. 31
In the wake of Jutland, Jellicoe modified his battle orders. Where he had initially assumed that the Germans wanted to fight at short range, now he believed that they would keep their heavy ships either out of range or at long range, relying on their torpedo craft (destroyers), He thought that such tactics were facilitated by the high speed of some of the German battleships.32 This was a grossly inaccurate reading of the battle, in which the only mass torpedo attacks were mounted as part of Scheer’s effort to extricate his fleet. It shows just how obsessed Jellicoe was with the underwater threat.
During the war the basic battle hints were supplemented by other tactical orders, so that by the end of the war the list had grown to twenty-four and the original hints had been replaced by a report of methods of screening the fleet dated 9 April 1918.33 The most striking feature of this instruction was that, instead of concentrating on the area ahead of the fleet, the screen should normally completely surround the fleet. The pre-war idea had been that the German fleet would have the initiative, steaming towards an oncoming enemy. War experience showed that the enemy might well be able to get around to the fleet’s rear. As ASW screens, destroyers had to cover both the front of the fleet and both its flanks. Again, this was very different from pre-war thinking.
After Jutland, it was generally agreed that the Germans had invested much more effort in night fighting, which meant much better techniques for recognition signals. Remarkably, their 18 October 1915 tactical orders for such signals were still in force at the end of the war.34 In theory all such systems relied on a challenge and answer, but by this time the Germans had learned that challenges by a force or ship approaching one already in action were often unanswered. At the discretion of the CO, each ship would, without waiting to be challenged, continually show the new challenge as long as he was burning his searchlights or firing. The ‘challenge number’ would be shown using the forward NSA (Night Signal Apparatus – a designation which suggests an IR device). If an approaching force was unable to make out the NSA before joining the action, it might make the signal FSL (‘make recognition signal by searchlight or Morse lamp’). Other emergency signals were FTB (show the ‘new challenge’ continuously) and FTC (show the ‘new challenge’ until further orders). There was also an emergency signal using a searchlight pointed up, in case a ship was being fired on by friendly forces.
As fleets grew larger, an important issue was how to concentrate the firepower of multiple ships against single enemy ships. To do that, ships needed some means of passing range data. The Royal Navy introduced range drums, as shown here on board the battlecruiser HMS Inflexible, visiting New York in 1910. The drum is the horizontal cylinder visible on the fore side of the foretop. These drums were widely distributed in the pre-war Royal Navy.
The corresponding British night recognition signal was two green lights 2m apart horizontally on the starboard yardarm and two red lights on the port yardarm. There was also a two- (later three-) letter recognition signal made by searchlight or flashing lamp, changing at midnight local time, as ordered at the end of March 1915 (the Germans concluded that to deceive the British they should use a badly-manipulated searchlight sending out a two-letter challenge in hopes that no special care would be taken to make sure it was the right challenge). By 1917 typical night recognition lights were red and white horizontal lights on the foremast. By day ships relied on flags. For example by order of 16 November 1914, to avoid confusion due to the similarity between British and German naval ensigns, British ships were also to show at least one red ensign in a conspicuous position. Destroyers and torpedo boats generally flew the Red Ensign at the yardarm in such a place as not to interfere with signal flags.
The Radio Revolution35
Looming over the visible changes to navies was a subtler but perhaps even more important one: wireless (radio), in both strategic and tactical aspects. Radio systems were described as either wireless telegraphy (dot-and-dash transmission of individual letters and numbers: W/T) or radio telephony (voice: R/T). Radio was first demonstrated (over a very short distance) by the German physicist Heinrich Hertz in 1887. Navies quickly understood its potential: for the first time it might be possible for ships to communicate beyond the horizon. Wireless was also the only way for a command centre ashore to communicate instantly with ships at sea. Underwater (acoustic) signals, which many navies adopted about 1906–7, was a short-range competitor, but only wireless offered really long range. By 1901 the Royal Navy was testing short-range tactical sets on board ships.
Already very interested, the Royal Navy learned what wireless could do when it was used to gather destroyers in fog during the 1904 Torpedo Manoeuvres. Wireless was vital to Admiral Togo at Tsushima in 1905. His deployment was keyed to the Russian choice of which side of the Straits through which to pass – which was transmitted to him by W/T from a Japanese scout cruiser. On the other hand, wireless did not help Togo maintain control of his two-division fleet once the battle began.
Togo’s use of wireless exemplified a radical change in scouting. Now a scouting screen could be maintained without the large numbers of repeating ships formerly needed to transmit what the scouts saw back to the main body of the fleet. Once scouts could report back from beyond the horizon, there was a different issue. They had to know (and to report) where they were, so that the command to which they reported could turn their information into something useful. That was not the case for a scout reporting visually, since those on board a flagship could see where the scout was. Incomplete reporting (no position) was such a problem during the action off Scarborough in December 1914 that Admiral Jellicoe had to send his fleet a message emphasizing the need for position reports.
During the early part of the twentieth century the Royal Navy, the largest in the world, led the world’s navies in W/T installations and technology, although the German navy and the US Navy were rapidly catching up.36 During the war, German sets (using tubes) were considered superior technically, but it appears that the British approach to wireless within the fleet was substantially more sophisticated, having been developed through years of large-scale fleet exercises. Moreover, W/T was integral to the new concept of centrally-controlled fleet operations developed by Admiral Fisher. When he took office in 1904 the British already controlled much of the world telegraph network. Fisher realised that by adding long-range radio stations to key telegraph stations he could produce just the global command net he wanted. It took cables and telegraph to send a message halfway around the world to a distant station, after which long-range radio could take over.
Inflexible is shown at full speed, her coal-fired boilers pouring out smoke which makes her after spotting top nearly useless. During the battle of the Falklands, her spotters suffered from vibration despite the support offered by the heavy tripod masts, and at times they were badly smoked. Note the director on the platform below the foretop, and the spiral surrounding the fore topmast. It was an anti-rangefinding measure, intended to break up the visible vertical mark represented by that mast. The British were unaware that such measures were useless against the stereo rangefinders the Germans used. Although often captioned as a photo of the ship at the Falklands, features such as the extended bridge and the director indicate that it was taken later.
Among Fisher’s first investments as First Sea Lord was a Navy-owned shore wireless network, whose main stations (Cleethorpes and Horsea in England and Malta and Gibraltar in the Mediterranean) could transmit to ships 500 miles out at sea. All were connected to the Admiralty by cable. Wireless was so important that, despite this investment, Fisher bought new stations in 1908. Cleethorpes and Horsea were upgraded to a range of over 1000 miles. A powerful station was installed at the Admiralty. New transmitters were ordered for British warships. The meaning of this combination was not lost on the Germans. In 1914 the German cruiser Emden was destroyed while her crew tried to disable a British wireless and cable station in the Pacific. Conversely, British attacks on German cables drastically reduced the ability of the German Admiralty to contact distant forces – such as Admiral von Spee’s Asiatic Squadron, once it left the German base at Tsingtao in China. Radio was also developed for use within a fleet, although visual signalling was preferred. To that end the Royal Navy developed techniques it hoped would limit the reach of tactical sets. By 1914 it had a considerable variety of sets in service.37
Having gained considerable experience with radio, the Royal Navy was well aware of the potential advantages of an enemy intercepting its signals. One recommendation after a major fleet exercise testing radio jamming was that the navy should adopt multiple special wartime frequencies, not to be used or disclosed in peacetime, as well as a system of wartime call signs, again not to be used in peacetime.38 A special radio frequency (tune) was assigned to the Red Fleet in the major exercise later in 1906. By 1907 planned W/T watches were one to operate the set in the silent cabinet, the other to cipher/decipher. All ciphering was by book code, but the Royal Navy was seeking a mechanical device.39
Wartime W/T organisation placed the Admiralty at the head of the entire shore system of the British Isles and the Mediterranean, with each deployed fleet focussed on the Admiral afloat.40 At least in theory, the Admiralty transmitted to the Admirals, who commanded their fleets; there was no option for the Admiralty to transmit direct to any subordinate. That created a problem in with the Grand Fleet, as Admiral Jellicoe was C-in-C Home Fleets, commanding widely-dispersed forces. After Beatty’s battlecruisers were moved to Rosyth, Beatty was given, in effect, a separate line from the Admiralty. Separated ships would not normally communicate with each other directly. Instead, a ship would normally communicate with the nearest low or medium-power shore station, which in turn would communicate with the shore station nearest the addressee. A fleet had a designated W/T guardship intended to listen for Admiralty messages addressed to the fleet commander. Within a fleet, there was also a dedicated Destroyer Wave. It might be used directly, or the depot ship or flotilla cruiser (later, leader) could use the Destroyer Wave to transmit messages on the Admiral’s wave. For example, flotilla cruisers (leaders) would listen on the Admiral’s wave for five minutes at a time at quarter past and quarter to every hour.
The Admiral of a fleet or squadron was assigned his own frequency, which was ‘guarded’ (listened for) by the senior officer of each group. It was also guarded by the Second in Command. The main scouting group would keep watch on (and use) the Admiral’s wave. Ships within the Admiral’s force were broken down into groups, each with its own senior officer (plus not more than eight ships). That included all but the most important cruisers. Principal Grand Fleet squadrons all functioned as groups in this sense. Each group kept a guard on the Admiral’s (or other senior officer’s) frequency, retransmitting on the group’s own frequency (which designated the group, frequencies being identified by letter). An Inter-Group Guard ship kept watch on the Group Wave (group frequency) and collected messages from ships in her group intended for the senior officer and for ships in other groups, retransmitting as required. A separate Group Repeating Ship read all messages sent by the Group Guard and kept track of any failures to answer the Group Guard’s messages. Whenever possible, communication between Senior Officer and Group Duty ships should be by auxiliary (low-power) radio, to limit the enemy’s ability to intercept signals. The point of the entire scheme was that there were so few available frequencies, as the spark transmitters in use were fairly broad-band.
Photographed in 1918, Inflexible shows how pre-war gunnery ideas played out in wartime. The most dramatic visible change was greatly increased range, which explains the big rangefinder atop her foretop. The platform below the top, which before the war accommodated a searchlight, now carries a cylindrical director to co-ordinate the fire of the ship’s turrets. Concentration fire, abandoned in 1914, returned in 1917. The front face of the foretop carries a range clock, which displays the range at which the ship is firing (in effect it replaced the earlier range drum). The ship’s foreturret carries bearing markings so that other ships can fire at the same target. The smaller rangefinder atop the bridge is for plotting and also for long-range torpedo control. The structure below the open bridge is a charthouse and plot. Another wartime feature is the big life raft carried abaft the bridge structure. Such rafts could float off a sinking ship, saving lives even if boats could not be launched (or did not survive a battle). The 4in anti-destroyer guns initially atop ‘A’ turret were moved into the superstructure when it became clear that destroyers would join in a day action. Late in the war that made the top of ‘A’ turret available for the anti-aircraft gun shown. The tripod foremast of the light cruiser in the background shows multiple range clocks, because other cruisers co-operating with her might not be in line ahead, hence might not be able to see a clock oriented fore and aft.
Special provision was made for radio silence. The Wireless Instructions referred to the possibility that the distance to a ship could be estimated from the strength of her signals, but the issue was the more general possibility of interception (it turned out that signal strength could be deceptive). A fleet commander could order all of his ships to continue W/T watch but would not answer any messages.
Wireless works because a fraction of any varying electrical signal coupled to an antenna will radiate into space, generally in all directions. The lower the frequency, the further it goes, with some qualifications (at particular high frequencies signals bounce off the ionosphere and can travel around the world). Frequency means wavelength: the lower the frequency, the longer the wavelength. How much of the signal goes out into space depends on the antenna (gain), which in turn depends on the size of the antenna in wavelengths. That is why the US Navy needed tens of miles of antenna to transmit the extremely low frequency (ELF) signals it used to communicate with submarines in the 1980s. Before the First World War signals were usually designated by wavelength, typical Royal Navy waves being 400ft or 1025ft or 2760ft or 4250ft long (in modern terms 2.5 MHz, 950 kHz 350 kHz and 280 kHz meaning thousands of cycles per second).41 These frequencies were chosen because they were the lowest that could be picked up by shipboard antennas of reasonable size, it being difficult to generate high-frequency signals.
The Germans did not initially use spotting tops, because they assumed gun range would be short. The top (with a more rigid tubular foremast) was fitted to Friedrich der Grosse in 1918; she is shown at Scapa Flow after the German fleet surrender. The Königs were similarly fitted in 1918 (but Kronprinz Wilhelm may have had a tubular foremast and gun control top on commissioning in 1914 and some battlecruisers had foremast spotting tops in 1914). The upper top, with rangefinder, was for main battery control. Below it was a spacious torpedo control level. German director control was simpler than British because the Germans relied on gyros at the guns to fire them at the right time in the ship’s roll. Their director was really a pointer which co-ordinated the bearings of the guns. That suggests no interest in counteracting the pronounced effect of roll on gun bearing, particularly on bearings near dead ahead and astern. After the ship was surrendered, her officers said that the foremast vibrated, even at anchor in harbour. It is not clear why tripod legs were not added. The main battery was usually controlled from a position above and abaft the conning tower. In contrast to the British, the Germans had no interest in trying to conceal a ship’s course either in day (by using a single mast) or at night (by concentrating searchlights more or less amidships with remote control). (Naval Institute collection)
Like its wired equivalent, wireless started with a signal, a long (dash) or short (dot) pulse or, in a more sophisticated version, a voice signal. None of these signals worked at the kind of frequencies a shipboard antenna could transmit. The antenna sent, for example, a short burst of high-enough frequency electrical noise. To create that, a shipboard wireless set had to turn on a source of such noise for a short time (equivalent to a dash or dot). In early systems that was created by a series of sparks. Each spark mixed components with a wide variety of frequencies.42 The sending antenna filtered out most of them; to a limited extent it was tuned. As might be imagined, most of the energy involved was at the wrong frequencies; it was wasted. It took a lot of spark to create a relatively weak signal. The signal was, moreover, mixed with background noise. At radio frequencies the world is not at all silent. The signal spreads out from the sending antenna, so that only a small fraction reaches the receiving antenna. Like the sending antenna, the receiving antenna chokes out all but a narrow range of frequencies. It too has a gain.43
The Royal Navy continued to use spark radios during the war. They were robust and durable, easily repaired and their broadband signals could force through interference. Against that, they were inefficient, their broadband signals caused serious interference (which was particularly evident at Jutland, when many ships attempted to transmit more or less simultaneously) and they and their antennas required considerable insulation. A late-war account of submarine W/T transmission, for example, mentions that if the insulation on a mast was ruined, the spark could not go up the mast into the aerial.
The alternative to a spark transmitter was a means of creating a continuous wave at a more or less pure frequency.44 The first such device used by the Royal Navy was the Poulsen Arc, which offered by far the longest-range performance before and during the First World War. The alternative way to produce a continuous wave was a vacuum tube transmitter. Because its signal was a clear tone, the Poulsen could not be heard by a conventional receiver; it needed a special device.45 In 1914 the Royal Navy installed Poulsen sets on board several cruisers, to function as linking ships between long-range stations at home and deployed ships with shorter-range spark radios.46
By 1907 the Poulsen Arc had largely been perfected and many US battleships had it. It produced a narrow-band continuous-wave signal, which could be applied directly to an aerial. Dots and dashes were applied to the continuous wave as a series of tones (which could not be heard as such in First World War receivers).47 Arc transmitters were also liked for their robustness and durability and they could easily be built to produce very powerful signals. However, they were slow to start, as the arc had to form. The transmitter had to be monitored to keep to the desired frequency. Maximum frequency was limited (to 250 kHz). The transmitter produced not only the desired frequency but also harmonics (multiples of the basic frequency). Arc transmitters were considered unsuitable for fleet use because it was impossible to ‘listen through’ to hear the Admiral’s signal.
Vacuum tubes were introduced for a variety of roles, beginning with diode detectors, but the most important type was the triode, the basis for amplifiers and for transmitters.48 Tubes started quickly and they were easier to key than an arc transmitter. Against that, they were fragile and it was more difficult to trouble-shoot vacuum tube sets. The British tested their first tube in 1908, but as of 1914 they had not adopted tubes.49 The Germans adopted tube radios, which offered much cleaner (less interfering) signals than wartime British spark transmitters. After the war the British adopted tube radios,50 while during it they used vacuum tube amplifiers to detect German radio signals at long range.51
Typical shipboard antennas were vertical wires connected at one end to a ground (such as the metal in a ship) and at the other to a horizontal element, in either a T- or an inverted-L or an inverted V-pattern. Performance depended on how long the vertical and horizontal elements were. Typically the horizontal part consisted of parallel wires whose total length was a larger fraction of the wavelength to be used. British practice was to create a cylinder of parallel wires, each insulated from the others. Other navies used multiple wires side by side. Relatively low (by later standards) frequencies made for either long wires or inefficient antennas coupled to higher-powered or more efficient transmitters. The first British battlecruisers had pairs of high masts specifically to attain the highest possible wireless performance, in this case reception rather than transmission. That had tactical consequences. High masts could be seen further beyond the horizon. Paired masts gave away a ship’s inclination, the angle between her course and an enemy’s. During the First World War, the Royal Navy found itself cutting down topmasts and wherever possible eliminating one of the earlier pair of masts altogether.52
Whatever fraction of the original signal which survives the filters represented by the sending and receiving antennas and dissipation due to distance has to be recognised against the noisy background. That means converting something varying thousands of times a second into a dot or dash (or, later, voice) which an operator could use. That was called rectification (stripping away the radio-frequency oscillation). The earliest device to do that was the coherer, a glass tube with metal filings inside. When a signal passed through it, the filings lined up visibly. The lined-up filings of the coherer reduced the wireless signal to a current which could be detected.53 Later more reliable rectifiers were fed into earphones. Thus the dot or dash in the signal became a sound which an operator could use the same way a telegraph operator used the tap of the key. Better rectifiers were soon invented (the coherer had to be tapped to recover after each signal, which seriously limited it). Rectification was not perfect, however, so the same earphones picked up the noise (‘atmospherics’) accompanying the radio signal. Often the brain of the operator was the key component which could pick dots and dashes out of the whistles of static. Much the same could later be said of underwater sound.
During the war the Royal Navy continued to rely on wireless for longer-range messages, but to a considerable extent on flags within fleets and formations.54 Flag hoists offered a considerable command vocabulary and answering hoists made it clear that signals had been received. The system placed signallers (officers) in an important position: they generally decided how to convey a commander’s message. To do that, a signal lieutenant had to have some idea of what sorts of commands would be sent and therefore an education in fleet tactics. Conversely, a signal lieutenant who did not have the full vocabulary at his command could not express his commander’s intentions properly. That seems to have been the case with Admiral Beatty’s flag lieutenant Ralph Seymour in several battles. The standard signal book included a discussion of how a fleet should function in battle; there is some question as to whether it was updated as British tactics changed radically in the run-up to 1914. The Germans never produced so elaborate a vocabulary.
After Jutland, among the many fleet committees formed to report the lessons of the battle were a W/T Committee and a separate committee intended specifically to evaluate interference from German ships and the potential of radio direction-finding at sea.55 Given accounts of widespread self-interference (‘pandemonium’ in some accounts), it is interesting that the committee report mentioned none at all; it was concerned more with ensuring that aerials survived in battle. Early in the action the Germans tried to jam British signals, but ‘not the slightest difficulty was experienced, the interference being almost completely cut out by slight alteration in tuning’. Attempts to jam Type 3 sets failed because the German jamming note was so high that loud signals from ships in company could easily be read through by an experienced operator. Admiral Jellicoe summarised the report: he wanted ships rigged for dual (frequency) reception by their main radio stations, instead of being limited to a single frequency. Priority should go to all flagships and then to the W/T Guard Ships in the battle squadrons. A few flagships were already fitted. The fleet flagship Iron Duke had three separate W/T (silent) cabinets (each with a receiving operator) and typically maintained lookouts in two of them, thus avoiding considerable use of auxiliary W/T and providing the Admiral with information on two or three lines without delays entailed by re-transmission from Guard Ships.
The great problem in the North Sea, understood by the Royal Navy but not by the German navy, was ocean surveillance: knowing when the enemy fleet was at sea and where it was. On the eve of war, the Germans had by far the greatest potential to solve the problem in the form of their airships (mainly Zeppelins), but they considered airships essentially tactical scouts tied to their fleet, substitutes for the light cruisers they lacked. There was no systematic search of the North Sea. Perhaps because it lacked effective airships, the Royal Navy was acutely aware of the need for ocean surveillance, in its case mainly achieved by exploiting German wireless messages. L 30 was the first of a class of ten delivered in 1916. They were 643ft long and 78–75ft in maximum diameter, powered by six 220hp Maybach engines, the exhaust from one of which is visible. They drove two pusher propellers and two propellers on ‘wings’ built out from the airship’s body. Full speed was 60mph and maximum altitude was 14,000ft. Four cars were suspended under the envelope, the forward one being divided in two. Payload was 27 tons, far beyond anything possible with a contemporary aircraft. (Philip Jarrett)
The most interesting innovation on the British side was the prototype radio direction-finder in HMS Lion, which had not yet been calibrated properly. Its operator, a Royal Marine lieutenant, was confident that once calibrated properly it could be used to plot the bearings of major enemy units. This D/F set was the first effective naval over-the-horizon sensor in the world. This device employed vacuum tubes for sensitivity and for precise tuning. Performance apparently depended very much on the operator’s skill. Work was ongoing to reduce interference from British auxiliary sets, which operated at much the frequency of the German ones. The final report to the Admiralty pressed for a D/F set capable of receiving at ranges up to 50 miles, with an aerial which could be used without interference from the main battery and with a particularly sensitive receiver for wavelengths between 2000 and 3500ft. It would be ‘of extreme value in locating the enemy and also our own ships in thick weather or at night’. All D/F gear tried so far required an aerial so large that it could not be kept up once guns began to fire. The second D/F system was fitted on board HMS Princess Royal during her post-battle refit (the post-Jutland file includes a report of initial operations dated 12 October 1916: average bearing error was 15° on Q wave). An extemporised set was installed on board HMAS Australia (completed November 1916). In December, Admiral Beatty asked for similar sets on board Barham and Warspite, having found the Princess Royal set extremely useful (he soon added his new flagship Queen Elizabeth). He also endorsed a request by 2nd Cruiser Squadron for a set on board HMS Minotaur for test purposes. In the latter case, the set was valued both for detecting and fixing enemy ships and for navigation using German shore stations. When the United States entered the war, the US Navy developed its own much more compact shipboard radio direction-finder, which was used against U-boats (the US Navy seems to have been unaware of the British capital ship set; the British thought it impractical to build a destroyer radio direction-finder).56
In 1903 the Germans formed Telefunken, which made their wartime W/T sets.57 Telefunken used alternators to produce distinctive high notes as its dots and dashes and during the war British operators reportedly considered its signals much cleaner than their own. It is not clear why this was so. The British considered German transmitters much better than their own and the Germans achieved remarkable ranges for the time.58 They were also much more willing to use low-powered tactical transmitters, as at Jutland. In general the Germans seem to have been unaware of the potential for interception, so they used wireless far more freely than the Royal Navy.59 The Germans used both flags and radio within their formations. The argument in favour of wireless was that flags might not be very visible in a smoky battle environment. The argument against wireless was that it might be difficult to distinguish near-simultaneous signals from many different ships.
The best indication of German surface combatant practice at the end of the war was what the British found on board the High Seas Fleet ships interned at Scapa Flow.60 The caveat is that the Germans stripped the ships of all fire-control equipment before steaming to Scapa and there was reason to think they had done the same with their latest radio equipment. Against that, examination of the larger ships showed no large empty spaces and one German remarked that the British would find nothing of interest. Transmitting offices all had a clock with an electric attachment to a box showing wavelengths, presumably to indicate when they should be used.
The standard installation for larger German warships (above light cruiser size) was three separate W/T offices: main (with separate spaces for transmission and reception), auxiliary and after-action. There was no separate British-style silent cabinet, but the whole receiving room was lagged for silence. The transmitters were a CW unit (equivalent, presumably to a Poulsen) and a quenched-spark unit. The sets were served by three parallel three-wire units, which could be used in different combinations covering the range from 350 to 600m (857 down to 500 kHz). The auxiliary office contained a spark set (130 to 600m) and the after-action office contained a portable set and a receiver (160 to 2000m) plus a small petrol generator like that in British ships.
There was no trace of tube transmission, radio direction-finding, separate fire-control sets, supersonic signalling (presumably a cover for sonar), or ultra-violent light receivers, all of which the Allies were trying out. The cruiser Emden did have a tube receiver and the cruiser Nürnberg had a heterodyne (tube) receiver. Several ships had two-tube amplifiers.
British Signals Intelligence
Radio (wireless) had important implications for any navy using it. Signals were generally transmitted in all directions, not only towards the intended recipient, so radio was always a potential source of intelligence. To some extent the sender could be identified: until well after 1945 most radio messages were sent by hand. Each of the few radiomen aboard a ship had a characteristic ‘fist’, which a skilled intercept operator could identify (during the Second World War this process was automated to some extent). Even without code-breaking, wireless offered some intelligence. Ships, both merchant and naval, use unique call signs (as addresses for radio messages). Because messages had to be addressed specifically, call-sign tracking and traffic analysis can continue even if enemy codes cannot be read. Range could be estimated, albeit crudely and not entirely reliably, by signal strength. In 1904 it was discovered that the direction to the transmitter could be measured (radio DF, or direction-finding).
A fleet relying on air reconnaissance needed some way for its airmen to distinguish its ships from the enemy’s. Beginning in 1917, the Germans painted large white circles on battleship turrets (typically at least the superfiring or highest turret forward, but often turrets forward and aft, as here). Smaller ships had circles on their foredecks. The fleet flagship Baden is shown. Her near-sister Bayern could be distinguished by her considerably less elaborate bridge, which did not incorporate facilities for a fleet commander. (Naval Institute collection)
The First World War inaugurated a new kind of intelligence exploiting wireless. In the past, intelligence was generally strategic because of the considerable delay between receiving information and providing it to ships at sea. As C-in-C of the British Mediterranean Fleet from 1899 to 1902, Fisher began the transformation to operational intelligence by exploiting the telegraph link through Malta. He still could not command the fleet from Malta once it left harbour. Within a few years, wireless changed the situation. A commander at the centre of an intelligence web could command a fleet in more or less real time, steering it towards contact with an unseen enemy. The British practiced exactly such tactics at least as early as 1913 and Fisher was writing obliquely about them as early as 1908.
Fisher’s account concentrated on the use of wireless to command forces. It referred to the mass of information accumulated by the Admiralty, without reference to sources. In 1908 the Royal Navy was very much interested in running down enemy commerce-destroying cruisers and the Admiralty was well-placed to use merchant shipping data, such as that collected by Lloyd’s, to estimate where such cruisers were operating. Admiral Fisher’s 1908 memo refers to the deliberate choice to give battlecruisers the highest possible masts so that their wireless antennas would be well placed to operate at maximum range. It was clear that Fisher appreciated the value of strategic wireless; he invested heavily in it.
In his memo Fisher made no reference to intercepting or decrypting foreign naval messages so as to keep track of foreign fleets. Nor does it seem that the pre-1914 Royal Navy built up a system of radio direction finders to track foreign message traffic. However, manoeuvre instructions do mention the common practice of decrypting messages (in one case the practice is forbidden, because the manoeuvre is intended to test the ability to intercept the enemy fleet without such information). Fisher himself later wrote about ‘listening to what all the Captains of warships at sea were saying’, which implies systematic wireless interception. A letter from a Royal Navy captain, offered for sale to the National Maritime Museum (and not bought) mentioned how, on the eve of war, the ship was listening to German naval messages to find out when war would break out. The implication was that this was standard procedure.
Well before 1914 the Royal Navy leadership realised that by monitoring and decoding German naval wireless messages, the British could achieve a degree of ocean surveillance, which in turn could be exploited.61 That the Germans enthusiastically used radio tactically helped. Instructions for the 1909 manoeuvre cautioned that officers should refrain from the popular pastime of code-breaking to predict enemy movements, since ‘in actual war the difficulties to be solved would be much enhanced by the lack of acquaintance with the enemy’s methods’. The British dilemma was to continue to pass vital information by wireless without offering too much to an enemy which might exploit British signals. Exploitation might include traffic analysis as well as code breaking. When war came, moreover, the Royal Navy expanded explosively, and many new-entry officers found it difficult or impossible to cope with coding and decoding aboard small warships. Messages were often sent in clear rather than in code, and the Germans could and did exploit this opportunity. The difference between British and German wartime practices was primarily that the British exploited code-breaking to deduce German intentions, whereas the Germans did not.
The British certainly were aware of the vulnerabilities W/T entailed. They invented a broadcast technique specifically to defeat traffic analysis, and they became interested in the deception opportunities presented by W/T. Grand Fleet W/T orders included several deceptive techniques.62 The Royal Navy also introduced W/T direction-finding, both ashore and at sea, as a way of tracking the German fleet. For their part, the Germans liked their radios and used them enthusiastically, both in port and at sea. Their dispersed anchorage at Wilhelmshaven practically required radio communication between ships. The Germans seem to have been unaware that even when they transmitted at much-reduced power, their messages could be (and were) intercepted at a considerable distance.
It is not clear to what extent the pre-1914 Admiralty relied on code-breaking and radio direction-finding. When the British finally released records of pre-Second World War signals intelligence, it appeared that they had introduced radio direction-finding in 1914. The British were certainly lucky to obtain all three of the German naval code books before the end of 1914, but they were also well aware of how to exploit that good fortune. Of the three, the German merchant fleet code (HVB), which was also used within the High Seas Fleet, was seized in Australia on 11 August, although the Australian Naval Board did not inform the Admiralty of its success until 9 September. The most secret code book (SKM) was obtained by the Russians from the sunken Magdeburg (lost on 26 August 1914) and then passed them to the British, who received it on 13 October. However, the copy in the PRO shows no sign of water immersion.63 The Germans suspected that the British had simply broken their code, which was apparently both cumbersome and poorly constructed.64 Very soon after the SKM was received, the British code-breakers intercepted signals indicating a German destroyer operation, which a British force was ordered to intercept. It sank the destroyers and in December a British trawler came up with the third of the German code books, the VB (to Room 40, the ‘miraculous draught of fishes’). The VB was apparently to have been used to communicate with the German army in Flanders. It was normally used by flag officers.65
The Germans seem to have assumed that their messages were secure because they re-enciphered (super-enciphered) their codes. That is, once a message had been rendered in code, each group was altered in a way determined by an additional key, which could be changed periodically. The great secret of Room 40 was not that the British had the code books, but that they learned to break super-encipherment so quickly that they could exploit German messages operationally.
Given their confidence in super-encipherment, the Germans used their three codes far too widely. The HVB was intended for merchant ships and so was widely distributed pre-war. It was also widely used within the Imperial Navy, both by minor units and, crucially, by U-boats and Zeppelins. The Germans never seem to have realised that a good operational intelligence organisation could piece together hints, for example minesweeping orders, to deduce fleet operations. Moreover, because U-boats stayed at sea for long periods, they were often still using an old key when a new one came into force, offering the golden opportunity of reading the same message in two keys.
In 1917–18 the Royal Navy took fighters to sea specifically to deal with Zeppelins and thus to blind the Germans. Existing aircraft could be launched from turret tops or from lighters towed at high speed by destroyers. In both cases, the aircraft ditched, its air bags keeping it afloat while the pilot was rescued and the aircraft salvaged (mainly to recover its engine). This fighter was on board HMS Tiger. (RAN via Dr David Stevens, SPC-A)
Probably the greatest lesson of wartime British signals intelligence was that code-breaking was not enough. Messages were hints of what was happening and it took someone knowledgeable, with a light touch, to turn them into usable intelligence. Despite the October success against the destroyers, the new code-breaking organisation failed to predict the German battlecruiser raid on Yarmouth in November 1914. Commander Herbert Hope proved brilliantly capable of understanding what the Germans were doing.66 Hope’s first success was the Scarborough Raid in December 1914. Others in the Royal Navy seem to have sensed the potential of signals intelligence. Admiral Jellicoe argued that it was worthwhile to mine the Heligoland Bight because the Germans would reveal their intention to sortie by sweeping an exit path.67 This warning would be made even if an operation were planned very secretly. On several occasions the Germans discovered that their efforts at security had failed and they typically imagined that they either had a traitor in a senior position or that officers were gossiping much too freely. On other occasions they imagined that fleet movements had been betrayed by North Sea trawlers secretly in British pay. When the Germans came to compile their own operational history, after the British official history had been published, they found out that between December 1914 and December 1916 the Grand Fleet had come out nearly every time the High Seas Fleet sortied. That was the most graphic measure of success of the British signals intelligence organisation. Early in 1917 Hope was promoted to Captain and sent to sea as captain of the cruiser HMS Dartmouth in the Adriatic; he was promoted to Rear Admiral after the war. By that time it must have seemed that the war he had brilliantly fought against the High Seas Fleet was more or less over; the great priority was the U-boats.
The other great success was in the U-boat war. U-boats used their radios far too freely and their messages could be deciphered and their positions revealed by radio direction-finding. That made evasion possible. It was also the basis for hunting tactics and the success of Room 40 may explain why the Admiralty considered hunting a viable alternative to convoy operations as late as it did (that is different from hunting as a valuable adjunct to convoy, in 1917–18, particularly after improved acoustic sensors entered service).
The Germans gradually changed their codes, beginning with the HVB used by minor units – and, unfortunately for them, by U-boats – early in 1916. It was used by and carried, by Zeppelins, so copies were recovered from crashed ones. However, before any had been recovered, Room 40 had reconstructed most of the codebook.68 In May 1917 the Germans finally changed their main codebook, replacing it with an entirely new one called the FFB.69 The VB was replaced by a new code which Room 40 called Nordo and this time it was restricted to flag officers. At the same time they changed all their call signs and Admiral Scheer drastically curtailed signalling. However, the Germans continued to use the AFB, the replacement for the HVB. The British could still tell, at least in theory, that the High Seas Fleet was preparing to come out through a freshly-swept channel, but Scheer’s improved security made prediction of what would come next a much more delicate proposition. Room 40 suffered three considerable failures. It failed to detect the attack by the two German cruisers against the Scandinavian convoy in October 1917 and, it appears, the German destroyer raid against a later convoy on 12 December.70 Worse, Room 40 failed to detect the major if ultimately abortive German sortie of April 1918, against a Scandinavian convoy.71 Ironically, this operation failed because of faulty German intelligence, which failed to inform Admiral Scheer that the British had changed convoy schedules.
The British were always acutely aware that the Germans could wreck signals intelligence, so during the war they sought back-ups. Since surface scouts could not work near the German coast, they used submarines. The surveillance role was considered so important that standing orders prohibited patrolling submarines from attacking any ships they saw until after they had radioed sighting reports.
For radio intelligence to work, the enemy has to co-operate. If he uses radio indiscriminately, he can be tracked, but if he realises how much he is giving away, even that valuable source of intelligence can disappear. It appears that before 1914 many naval officers were aware of the danger that their codes might be broken, but few seem to have realised that simply transmitting radio signals would allow their ships to be tracked. When the US Navy entered the war in April 1917, it used its tactical radios enthusiastically and indiscriminately. The British had to educate the Americans in the use of radio intelligence in order to convince them, successfully, to cut down on tactical radio. After the war, the British revealed that they had been reading German codes, an admission which led the Germans to adopt the ‘Enigma’ machine code they used during the Second World War. For their part the British seem to have concluded that since codes had been shown to be so breakable, no one would be foolish enough to use radio extensively in a future war. That led them to grossly underestimate the number of messages they would be sending, which in turn led them to adopt coding methods which proved insecure.
German Signals Intelligence
The Germans were listening to British signals, but in 1914 they were making no systematic attempt to exploit British W/T. According to their official history, in December 1914 British W/T was being intercepted mainly by designated ships (during the Scarborough Raid, for example, the battlecruiser Moltke was W/T intercept ship). Typically several ships in a force had their receivers tuned to a particular British frequency (‘wave’). The Navy Office furnished keys it thought the British were using, but decoding was said to be slow and clumsy and there is no indication in the official history that at this time it had any impact on operations. Messages could not be decoded quickly enough for the C-in-C to use tactically, whereas, unknown to the Germans, their messages were being turned around instantly, even in December 1914.
The first German successes were by a Bavarian army unit at an intercept station set up in November 1914 near Roubaix on the Franco-Belgian frontier.72 This was not a naval operation, but most of what Roubaix intercepted was British naval traffic, so it worked on that material. Although not trained in cryptology, the officers at Roubaix broke a simple British naval cipher at the end of December 1914. They mainly recovered reports on merchant ships stopped and checked for contraband at Dover, but they also broke messages to and from British submarines at sea. On 31 January Roubaix broke the British message advising all British merchant ships approaching home waters to fly neutral (or no) flags; the Germans used this information as part of their justification for the order to sink all shipping in the new ‘war zone’ they soon declared around the British Isles.73 In May 1915 Roubaix was routinely solving messages from all British coastal patrols, from Channel patrols and in the British Merchant Navy Code.
The following month Roubaix broke the Allied Naval Code used between British and French ships in the Channel. In a few cases Roubaix profited by British errors, such as a request to repeat a message in an earlier code because the message in the current code was unreadable. That brought information on the activities of the Grand Fleet (but this information is not reflected in Krieg zur See). In mid-July a naval officer was detached to Roubaix to arrange co-operation with the Imperial Navy. From the end of July 1915 decrypts were sent directly to the command of the High Seas Fleet, Naval Forces Baltic, MarineKorps Flandern, the Admiralstab, naval stations Kiel and Wilhelmshaven and the commander of U-boats. On 30 October 1915 High Seas Fleet commander Admiral von Pohl formally thanked Roubaix for its bimonthly bulletins, which gave a nearly complete picture of enemy patrol and home defence forces (but not, significantly, of the Grand Fleet). That in turn was very helpful to the U-boats and Zeppelins.
Given these successes, the Imperial Navy set up its own intercept and evaluation operation (Entzifferungsdienst) at Neumünster, with an outstation at Bruges.74 To operate it the navy withdrew its liaison officer from Roubaix in February 1916 so that it could exploit the techniques developed there. Roubaix continued to intercept and attack British codes. Late in February Neumünster read a British message passing intelligence that three raiders disguised as Danish merchant ships were breaking out (this seems to have referred only to Greif and Wolf).75 During the 1916 Lowestoft raid Bruges reported the messages recalling ships off the Belgian coast. However, the system operated sluggishly. The report from British submarine E 23 that she had torpedoed the battleship Westfalen during the August 1916 fleet sortie was promptly intercepted and broken by Bruges, but it waited five hours to send it on to Neümunster and the signal was further delayed en route to higher headquarters. Moreover, given the fragmented German naval command, it is not clear who could have made full use of such information.
Most of what was obtained was low-level information including minefield data.76 For example, Roubaix broke a series of British messages sent in 1917 indicating which German minefields had been swept. Earlier messages confirmed the German belief that British submarines were patrolling the Heligoland Bight and the North Sea. It helped that, like the Germans, the British did not change their two most-frequently used codebooks, one of which was for the Auxiliary Force.
The Germans also tried traffic analysis, which was useful because the Royal Navy was so huge that it had to send numerous operational messages. The German official history refers again and again to the use of the volume of British wireless traffic as an indicator of whether the British were aware of a German operation. For example, in April 1915 the High Seas Fleet sortied to lay mines on the Dogger and Swarte Banks. Until some hours after that British wireless volume remained normal, which suggested that the British were not yet aware of the operation. The Germans also tried to estimate range from the strength of wireless transmissions, but soon learned (in 1915) that this technique was useless.
In the first half of 1917, the British were becoming more aware of German decryption and were adopting new techniques which complicated analysis. Even so, Roubaix continued to forward large numbers of reports and it was successful enough for a second surveillance command to be set up.77 In the autumn of 1917 all German army signals intelligence was unified as a new signals intelligence service and Roubaix shifted mainly to army work. Presumably this fed into the build-up to the April 1918 army offensive. By that time Neumünster and the Imperial Navy’s signals intelligence arm were well established.
It seems clear that the Germans never made much use of what they obtained. They did not, for example, use W/T to find convoys. Probably most importantly, they did not have a Commander Hope or other analysts who understood naval operations well enough to use apparently low-level intelligence sent in breakable ciphers to deduce Grand Fleet operations. Again and again, the official German history refers to operations premised on vague agent reports from neutral capitals and on rumours in the British press. These were not covers for hard intelligence; the official history is quite willing to cite signals intelligence information.
Late in the war the Royal Navy made extensive use of kite balloons to extend ships’ vision. They are the balloons flying in numbers above the ships of the Grand Fleet at Scapa Flow. The absence of turret flying-off platforms dates this photograph to 1917. Although the balloons may resemble later barrage balloons intended to counter low-flying bombers, the spots below each balloon show that they were manned. (Naval Institute collection)
Neumünster turned out to be very much a double-edged sword, because it used radio to broadcast its own successes, alerting the British and the Russians of their own code failures. The worst example came in 1917 when Neumünster sent out its solutions of the current British fleet code in German code, each British solution followed by its German coded equivalent. In effect it was giving the British the current German FPB code.78
It is not clear to what extent the Germans were aware of British signals intelligence. The author of Krieg zur See (Nordsee) indicates both that they were unaware and that, on the eve of Jutland, the Germans were using special phrases to deceive any listeners. For example, U-boats watching the British coast were to be told that the High Seas Fleet was at sea by the phrase ‘take into account that enemy’s forces may be putting to sea’, and other phrases were to be used by the U-boat reconnoitring the target area, Sunderland.
The Aircraft Revolution
In 1914 navies generally agreed that aircraft had changed sea power, but they apparently over-rated what they could do. By 1918 aircraft technology had advanced to the point that expectations were being realised. Where in 1914 the British had seaplanes that could barely lift torpedoes from calm water, in 1918 they had carrier-based torpedo bombers which offered a credible way of attacking the High Seas Fleet in harbour (something proposed in 1914). The great surprise for a modern reader is how overrated (not by any means under-rated) aircraft were during the war. Their operations and their potential were described in much the way they might have been during the Second World War. For example, when a 1915 British raid on Zeebrugge damaged a U-boat there, the Germans concluded that the base was unsafe until it had been provided with massive anti-aircraft defence. In 1917–18 there was a massive fight to control the air near Zeebrugge, as though the base might be bombed out otherwise. The Germans built massive U-boat shelters similar to those at European U-boat bases during the Second World War, even though the bombs they faced were pathetic by later standards and bomb-aiming grossly inaccurate. For that matter, in 1917–18 the US Navy made air attacks on German submarine bases a key element of its strategy. The only reason that is not obvious nearly a century later is that the war did not last long enough for it to go into effect.
As understood before the war, the most striking aircraft role was reconnaissance. The pre-war Royal Navy was so impressed by German airship progress that in 1912 it decided that it was no longer possible for it to use uninhabited islands off the North German coast as temporary destroyer bases supporting an inshore squadron. At the time, the Germans did not yet have Zeppelins in naval service, but they were about to acquire some.79 The British found the potential of airships so impressive that the cruiser Hermes was converted into a seaplane carrier for the 1913 Manoeuvres specifically so that her aircraft could simulate the Zeppelins the Imperial Navy was about to commission. For their part the Germans were impressed by the demonstrated British ability to operate aircraft from ships. The British, not the Germans, conducted the first naval air raid, against Cuxhaven at Christmas 1914.80 Even the most primitive seaplane carriers gave the British deployable naval aircraft, so HMS Ark Royal was able to provide reconnaissance and spotting support at the Dardanelles in 1915. Later the Royal Navy used air spotting to support monitors shelling Zeebrugge and Ostend.
From about 1915 on, the High Seas Fleet frequently took a few floatplanes to sea on board cruisers, which could hoist them out for launching. At this time the British were doing the same on board dedicated carriers, but until 1916 these ships were not integral with the Grand Fleet. The German official account suggests that the aircraft on board the cruisers were rarely if ever launched at sea, no great surprise in the rough North Sea. Only at the end of the war did the Germans plan to convert cruisers into seaplane carriers.81 By that time the British had a large carrier force and they were deploying a large force of wheeled – higher performance – seaborne aircraft.
Given the failure of British wartime attempts to develop viable airships comparable to the German Zeppelins, the Grand Fleet sought alternatives. One was reconnaissance aircraft launched from the take-off decks of carriers such as the converted liner HMS Campania. She demonstrated that wheeled aircraft could take off from a deck and she was rebuilt as shown to lengthen the deck so that she could fly off two-seat reconnaissance aircraft. Typically they were seaplanes on wheeled trolleys, not land planes which might ditch alongside on their return. On that basis she could accommodate ten large aircraft. In 1917 Campania was credited with seven deck-launched reconnaissance aircraft (Furious had four). The next ship to have that capacity was the modified ‘large light cruiser’ Furious. Due to her age and poor performance, Campania was not included in the Grand Fleet carrier force in 1918. Campania is shown on 5 April 1916. (Author’s collection)
On the eve of war the Royal Navy saw the Zeppelins as the eyes of the German fleet, but it seems to have taken some time for the Germans to appreciate what they had. The German navy became interested in Zeppelins due to the success of the first German Zeppelin airline, Deutsche Luftfahrt (DELAG) and after the army bought some.82 Admiral Tirpitz initially resisted the idea.83 As C-in-C High Seas Fleet, Admiral von Ingenohl did not use airships to support the fleet, but his successor Admiral von Pohl did so for the first time on 29 March 1915.84 On 4 June he explicitly recommended airships as cruiser replacements, something the British thought the Germans had already done.85 He pressed for considerable enlargement of the airship force and for a corresponding increase in bases (big sheds [hangars] were necessary to shelter the huge airships from the wind). Among the results was a programme to build six two-airship sheds. The official German history of the war at sea commented that in fine weather an airship might be considered equivalent to two light cruisers (as a scout, as it would not have anything like their firepower).86 Before the war the Germans expected the British to come to them, into the Heligoland Bight. They wanted intensive reconnaissance of the Bight, which was practicable using airships and some aircraft. They had far less interest in what the British were doing further out to sea, with one important exception. When they went to sea, they needed reconnaissance around the path it would take. It appears that Admiral Scheer was the first to seek wider-area reconnaissance by the navy’s airships. By 1916 the British saw German airships as Scheer’s insurance against engaging the Grand Fleet and they were increasingly interested in destroying them, using either anti-aircraft guns (an airship was not so very much faster than a destroyer) or shipboard fighters.
Initially the Germans were unlucky. Their first airship L 1 crashed into the sea near Heligoland in September 1913. Their second (L 2) burned at Fuhlsbüttel on 17 October 1913, killing nearly their whole naval aeronautics staff. With only L 3 in service at the outbreak of war and both the North Sea and the Baltic to cover, the German navy had to press a non-rigid (Parseval type) into service.87 However, it reached an agreement with the army that the navy would receive every second Zeppelin and also some Schütte-Lanz airships. Bases, nearly all with anti-aircraft weapons, were built at Tondern and at Haage in East Frisia and at Seddin in Pomerania. A network of meteorological stations was set up, communicating with a central station at Wilhelmshaven. During the summer of 1917, with the navy regularly bombing Britain with airships, the army turned over its remaining Zeppelins (LZ 111, LZ 113 and LZ 120) to the navy.
By the end of 1914 the navy had four operational airships of the improved L 3 class and two named training airships. As an indication of what airships could already do, these ships, which were considered small by later standards, had a payload (including fuel) of 8700kg, at a time when aircraft payloads were in the low hundreds of kilogrammes. Proposals to raid England with Zeppelins were made by the Admiralstab as early as 20 August 1914, but its chief Admiral von Pohl resisted on the grounds that they were desperately needed as High Seas Fleet scouts. Only with the appearance of the fourth Zeppelin (L 6) in October 1914 could one be spared for raiding, which began in 1915. The initial Zeppelin attacks on London, executed by naval airships, were allied to the new campaign of unrestricted submarine warfare.88 In both cases, the hope was not so much that the British would be knocked out of the war (as the Germans hoped in 1917), but that they would find their vulnerability, both to bombs and to torpedoes directed at merchant ships, so shocking that they would elect to come to terms. From a German fleet point of view, the raids on London seriously detracted from the air reconnaissance on which the High Seas Fleet depended. Dogger Bank was another milestone: for the first and only time during the war, an airship (L 5) was present above the German fleet during a battle.89
As an indication of how quickly Zeppelins were pressed into service, L 20 was commissioned at the end of 1915, L 39 at the end of 1916 and L 60 at the end of 1917. The total acquired by the Imperial Navy was sixty-five Zeppelins, nine Schütte-Lanz airships (which had wooden rather than metal structure), three Parseval semi-rigids and one M-ship. Deducting special and training ships, seventy-two of the seventy-eight were operational, for reconnaissance and attacks. On average each performed sixteen reconnaissance missions and three attacks. In the North Sea, seventy airships performed 926 reconnaissance flights and 159 attacks. In the Baltic, twenty-four airships performed 220 reconnaissance flights and delivered forty-one attacks. The first thirty attacks were carried out in 1915, followed by 107 in 1916, forty-six in 1917 and seventeen in 1918. All were against fixed targets; an attempt to develop a remote-controlled anti-ship weapon for use from airships failed. The elements and the explosive potential of the hydrogen in the hull destroyed about as many airships as the enemy. Losses to explosion amounted to two in 1915, four in 1916 and six in 1918. Losses to weather: four in 1915, four in 1916, five in 1917 and one1 in 1918. Losses to enemy action: four in 1915, eight in 1916, nine in 1917, five in 1918.
A July 1918 German tactical instruction describing airships gives some idea of their capabilities and limitations as understood after four years of war.90 At this time the main operational type was the L 50 and a new high-capacity L 70 was about to enter service. An L 50 had five high-efficiency 260hp engines with ‘super compression’ (i.e., superchargers for high altitude) driving four propellers (the two rear engines were coupled to one propeller). It was 664ft 8in long (beam 78ft 5in) and could reach 67mph. An L 70 had seven such engines and six propellers; it was 693ft 10in long with a beam of 78ft 5in and it could reach 78mph. A typical crew was twenty-one (twenty-five in L 70 and later ships), including two elevation and two directional helmsmen. Armament was two machine guns plus a 20mm cannon (being introduced) and bombs weighing up to 300kg (660lbs). W/T range using a 1.2 kW Telefunken set was 500–600nm (300nm for the emergency set). The set used a 100m antenna hung below the ship. However, the W/T could be used only when the gas bags were full. When hydrogen was released, it could form an explosive mixture under the airship. ‘In the majority of cases the ship will not be able to inform the station with which it is in W/T communication that she dare not transmit any further’, e.g. might not be able to acknowledge a long message. In that case she could still receive. Airships typically carried both the AFB and FFB code books with transposition table for the FFB, but if flying over neutral or enemy territory they were limited to the AFB with the current transposition table and that for the next day.
At about 200m visibility was about 30nm; at about 4000m it was about 60nm. Ceiling depended on weather, but the North Sea fighting airships could generally reach 6000m (19,685ft) during an attack. The Germans considered an airship safe from anti-aircraft fire at a range of 4–5nm and an altitude of 3000m. Endurance was normally twenty-four to forty-eight hours, but with special preparation it could be extended to 100 hours. Over the sea, airships fixed their positions by using dead-reckoning and directional W/T. Procedure as given in September 1918 was for an airship in contact with an enemy force to transmit every 15 minutes so that her absence could be noted. Only the shadowing airship was to signal. Ships could not be berthed or brought out of their sheds with wind blowing at more than Force 3 (5m/sec). Head winds of Force 4–5 halved the speed.
The Royal Navy also wanted airships, but it was much less successful developing them. It seemed to have had a much clearer idea of the potential value of airship reconnaissance for naval forces. That led it both to an extensive airship programme and to take fighters to sea specifically to deal with enemy air reconnaissance. DNO Captain H S Bacon made the first proposal to build a British airship on 21 July 1908; on 7 May 1909 Vickers received a contract to build an airship using Zeppelin technology as part of the 1909–10 programme. At this point the Royal Navy was interested mainly in airships as fleet scouts and a January 1909 CID report it favoured rigids over non-rigids due partly to their higher speed (hence better resistance to wind).91 A rigid was easier to moor and it was considered easier to navigate, as a crew member could climb to the top of the hull to take navigational fixes. The Germans had not yet integrated airships into their fleet, but the CID reported evidence that they planned to do so. It compared a £35,000 airship with an £80,000 destroyer and a £400,000 light cruiser. This was HM Airship 1, Mayfly. Unfortunately she was wrecked on 24 September 1911 by ground wind before she could fly (she was broken in two). Admiral Sir A K Wilson headed the inquiry and he became convinced that aircraft were much better; First Lord Winston Churchill backed the airships.
In December 1911, however, the naval and military attaches in Berlin submitted an alarming report of German airship progress. Based on what German airships had already done, it seemed that they could reconnoitre the whole North Sea
except in foggy or stormy weather, it is probable that no British war vessels or torpedo craft will be able to approach within many miles of the German coast without their presence being discovered and reported . . . Unless we had obtained the command of the air, any idea that our torpedo craft could seek shelter among the Frisian Islands and remain there undetected must be abandoned . . . It is difficult to exaggerate the value of this advantage to Germany. 92
Hence the urgent need to simulate a Zeppelin on reconnaissance duties in the 1913 Manoeuvres. The CID also noted the potential for a Zeppelin bombing raid on the United Kingdom, as Zeppelins had already flown the equivalent of the distance from Germany to England. The clearest evidence of a purely naval role for Zeppelins was the erection of the necessary protective sheds (hangars) at Hamburg and Kiel. At this stage the army had much more experience with airships, but the CID pointed out that they were much less suited to expeditionary operations, so in October 1913 the Royal Navy took over all British airship operations, including the army’s stock of airships.93
The Mayfly disaster showed how difficult it was to handle airships on the ground, so in June 1913 the Admiralty proposed that it begin with smaller non-rigids, to gain experience.94 Bids were requested and Vickers offered one rigid airship based on Zeppelin technology and four Parseval-type non-rigids (it had a licensing agreement with Parseval).95 Other major shipbuilding firms were also approached. Armstrong (Elswick) reached a licensing deal with the other major non-rigid firm, Forlani (Italian). Vickers received a contract for the Zeppelin and three Parsevals under the 1913–14 Estimates and Armstrong one for three Forlanis (one to be deferred) under the 1914–15 Estimates, the navy deciding not to buy a rigid airship (presumably of Schütte-Lanz design). A second Astra Torres was also ordered. Rigid development proved difficult, the first successful one (Vickers’ HMA 9) not leaving her shed until 16 November 1916. As a result, the Grand Fleet never enjoyed the advantage of airship reconnaissance it believed the High Seas Fleet had.96
To post-war navies, the German successes outweighed wartime British failures. It seemed that airships were the ideal long-range scouts of the future, limited only in that fighters could shoot them down. Airships could lift so much weight, however, that they could carry fighters for self-defence. The British tried this idea even before the end of the war and the US Navy later built and deployed the airship-borne F9C Sparrowhawk, on board its two airship ‘aircraft carriers’ Akron and Macon. Despite their loss, as late as 1940 the US Navy was still interested in huge airships for scouting. It shared the Royal Navy’s need for strategic scouts capable of watching an enemy fleet sortie. Like the Royal Navy, the inter-war US Navy made up for the lack of such scouts by investing heavily in an entirely different surveillance technology, signals intelligence. Unlike the Royal Navy, the US Navy discovered to its great misfortune what would happen when the enemy suddenly changed codes and ruined its electronic surveillance system – Pearl Harbor.
Although the Royal Navy could not obtain the ocean surveillance it sought in the form of airships, it went much further than the Germans in taking aircraft to sea. The key development, demonstrated on board the carrier Campania in August and November 1915, was that a wheeled aircraft could take off from a flat deck on shipboard. Initially that was a short inclined deck forward of the ship’s bridge, which sufficed for a Sopwith Schneider floatplane riding a wheeled trolley. There was no attempt to recover the floatplane on deck; it had to land on the water.97 The deck was too short for the desired W/T-equipped two-seat reconnaissance aircraft, but it was extended sufficiently by modifying the ship. In this form Campania barely missed Jutland, not having sailed in time and being recalled because she was not sufficiently screened against submarine attack. Given the success of this project, two fast merchant ships were converted into carries with take-off decks (Vindex [commissioned 26 March 1915, conversion completed October 1915] and Manxman [commissioned 17 April 1916, but conversion not completed until December 1916]). Vindex was the first ship ever to launch a landplane (3 November 1915). She was assigned to the Harwich Force. Manxman was assigned to the Battlecruiser Fleet rather than the Harwich Force, but transferred to the Mediterranean in October 1917. She first operated the Sopwith Pup, which became the standard fleet fighter.
The first experiments with arrester gear were conducted about February 1916, leading ultimately to the decision to complete a suspended liner, laid down for Italy, as the full flush-deck carrier HMS Argus. Captain Sueter of the Air Department pressed for a 27-knot cruiser-carrier, for which a design was completed, but given the urgency of other projects, it was not laid down during 1916. In contrast to other naval officers writing at this time about carrier requirements, Sueter included among their functions attacks against both land targets and ships at sea.98 Sueter wanted three carriers. Director of Air Services agreed that Sueter’s ships would be well worth while, particularly to counter both Zeppelin reconnaissance and the threat that Zeppelins might bomb the fleet in the latter stages of a battle. Third Sea Lord referred to a design offered to Sueter in 1915. Consideration had been suspended while experiments in landing-on were conducted, but it was revived in February 1916. A modified design was submitted and now it was estimated that such a ship would take 15 to 18 months to build. Third Sea Lord wanted to wait for practical trials by Campania and in that case trade some speed for bulges and a serious gun battery (5.5in guns). The initial result of all of this discussion was to take over a suspended liner and order her completed as a carrier (Argus was purchased in August 1916). After that the landing-on experiments resumed. Completion was forecast in about a year.
In the spring of 1918, Grand Fleet air requirements cantered on reconnaissance: to find the High Seas Fleet and to deny it Zeppelin reconnaissance using organic fighters. The converted cruiser Cavendish (Vindictive) could operate six reconnaissance aircraft. At that time the projected Grand Fleet carrier force consisted of Cavendish, Furious and Argus, with fighters also launched by the ‘large light cruisers’ Courageous and Glorious and the cruiser Caledon: a total of twenty-five deck-launched reconnaissance aircraft and thirty fighters. In fact many battleships and cruisers also carried fighters atop their turrets, so the fleet fighter figure was considerably higher. Note that at this point the fleet air requirement did not include torpedo bombers (Argus was credited with fifteen reconnaissance aircraft and ten fighters). Cavendish was renamed Vindictive to remember the cruiser which fought heroically at Zeebrugge in April 1918. Note her separate flying-off and flying-on decks. Vindictive was completed on 21 September 1918 (commissioned 1 October). She was used only for trials (on 1 November a Sopwith Pup was successfully landed on her after deck). After the war she ferried aircraft to North Russia, but they operated only from ashore, perhaps because she grounded almost immediately after arriving. (Author’s collection)
In October 1916 Admiral Jellicoe complained that he still had far too few aircraft, his only carriers being Campania and Engadine, which had been taken over as a seaplane carrier in 1914. As an interim solution, two more merchant ships were taken over while building and completed as carriers with flying-off platforms: Nairana and Pegasus. They were completed on 14 August and 25 August 1917 respectively. By this time there were effective torpedo bombers and Sueter was trying to raise interest in operations (see the chapter below on the fleet in battle).
In October 1916 Jellicoe was superseded by Admiral Beatty, who shared his urgent need for air services for the fleet.99 In January 1917 Beatty wrote to the Admiralty that he badly needed more aircraft to deal not only with the Zeppelins but with German seaplanes carried, he claimed, by six seaplane carriers (which did not actually exist). He had only three carriers, of which Engadine had no flying-off platform, hence was useless in anything but the calmest weather, Manxman was too slow and short-legged and only Campania, which was very old and hence unreliable, was at all what he needed. He hoped the new seaplane carrier (presumably Argus) would help, but he had no idea of her capabilities or of her likely date of completion. In February the Board answered that in addition to his own ships, the long-range bomber wing of the RNAS, which was attacking blast furnaces and munitions plants in Alsace-Lorraine, would also be hitting Zeppelin bases. Argus was described with an air group of eight reconnaissance aircraft, six anti-Zeppelin aircraft (i.e., fighters) and, if necessary four torpedo bombers. It was hoped that the reconnaissance aircraft could land on; experiments were most encouraging. Argus should be ready in November.
Beatty had convened a Grand Fleet Committee on Aircraft Requirements, which reported on 5 February.100 It stated two requirements for aircraft to work with the fleet: reconnaissance and anti-Zeppelin. Shipboard reconnaissance would mesh with long-range reconnaissance conducted by seaplanes based ashore (the Royal Navy was taking delivery of new ‘Porte Boats’ and ‘Large Americas’). By this time arrangements had been made to carry two seaplanes and a portable flying-off platform on board the large light cruiser Furious, still under construction; in March it was decided simply to replace the ship’s single forward 18in turret with a flying-off deck and hangar. There may have been proposals to convert some light cruisers (Beatty opposed them, but he may have referred to a garbled version of the Furious proposal).
The ‘large light cruiser’ Furious was modified while under construction, her forward single 18in gun replaced by a hangar and flat flying-off deck. Admiral Beatty approved reconstruction because the loss of a single slow-firing gun would mean little to the firepower of his fleet. He rejected similar conversion of her half-sisters Courageous and Glorious on the grounds that losing a twin 15in turret was a very different matter. They were given the usual turret flying-off platforms for fighters. Aircraft were lifted from the hangar by crane. Furious is shown as initially completed. (Author’s collection)
After a successful landing onto the flight deck on 2 August 1917, Admiral Beatty recommended replacement of the after 18in gun by a landing-on deck. Work was simplified by duplicating the deck and hangar already designed for HMS Vindictive. In addition, the forward hangar was given a hydraulic lift. This work was considered urgent; it was completed in March 1918. The flying-on deck proved useless due to turbulence created by the ship’s superstructure. In this form HMS Furious raided the German Zeppelin base at Tondern on 19 July 1918 – an attack at source to blind the German fleet – using seven Camel fighter-bombers. Two airships (L 54 and L 60) were destroyed in their sheds. This was the first carrier air strike in history. Furious is shown as completed in March 1918, with a flying-on deck aft and a net to act as back-up arrester gear. (Dr David Stevens, SPC-A)
An Admiralty Air Policy was laid down early in 1917. The existing seaplane carriers would be maintained and Furious modified. Work on Argus would proceed, but the hope of completing her by November 1917 would be abandoned. Four new carriers, two large and two small, would be built (the two smaller ships were Nairana and Pegasus).101 By September, plans for one of the seagoing carriers had been abandoned, but it had been approved for the cruiser Cavendish (later Vindictive) to be completed as a carrier and the two cruisers Cassandra and Yarmouth had been provided with flying-off platforms for single aircraft (fighters). At a meeting in September, Beatty asked for aircraft arrangements in the ‘large light cruisers’ Glorious and Courageous and in the cruisers Caledon and Dublin. The question was whether these ships provided enough reconnaissance and fighter aircraft to give the Grand Fleet superiority over German aircraft. At this time the Grand Fleet was credited with the carriers Nairana, Furious, Engadine and Pegasus and the air-capable cruisers Cassandra, Dublin and Yarmouth. Only Furious could fly off her four reconnaissance aircraft from her deck; the other three carriers had to fly theirs off from the water. Campania had seven deck-launched reconnaissance aircraft, but was much slower than the others. Furious had four fighters, Nairana three and Pegasus five, all flying off decks. None of these ships was credited with torpedo bombers. In the spring of 1918 the carriers Cavendish (six reconnaissance aircraft) and Argus (fifteen reconnaissance and ten fighters) would be added. Courageous and Glorious would each be able to fly off two fighters and the cruiser Caledon would add another, for a total of twenty-five deck-launched reconnaissance aircraft and thirty fighters.
Grand Fleet Battle Orders envisaged keeping five reconnaissance aircraft continuously airborne during the approach and in action. Allowing ten for the approach and fifteen for the battle, the numbers available were just sufficient. However, the thirty fighters might well not be sufficient, so the Admiralty proposed fitting flying-off decks to as many more light cruisers as possible, pending Beatty’s approval. ‘As for his claims about the Germans, it is not known to what extent seaplanes or aeroplanes are being carried with the High Seas Fleet.’
As of September, it was proposed to fit an after landing-on deck and additional hangar to Furious and to convert Courageous and Glorious with forward decks and hangars as in Furious as already converted. The Admiralty analysis pointed out that even the largest carrier could launch only one aircraft at a time and each time it turned into the wind to do so it lost position in the fleet formation. On that basis it was by no means clear that adding aircraft capacity would also add to the number which could usefully be launched. Converting two more ‘large light cruisers’ would triple the number launched at one time and would add ten to twelve reconnaissance aircraft. This was by far the quickest way to gain more carrier capacity, as conversion would take no more than three or four months. The sacrifice of forward turrets was acceptable given the crushing superiority of Allied battleship gun power over the High Seas Fleet. Given the additional reconnaissance aircraft, it might be possible to release the new Argus from that role and devote her completely to carrying torpedo bombers. Her hangar could accommodate more than twenty of them and three or four reconnaissance aircraft could be lashed to her flight deck for pre-strike reconnaissance. In that case Argus could be employed on special operations against the enemy’s ships in harbour or in support of the battle fleet. Given her comparatively low speed, Argus would find it difficult to remain in position in the fleet as she flew off many reconnaissance aircraft, so she might be particularly suited to torpedo attack.
By the end of the war, carriers and their aircraft were firmly embedded in British naval thinking, to the extent that it was the British delegation to the Washington Naval Conference (1921) which ensured a rather large tonnage (i.e., potentially a large number) of carriers in the Royal Navy and the US Navy. Aircraft were also key to the Allied ASW campaign.
Smaller wartime carriers could launch only small single-seaters. This is HMS Nairana in 1918. She carried three fighters forward and four seaplanes aft, serviced by the massive gantry. (Allan C. Green collection via State Library of Victoria)
Navies favoured line-ahead formation because of its simplicity. Unlike all other formations, line ahead required virtually no signalling, as each captain could tell what he had to do by looking at the ship ahead of him. Conversely, if a line-ahead formation was disrupted, ships might easily be thrown into disorder. The German fleet is shown in line ahead before the war, with pre-dreadnoughts in the foreground. The Germans sought to organise their fleet in homogeneous eight-ship squadrons. (Naval Institute Collection, photo by Brown Bros., New York)