1. Introduction
1. For an example of this practice, see Captain Peter Hore, RN, Sydney, Cipher, and Search: Solving the Last Great Naval Mystery of the Second World War (Seafarer Books, Rendlesham: 2009). Captain Hore describes the way in which the Admiralty and the Australian naval intelligence centre tried to locate the German raider operating in the Indian Ocean. This was the German Kormoran, which sank (and was sunk by) HMAS Sydney.
2. See P J Cain and A G Hopkins, British Imperialism 1688–2000 (Longman, London: 2000 [second edition; first edition 1993]). The position of the United States within the informal empire but also as a force attempting to disrupt the formal empire gives some idea of the complexity of informal empire. Much of the formal empire was obtained to support the trading requirements of the informal empire; places like Hong Kong were valuable as trading ports, not in themselves. The British (or at least some of them) seem to have been unique in the nineteenth and early twentieth centuries in accepting the modern idea that investment and return were what counted, not physical control; hence many modern claims that conquest does not pay. Of course the British (or at least some of them, in government and the City) well understood that control of some territory made it more attractive for informal-empire partners to work with the British.
3. Staff development up to 1929 is described in the Naval Staff Monograph (Historical), ‘The Naval Staff of the Admiralty: Its Work and Development’ (ADM 234/434).
4. Unfortunately the Notebook collection is incomplete. The greatest apparent gap is in the series of wartime light cruisers. The collection also seems to lack almost all references to cruiser reconstruction or major modernisation; for example, there is no reference to the London reconstruction. Reconstruction designs may have been assigned to constructors at the dockyards performing them, and the Notebook series seems to include only work done in the central DNC establishment in London, and later in Bath (with some material from the London headquarters). It includes virtually nothing from constructors assigned to the fleet.
2. Protecting Trade
1. ADM 116/1013A, Case 3749 Vol. 2, consisting of notes on various designs. Controller’s note of 13 January 1908 is numbered CN.024/1908. The general design concept was approved on CN.9153/07, but ‘several subsequent discussions showed there was a general feeling of building rather a better class of vessel, in view of its principal role being to meet the German 3rd class Cruisers’.
2. The Cover shows two alternatives for the ‘New Boadicea,’ a preferred 420-footer (420ft x 43ft x 14ft 9in, 4,000 tons, 19,000shp) and a 410-footer (410ft x 43ft x 14ft 9in, 20,000shp), both capable of 25kts, and both armed with twelve 4in guns (twice the Boadicea battery) and two 18in deck torpedo tubes. An alternative version of the 420-footer had four 6in guns instead of twelve 4in. Protection was limited to a ½in deck with 1in slopes over the engine room and a 4in conning tower, as in Boadicea. Both designs showed 650 tons of fuel in normal condition, compared to 450 tons for Boadicea.
3. On 17 January 1908 Watt offered Designs A and B. A was armed with twelve 4in guns, B with two 6in and eight 4in guns (each had two submerged 18in torpedo tubes). As in the later design, each had much more fuel than Boadicea (650 tons normal, 1,600 tons fully loaded, compared to 450 tons and 1,050 tons). Required speed was 25kts, and power was given as 21,000shp. Displacements were given as 4,325 tons and 4,400 tons respectively, somewhat below Controller’s estimate. Estimated costs were £400,000 and £410,000.
4. The 6in capped AP shell could penetrate 4in KC armour at 6,000yds and 6in at about 3,500yds under optimum conditions. Because the gun was not very effective against armour, the planned capped common (higher capacity) shell had not yet been developed as of August 1908; it would probably penetrate 4in KC armour at 4,000–4,500yds. The most effective shell against unarmoured sides was the HE (Lyddite). As of August 1908 the main role of these cruisers was to fight torpedo craft and cruisers of about their own size, so DNO (Captain Roger Bacon) proposed arming them only with Lyddite (75 per cent) and pointed common shells. Controller suggested that, as the ships were intended for foreign as well as home service, they should have some shrapnel shells as well, presumably for anti-personnel use (Imperial policing). The approved outfit was then 70 per cent Lyddite, 25 per cent common and 5 per cent shrapnel (approved August 1908).
5. These became Designs I, II and III, respectively; unfortunately the Legend is undated. However, a comparison of the original twelve 4in design (A) with these three (B, C and D) was dated 17 January 1908. Design I (two 6in) was a 4,400-tonner (21,000shp). Design II (four 6in) was a 4,600-tonner (21,500shp). Design III was a 4,800-tonner (22,000shp). All were expected to make 25kts, and all had a total fuel capacity of 1,600 tons (coal and oil), of which 650 tons were carried at normal displacement (Boadicea’s capacity was 1,050 tons, 450 tons of it carried at normal displacement). Ammunition was the normal 150 rounds per 6in and 4in gun.
6. A Legend for Designs B and E was dated 23 January 1908. The ship was lengthened (430ft x 46ft x 15ft 3in, 4,650 tons rather than 4,400 tons) and power was increased from 21,000shp to 21,500shp.
7. These ships, like earlier turbine cruisers, had three boiler rooms, all adjacent, arranged fore to aft, with engines abaft them. Bristol (Curtis turbines) had two adjacent engine rooms, each with a condenser on one side (to port of the forward engine room, to starboard of the after one). HMS Newcastle of this class, with Parsons turbines, had three engine rooms, all abeam of each other, with condensers abaft them. Similar arrangements were planned for the Parsons and Curtis cruisers of the 1909–10 programme. All had four funnels, the middle two wider than those at the ends. Each boiler room had uptakes at both ends, the after uptake of the forward room being combined with the forward uptake of the middle room, and the forward uptake of the after room being combined with the after uptake of the middle room.
8. The papers of the 1909 Imperial Conference are in ADM 116/1100B. Given the German threat, on 22 March New Zealand offered a modern battleship, which became the battlecruiser HMS New Zealand. On 29 March the Canadian House of Commons passed a resolution calling for the establishment of a naval service. On 4 June 1909 the Australian premier offered a battleship for Empire Defence. Given the New Zealand offer, the Canadian decision and ongoing Australian offers of assistance, on 30 April the British government invited the Defence Ministers of the four Dominions and of the Cape Colonies (South Africa) to an Imperial Conference. In the Australian case, the effect of the 1909 Conference was dramatic. As late as 1908, the Australian government was interested in a coastal navy: six ‘River’ class destroyers, nine ‘C’ class submarines, and two depot ships. See document 45, a draft letter dated 20 August 1908 to the Australian premier, in Nicholas A Lambert, Australia’s Naval Inheritance: Imperial Maritime Strategy and the Australia Station 1880–1909 (Royal Australian Navy, Canberra: 1998; No. 6 in the series of Papers in Australian Maritime Affairs).
9. The difference was two guns (abreast) on the forecastle rather than one.
10. A 6 May 1915 note by E-in-C estimates that, if they burned only oil fuel, the boilers in the middle boiler room would produce an additional 3,500shp.
11. Folio 15 of Cover for ‘Atlantic cruiser’. This memo is dated 8/5 (1914) and signed by S V Goodall. It replied to a May 1914 telegram from the Australian Naval Board: the Commonwealth had decided to build two additional ‘Town’ class cruisers at Sydney dockyard and wanted to order material for one immediately. ‘As machinery of Brisbane is now obsolescent, it is proposed to substitute a two shaft arrangement with Parson’s impulse reaction turbines. Boilers of Brisbane type have proved to be of insufficient capacity to obtain full results using Australian coal. It is therefore proposed to fit ten boilers in lieu of twelve. Four of these to burn oil fuel only … Hull and armament same as Brisbane.’
12. With these figures radius of action would be slightly superior to that of Birmingham when burning Welsh coal, but appreciably less if burning Australian coal; E-in-C suggested carrying more oil at the expense of coal. When the ship was lengthened, oil capacity increased to about 560 tons.
13. DNO had recently recommended that all light cruisers be fitted with an alternative director position and also with a rangefinder specifically for torpedo control. Although this initiative had not yet been approved, the necessary arrangements were provided in the new cruiser. A provisional armament statement dated 1918 showed nine 6in (200 rounds per gun), one 3in HA (300 rounds), three 2pdr pompom anti-aircraft guns (800 rounds each), and eight Lewis guns in twin mountings (5,000 rounds per pair). There were two submerged 21 in torpedo tubes for seven Mk IV or Mk IV* torpedoes.
14. Vickers Estimate Book 3 (Brass Foundry) lists a Vickers design (706) dated 10 July 1914 for a second-class cruiser for Spain which may be Reina Victoria Eugenia. She is described as similar to Birmingham class, 440ft (pp), 460ft (wl) x 49ft 6in (50ft over armour) x 25ft 7⅜in moulded depth x 15ft 9in draft, 5,500 tons (normal), armed with nine 6in guns, four Maxims, one 12pdr and two submerged torpedo tubes, 25,500shp for 25.5kts. At this time Vickers effectively owned the Spanish shipbuilding industry. Armstrong was probably acting as sub-contractor.
15. Vickers Estimate Book 3 (Brass Foundry) lists Vickers designs of light cruisers for Spain: designs 730 to 733 of 10 March 1915, and designs 731 (modified) and 741 of 7 June 1915. All were armed with six 6in guns and either two 21in or four 18in torpedo tubes. They differed in displacement: 4,500 tons (730), 4,700 tons (731), 5,000 tons (732 and 733, with four 21in tubes), 4,650 tons (730 mod), or 5,100 tons (741). Mendez Nunez as built displaced 4,650 tons. Designs 730 and 731 were about the size of the Birmingham class, 440ft (pp) x 45ft 3in x 24ft 10in (depth) x 14ft, with four-shaft turbines (43,000shp for 29kts or 37,500shp for 28kts, depending on whether all boilers were oil-fired). They had 2in side plating. Designs 732 and 733 were of similar size but heavier, with coal- and oil-fired boilers. The modified version of 731 had two twin deck tubes and 47,000shp machinery. It appears to have been the design chosen.
16. Policy laid out in Captain Ballard’s (and others’) remarks on the ‘peace cruiser’ in the Cover for that design and for the ‘Atlantic cruiser’. Ballard’s comment that the foreign service cruiser was already being provided for Australia makes it clear that he had the ‘Town’ class in mind. One justification for the small colonial cruiser was that it could be built in large numbers, to deal with expected attacks by numerous German armed merchant cruisers converted from medium-sized merchant ships. In rejecting this idea, Captain Ballard referred to the conclusions of a Special Committee appointed nearly a year earlier (i.e. about February 1912), whose advice the Admiralty Board had accepted. It claimed that the best and cheapest form of protection would be to provide British merchant ships with guns on a large scale; ‘arrangements to that end are already far advanced.’
17. The Cover includes a Legend dated July 1913 for this B3 design, the culmination of the design process. The ship would have been 500ft (pp) x 52ft x 18ft 6in (fwd) and 21ft 6in (aft), displacing 7,400 tons at normal load, with freeboard 26ft forward, 12ft amidships and 13ft aft (with a long forecastle); on 30,000shp she was expected to make 26kts when loaded. Fuel would have been oil only: 500 tons normal, capacity 1,200 tons. She would have carried the usual 150 rounds per gun, plus twenty torpedoes. An unusual feature was four anti-aircraft guns. Deck armour would have been limited to 1½in over the steering gear. Barbettes would have been 4in, and gunhouses 4in and 3in. The note of 4 August from Churchill to DNC asked specifically for two 7.5in fore and aft and six 6in (centre line or en echelon), the whole to fire on either beam (B4 could fire only five 6in on either beam), with four anti-aircraft and four anti-destroyer guns and four 21 in torpedo tubes (‘rather a long ship’) with a speed of 26-27-28kts, using either coal and oil fuel or oil fuel only, with maximum 4in armour, and displacement of 6,500 tons. Cost should be £500,000 if oil only, or £550,000 burning both coal and oil. This note is Folio 2 in the Cover. The design showed anti-aircraft but not anti-destroyer guns. Magazine capacity was 200 rounds per 6in gun.
18. The same Cover giving details of the big fast commerce protection cruiser includes notes of a conference on 14 August 1912 to define the characteristics of a vessel for service on foreign and colonial stations. It in turn reflected the conclusions of a committee (Sir Francis J S Hopwood, Rear Admiral Sir Edmond J W Slade, Captain G A Ballard [Director of Operations Division], Mr V W Baddeley [Asst Secretary for Finance Duties], and Cecil Perham [Secretary]) responding to the First Lord’s minute of 25 April 1912. It was expected to report which ships on colonial duties would have to be replaced during the next eight years and propose replacements. Such ships would undertake ceremonial and other visits and police patrols, ‘but without any appreciable equipment as a fighting machine’. However, a ‘peace ship’ was of no value; the Foreign Office and the Colonial Office repeated their 1907 advice that they needed ‘warships, looking the part, flying the White Ensign, carrying several effective guns and capable of disembarking landing parties’. However, a modern light cruiser was far more than what was needed. The Committee therefore argued that it wanted an economical ship which would nonetheless have a useful wartime function, e.g. in enforcing blockades and in other duties in remote areas. She would replace the small cruisers, sloops and smaller vessels on foreign stations ‘which are rapidly wearing out’. The War Staff pointed out that a slow (17–18kt) ship would be ‘a gift’ to enemy cruisers, to which the Committee answered that in much of the world the enemy would have no cruisers (German overseas cruiser operations in 1914 suggested the opposite). Higher speed would be attractive but too expensive. To support the Committee, the August conference laid down tentative characteristics. Present at this conference were A C Lord, Admiral Slade, and DNC (d’Eyncourt). It was apparently intended to meet requirements framed by a Committee on the Design of Cruisers for Foreign and Colonial Stations, which envisaged an economical ship specifically for peacetime operations. Admiral Slade gave the requirements as: 2,000 to 2,500 tons, draft less than 14ft, speed 17 to 18kts on service (trial speed 20kts), to burn coal and/or oil, with a straight stem and a two-mast fore and aft rig with sails (for additional endurance, and as insurance if the engines broke down very far from a port). Armament would be six 4in (four on the broadside, one each on poop and forecastle) and no torpedo tubes. There would be a protective deck. The ship would carry stores for four months and would have maximum radius of action at economical speed, the figure given being 5,000nm. She would have turbine machinery. Desired unit cost was £100,000. DNC offered two guns abreast on poop and forecastle, and two in the waist, to give the maximum number the best command, and to provide more ahead and astern fire. To provide a double bottom and to minimise draft, he had to give up the protective deck above the boilers. Instead he offered a 2½in side (including hull plating, hence contributing to overall strength) extending from 2ft 3in below the waterline to the upper deck over the machinery; he considered this better than a 2in deck. The ship would have four boilers in two boiler rooms, the boilers on the centreline with stoking spaces at the ends of each boiler room. That made for a longer boiler space, but all coal would be stowed abreast the boilers, offering protection. Based on experience with recent third class cruisers, DNC expected to require 400 tons of coal and 100 tons of oil for 5,000nm at 10kts. He commented that the ship could accommodate more economical machinery, such as geared turbines. Power (5,000shp) would suffice for 20kts at 1,750 tons displacement. Estimated cost (as of 6 September 1912) was £115,000. The ship was smaller than Admiral Slade had expected, 300ft (pp) x 35ft x 10ft 9in (fwd) 12ft 9in (aft) and 1,750 tons; DNC referred to a deep draft of 13ft 6in. Deck protection was limited to 1in over magazines and ½in elsewhere, with nothing over the machinery. A further conference (on 4 October) called for enlarging the ship to about 2,000 tons and a length of 320ft to increase armament to two 6in guns (at the ends) and four 4in, the belt being carried aft to protect the steering gear, and three rather than two funnels (presumably to make the ship more impressive visually). A new Legend was accordingly developed (330ft x 38ft x 10ft 9in/12ft 9in, 2,000 tons, with the same machinery). The Committee turned in its report on 1 February 1913, the main difference from the design being a desire for 8,000nm endurance. The new War Staff was unenthusiastic. Its chief, Admiral H B Jackson (later First Sea Lord) commented that the Royal Navy currently used sloops, gunboats, and obsolete Second and Third Class Cruisers for such duties – it used ships designed for war for peacetime duties. He considered it a false and uneconomical policy to build inferior fighting vessels ‘in these days of rapid advance in naval engineering and architecture’. Only ships with very special roles, such as surveying or river work, should sacrifice fighting power. Even so, it seemed wise to develop a design for a ‘peace sloop’, which would be laid up in wartime. The idea survived until about August 1913 as a design for a replacement sloop for the Pacific.
3. Destroyer-Killers
1. On trials, Undaunted made 28.47kts at 3,480 tons (i.e. slightly light) on 38,252shp. Aurora made 28.93kts at 3,495 tons on 41,786shp. Data from DNC First World War Cruiser History. This class was designed by Stanley V Goodall, later DNC. Quoting from an unpublished lecture Goodall delivered to US constructors in 1918, D K Brown wrote that Goodall knew the ship could never make the desired 30kts, but he and DNC believed Churchill would be out of the Admiralty before the ships were completed. The reference to propeller experiments was intended to cover this reality. Churchill was not gone in time, and there was considerable embarrassment. No propeller trials were ever run. The time scale was so tight that the hull form had to be chosen before any tank tests could be carried out; the next class was considerably better. ‘The Design of HMS Arethusa 1912,’ Warship International No.1 (1983).
2. Early in April 1912 Churchill asked Controller to begin one of the Super Actives as soon as possible so that this radical type of ship could be tested before the others were complete. He added that it was essential not to underspend in 1912–13; ‘it would be much better to spend more than we have asked for and obtain a supplementary estimate [appropriation], all of which would relieve 1913-14. I must therefore see the design of a Super Active this week [written 3 April] …’
3. According to a 20 November 1912 note by DNO to Third Sea Lord, the modification was in response to (undated) verbal instructions. DNC wrote that ‘personally I always favoured this arrangement – I do not attach very great importance to the mixed-armament theory of objections in these small craft but armed as at present designed they are merely large destroyers and could not face a cruiser with one 6in gun. It is true they could probably save themselves by running away, but this is a most undesirable thing to do on many occasions of a cruiser service, although under certain circumstances the power to do so is necessary, where the conveyance of information may be of more value than to fight a successful action.’ DNC resisted on the grounds that the change would be considerable and that it would add a total of about 86 tons, costing a loss of metacentric height (about 3in) and an increased trim by the stern (about 9in), resulting in the loss of about half a knot of speed. The change was formally approved at a conference on 7 January 1913. The 6in guns were provided with 150 rounds per gun; the 4in had 200.
4. The DNC cruiser design history attributes the Falmouth experiments to a committee on light cruiser armament presided over by Captain Grant.
5. The existing twin tube design had to be modified because its racer (training circle on deck) was too large. Stowage had to be arranged for three more torpedoes. With space for only one further torpedo in the torpedo body room, two had to be stowed in the tubes, a practice previously avoided. It was not yet accepted that nothing but torpedoes already in tubes could be ready for quick use. Total outfit was ten torpedoes. Modified tubes could simply replace the singles, or twins could replace the two aftermost 4in guns (which were relocated to the former single-tube positions forward of the new twin tubes). The latter seems to have been adopted.
6. Atlantic Cruiser cover, Folio 19. Controller (Rear Admiral F C T Tudor) pointed out that neither the Arethusa nor the follow-on Calliope could be altered, but that DNC should take the problem into account in new designs. The second Arethusa Cover includes a report of her experience at Heligoland Bight, where she was hit by German shells but not badly damaged: ‘It is satisfactory to know that our shells are so much superior to the German shells in destructive effect. Ours would be still better if all Lyddite [HE] shell were fused with No. 18 detonating fuse, which I believe is unfortunately not the case.’ The ship did suffer considerable damage to electrical leads, which might better be carried below the waterline.
7. The Calliope class Cover does not indicate when or why this change was made. However, this logic was laid out in DNC’s December 1914 comments on the proposed change to all-6in armament.
8. Designed revs for full power were 650rpm in direct-drive ships, 340rpm in Champion, and 480rpm in Calliope, Champion (two-shaft Parsons pinion geared turbines) being the most successful. Champion made 29.5kts at 41,188shp (337rpm) at 3,850 tons. There was a noticeable absence of vibration, and machinery spaces were roomier and access better than in direct-drive ships. Caroline (direct-drive) made 29.1kts at 40,780shp (586rpm). As a measure of efficiency, on ⅜power (16,000shp), a direct-drive ship (Centaur class) burned 260 tons in 24 hours, but a geared-turbine ship (Ceres class) burned 200 tons. At full power a direct-drive ship (Arethusa class) burned 550 tons, a geared-turbine ship (Calliope) 435 tons.
9. CAB 37-122, issued to the Cabinet in December 1914, comparing the war programme with what would have been ordered under the 1914/15 and 1915/16 programmes. Actual cruiser orders amounted to six ships. Cuts in the 1914/15 building programme saved about £1.5 million out of the usual £16 million. The usual number of capital ships (four) and torpedo craft was unchanged. Presumably new construction was being traded off against some other expense, such as new facilities (Churchill’s Cabinet paper did not say).
10. At this time generators were rated in amps (of current) rather than, as later, in kilowatts (power). Arethusa and Calliope class cruisers had two 500-amp generators; DEE wanted two 700-amp in the new ships. He also proposed that the after capstan be electric. DNO pointed out that four searchlights would require 600 amps plus at least another 150 amps to handle fluctuations in searchlight power. A thousand amps would give no margin for ventilating fans, for the alternators which powered current British wireless sets, and for other necessary motors. As of July 1914 trials were being conducted to determine whether even larger searchlights were needed.
11. Nicholson found his ship ‘remarkably steady’ and a great advance on all previous light cruisers. He feared that raising weights in his ship would reduce that steadiness. Presumably he meant stiffness, because raising weights would have made for a longer, slower roll due to reduced metacentric height (GM).
12. The modified sketch design had been submitted to the Third Sea Lord by 5 December 1914, according to a note by DNC; the Cover dates comments by Fisher and Churchill to a few days later. DNC found the all-6in proposal attractive because the entire battery could be fought in heavy weather (and was relatively dry), because it offered the five-gun broadside, and because it made director firing easy to fit. DNC added that the existing mixed armament was best if ‘as prescribed, the ships are to be employed as Flotilla Cruisers or attached to Battle Squadrons, where one of their roles would be the attack of enemy’s destroyers, I am strongly of opinion that the 6in gun is not a rapid or handy enough weapon for this purpose.’ The DNC history describes a debate over 6in, 4in, and mixed battery continuing until the Calliope design was finally approved by the Board on 5 May 1913. It is not reflected in the Cover, but Covers of this era often omitted discussions prior to final submission to the Board. The argument favouring an all-6in battery may be what Admiral Jellicoe referred to when the question was raised again in 1914–15. Arguments cited in the DNC history in favour an all-4in battery were that it sufficed to defeat destroyers, the ship’s intended targets; that with 6in guns the ship’s commander might be tempted to engage a superior enemy instead of performing the vital scouting function; that the evils of a mixed battery on a small ship outweighed any advantage of heavier guns; and that 6in guns would be ineffective on board a small ship. The mixed battery was chosen because the two 6in aft could deal with a pursuing cruiser. The ships suffered the same spray problems as the Arethusas, and were fitted with sponsons for their waist 4in guns.
13. The forward gun in Arethusa was 98ft from the fore perpendicular, and that of Calliope was 105ft back. This was reduced to 97ft in Centaur and further reduced to 96ft in Calypso. Presumably the latter figures were acceptable in view of Arethusa. Of earlier cruisers, Fearless had her fore gun 92ft from the bow, and Melbourne (Birmingham class) 109ft back.
14. Churchill asked for a proposal for six 60pdr on the centreline on 8 December 1914. The next day DNO wrote that the only existing design was for a QF gun (semi-automatic) using a brass cartridge case. The Royal Navy was experiencing difficulties with its new 4in QF gun. A QF gun had to be tested in prototype form first, whereas a conventional gun with bagged charges could be ordered from a design. DNO therefore called for a designs of a 5in gun with a service breech mechanism (i.e. bagged charges) with percussion firing. Even then the armament firms would need time to make a new type of gun and mounting. DNO badly wanted two guns firing right ahead. The same day DNC submitted a sketch of a cruiser with six 60pdr on the centreline, presumably with two guns firing forward and two superimposed aft.
15. Ships were renamed on 4 October 1916.
16. Presumably the Germans had to fill the centrelines of their cruisers with far more massive coal-burning machinery. Guns could be mounted over engine rooms, but their supports could not be allowed to penetrate the stokeholds of coal-fired boilers.
17. In submitting the design, DNC referred to ‘yesterday’s directions for a modified “C” class cruiser with 6 x 6in guns’ but that may have been Controller’s response to DNC’s initial proposal. DNC’s remarks on the logic of such a design suggest that the initial request was broader. At this stage the ship was 440ft (wl) x 44ft 6in x 14ft 2in, 4,500 tons; Calliope was 3,750 tons (Arethusa was 3,500 tons) and the all-6in version displaced 4,120 tons. The beam was 2ft wider than in Calypso, and the hull was a foot deeper. Machinery was unchanged (40,000shp). As revised, the ship was 445ft x 45ft 6in x 15ft 5in (fwd) 15ft 3in (aft), and would displace 4,650 tons, i.e. nearly a thousand tons more than Calliope. Oil capacity in normal condition was 329 tons rather than 300 tons (maximum 1,000 tons).
18. This probably reflected experience at Jutland. A report from the Captain of HMS Conquest (Roger Backhouse, later Controller and then First Sea Lord) in the third Arethusa Cover (pp 145–8) suggested that all future ships have 1in decks over their magazines and ½in bulkheads around them, with similar protection to lower conning tower and auxiliary wireless office (the main wireless office on the main deck should have a ½in screen around it). Backhouse also wanted a 1in HT forward machinery space bulkhead like the after one. He also strongly recommended 1in decks for bridges and control positions, to stop splinters from shells bursting below them. The ship took only one or two (Backhouse thought one) direct hits. This shell struck the forward edge of the flash screen aft (separating the two 6in guns) and passed through to detonate against a 6in gun support. The explosion blew holes in both the superstructure deck above and the upper deck below and jammed the 6in gun, damaging nearby equipment. Fragments flying forward pierced as many as five light (7½lb) bulkheads. They also penetrated the lower deck and even the platform deck. Some of the holes were large (one was 3-4ft2). The fragments were stopped by the 1in machinery bulkhead, which was dented. Backhouse thought this bulkhead saved the engine room from damage as well as limiting the effect of the burst on the lower deck. The burst destroyed everything within a radius of 10-12ft and caused damage at distances up to 60ft. The fire from its flash burned men in the 6in space on the platform deck and in the 6in shell room.
19. Folio 23 in the Atlantic Cruiser Cover rather than in the ‘D’ class Cover.
20. DNC pointed out that the current bridge had been designed after full discussions with DNO, Superintendant of Charts (S of C) and DNE and also with Commodore (T). DNC also pointed out that the Officer of the Watch, particularly at night, would never leave the compass platform, so he had to have the instruments for night attack close by (gun, torpedo and searchlight control); the committee’s vision of an entirely open navigating platform was nonsense. The problem was solved in the new battlecruisers Renown and Repulse by raising the fore part of the manoeuvring or standard compass platform higher than after part, with no gunnery or other armament instruments placed in it, though they were near enough for immediate control, but cruisers had nothing like enough space for this.
21. According to a letter from a constructor on board HMS Comus dated 24 October 1916, these weights included an extra generator, the third 6in gun, paravanes and their attachments, fifty extra men and kits, extra 6in ammunition (200 rather than 150 rounds per gun), oil fittings, handing rooms, and protective mattresses (5 tons). A list of possible weight removals followed. It was headed with the comment that, if tubes were substituted for the after 4in guns, it would be better to replace the remaining 4in guns on each side with one 6in, as the value of the remaining two 4in on each broadside would be almost negligible, and certainly so at longer ranges, given their small splash (difficult to spot) and the additional time of flight. Proposals included replacement of the existing coal-fired kitchen with an oil-fired one, eliminating the 15-ton culinary coal bunker at upper deck level.
22. In May 1915 the four Arethusa class cruisers of the Harwich Force (Arethusa, Aurora, Penelope and Undaunted) were all given downward-sloping flying-off tracks for fighters intended to shoot down Zeppelins. Although the Zeppelins were often sighted, they were not engaged, and the tracks were all removed by August 1916. These platforms do not figure in the Covers, but Alan Raven and John Roberts, British Cruisers of World War II (Arms and Armour Press, London: 1980), pp 30-1 show photos of a Deperdussin monoplane on board Aurora, on 30 October and 4 November 1915, and of a short Seaplane stowed aboard Aurora (presumably to be hoisted in and out, as it did not need a ramp) in March 1915. This was a national defence measure rather than the later attempt to use fighters to destroy Zeppelins on naval reconnaissance missions.
23. For example, the list of Admiralty Board decisions for the week ending 29 September 1917 includes a decision to fit HMS Cordelia with a flying-off platform during her current refit; her conning tower was to be removed and, if time permitted, a ¾in splinter-proof structure substituted. The only other change was substitution of two twin 21in torpedo tubes for her two after 4in guns. ADM 167/54. Raven and Roberts, p 48, refer to a 1917 decision to fit all the ships of the Caroline, Calliope and Cambrian classes, the prototype being Cordelia (refitted at Hawthorn Leslie between September and October 1917). After Cleopatra was inclined with her fourth 6in gun early in 1918, DNC asked that ships which already had flying-off platforms have them removed when the fourth 6in gun was added. Between April and August 1918 the additional gun was installed on board Constance, Conquest, Calliope, Cordelia, Champion and Canterbury.
24. Mixed-battery ‘C’ class ships with flying-off platforms at this time were Calliope, Caroline, Cordelia and Comus, with Constance about to be fitted.
25. Figure given by Cooper, Farnborough and the Fleet Air Arm, quoting an official report by R H Nailer, Flying-Off Platforms 1917-34. From 1917 onwards flying-off platforms were installed on board (by classes) Weymouth and Yarmouth; Dublin, Melbourne, Sydney and Southampton; Birkenhead; all seven surviving Arethusas; Calliope, Caroline, Comus, Constance and Cordelia; Caledon and Cassandra (with hangars alongside the bridge); Carlisle (with hangar under a raised bridge); Caledon (revolving platform); Dragon and Dauntless (hangars under the bridge); Delhi, Despatch, Dunedin and Dragon (revolving platform); and Emerald and Enterprise (revolving platforms); the thirty-second platform was the one aboard HMS Arethusa, sunk before she could be fitted with the later fixed platform. According to Raven and Roberts, the supports for the revolving platforms were retained in ‘D’ class cruisers until 1927-8 because they were to have been used as supports for catapults (apparently the plan to install catapults on board these ships was abandoned at that time).
26. Unfortunately the ‘E’ class Cover has been lost (a few papers not relevant to its origins were collected in a replacement Cover). Controller’s query is in the second Calliope class cover, together with Tyrwhitt’s ideas on greater forward firepower. The Board note, but not its content, is from 1917-18 Board Minutes and Memos (ADM 167/54). The account of the design is mainly from the DNC history of warship design and construction, the cruiser part of which was dated October 1918 (ADM 1/8547/34).
27. D’Eyncourt Design Notebook, National Maritime Museum, p 53. They were given on the same page as details of a proposed submarine cruiser armed with two 6in guns. Details of the two designs were: see opposite page. Side armour includes hull plating, as in earlier light cruisers. Note that ‘E’ class weights are for the Legend, not the deep load, condition. ‘D’ class data are as designed, before the decision was taken to arm them with triple instead of twin torpedo tubes. As completed, ‘D’ class displacement was 4,750 tons, increases being 5 tons for general equipment, 5 tons for armament (triple tubes), 35 tons for protection (magazine protection), and 100 tons for the hull. The additions to equipment, armament and protection consumed the 45-ton margin. DNC’s Notebook included the abortive 5,100-ton modified ‘D’ with 295 tons of armament, 955 tons of machinery, 350 tons of oil, 440 tons of protection, 2,725 tons of hull, and 50 tons for margin. Length would have increased to 487/460ft and beam to 47ft 6in. Commenting on a 1921 Japanese attempt to buy a fast light cruiser in the United Kingdom, NID likened it to the initial ‘E’ class sketch design (A) ‘which was found to involve so many sacrifices to get the desired speed, that it was necessary to about double the displacement to obtain the desired features’. The Japanese project began when the total tonnage budgeted for in the 1920–1 programme left about 4,000 tons, which they wanted to use for a light cruiser built in the United Kingdom. They told their London Naval Attaché to make inquiries. Like the recently-built Tenryu and Tatsuta, it would be designed to lead destroyers. It should displace not more than 3,700 tons, mounting four 5.5in/50 (120 rounds per gun), one 3in/40 HA (200 rounds), and four twin 21in torpedo tubes (two pairs on each side; twelve torpedoes), with 2½in side and 1in deck armour over the machinery. Speed should be 33kts using mixed coal and oil fuel, with a radius of 5,000nm at 14kts. NID doubted that the Japanese could get the desired speed using coal-burning boilers; they would probably have to accept a lightly-built hull with little protection. Although no cruiser was bought in the United Kingdom, this was presumably the origin of the cruiser Yubari. Reports of this cruiser project included a report that the Japanese had recently bought an air flask from Armstrong Whitworth for a 24in torpedo, which would probably have a range of 12,500yds at 29kts – probably the first Western report of this calibre in the Japanese navy.
28. The discussion of protection survives (as a carbon copy, marked ‘for ‘E’ class Cover’) as Folio 21 in the ‘Atlantic Cruiser’ cover. It began with DNO’s 14 March 1918 remarks, suggesting that the design was circulated at about that time. However, it is not at all clear to what extent (if any) the May 1918 design had been modified.
29. The version for the Board is described in the volume of Board Minutes and Memoranda for 1917–18 and is also reproduced (with a drawing) as ADM 1/9223.
30. The Cover on Foreign Cruisers from 1915 onwards includes as Folio 16 a copy of an extract from the 26 July 1918 issue of Engineer, dated 6 August in file. It claimed that the latest German light cruisers, to be named Mannheim and Köln, were a completely new design, said to be 520ft x 49ft x 17½ft, 6,300 tons, 33kts, with two 8.2in BL (21cm) at ends, six 5.9in QF, four TT. Oil-fired boilers, nearly 12,000nm at cruising speed. Some light side armour, probably about 4in thick, 320ft long with the usual protective deck. It claimed that the ships were armed with 8.2in guns because that was the next most powerful gun above the then-standard 5.9in in German service, the 6.7in (17cm) having been a notorious failure. No such ships seem to have been planned; Köln being leader of a class armed with eight 5.9in. Later it emerged that the Germans were interested in a much larger cruiser armed with 6.7in guns. Raven and Roberts, p 102, describe the two enlarged ‘E’ class designs. A substituted 7.5in guns in ‘A’, ‘B’, and ‘X’ and ‘Y’ positions for 6in and omitted the waist guns and ‘Q’ gun of the ‘E’ class. It would have had much the same performance, and would have displaced 7,700 tons. B was lengthened to provide space for a fifth 7.5in gun between the funnels: 570ft (pp) 607ft (wl) x 58ft x 15ft 9in, 8,850 tons, the extra length increased fuel capacity to an estimated 2,000 tons (from 1,600). Both designs showed four 4in HA guns and the four triple torpedo tubes of the ‘E’ class. By this time it seems to have been clear that the 7.5in gun was too large for a cruiser, because it could not be hand-loaded. These designs were not developed in sufficient detail to appear in D’Eyncourt’s Notebook describing his designs as DNC.
31. Commenting on the final report of the Fire Control Requirements Committee (March 1920) DNC stated that the test arrangements for Enterprise had already been proposed on paper G.719/20. Substitution of quadruple for triple torpedo tubes had already been provisionally approved.
32. ADM 1/8397/365, first docket in file (S.0796/1914).
33. Date given by DNC First World War cruiser history. Apart from the design alternatives, taken from the D’Eyncourt Notebook (p 43), details are taken from the DNC First World War cruiser history. DNC tried four versions of the same hull (560ft pp x 64ft x 16ft fwd/18ft aft, 9,100 tons, with freeboard 25ft forward, 21ft amidships, and 15ft 6in aft. Deep draft varied: 19ft 3in in Design A, 19ft 4in in B, 19ft 5in in C, and 19ft 8in in D. Two other designs (E and F) used a slightly larger hull, 565ft (605ft overall) x 65ft x 16ft 3in/18ft 3in, 9,750 tons, freeboard 24ft 9in forward, 20ft 9in amidships, and 15ft 3in aft, deep draft 19ft 4in. All versions had 60,000shp machinery for 30kts, with 880 tons of fuel at load draft and 500/1,750 tons of oil fuel. In all the versions, coal-burning boilers accounted for ½ of total power. Because they had different armaments, the versions varied in complement, from 440 in Design A up to 470 in E and F. Protection was on typical light cruiser lines: 3in side to the upper deck (but 2in to the forecastle deck), extending down to 3ft below the waterline, with 2½–1½in side armour forward and 2½–1¼in aft, a 3in conning tower with a 2in tube, a 1in deck over the steering gear, and 2in and 1in ammunition tubes. Torpedo tubes were submerged in each version of the design. The design chosen used the same hull as alternatives E and F, but with a load draft of 19ft 3in and a slightly reduced fuel load. Instead of eight 7.5in guns and four 3in, it had seven 7.5in and ten 12pdrs, four of them HA. Details of alternatives, other than dimensions and horsepower and speed, were: see table below.
34. ADM 1/8424/171. As with the ‘E’ class, unfortunately the Cover for these ships was lost. The Minute described the (non-cruiser) programme recently approved, which included thirty-six sloops (mainly minesweepers) as first priority, followed by twenty-four destroyers, three leaders, thirteen minesweepers, and one small floating dock.
35. Principal Questions for DNO 1918, p 16, Naval Historical Branch. The reference was to a Minute by Controller dated 12 December 1917 citing an earlier query by First Sea Lord. DNO stated that in September 1917 Portsmouth Dockyard was asked to design a 6in gun capable of firing at any elevation from –5° to +90°. Existing land-based mountings were too cumbersome for shipboard use, and the 6in QF used had too low a muzzle velocity (2,150ft/sec). DNO envisaged a new high-velocity QF gun. CSOF, Vickers and Elswick were all being asked for designs of 4.7in, 5.5in, and 6in anti-aircraft guns, of which the 4.7in weapon eventually armed the Nelsons. DNO pointed out that a dual-purpose mounting would not be as effective in either HA or LA fire as a specially-designed gun, but the idea survived, eventually producing the 5.25in gun and mount. A particular problem for LA fire was the increased height of the mounting (so that it could elevate), which would make loading more difficult. DNO concluded that 6in dual-purpose guns could not form part of the armament of the next light cruiser (‘E’ class), but that it would be wise to strengthen HA positions so that they could accept such guns if and when they became available.
36. The 1923 Admiralty Memoranda volume contains a discussion (item 1712, 26 July 1923) of plans to convert Frobisher and Effingham. The alternatives were either to convert the present after boilers to burn oil at a rate corresponding to their existing oil and coal consumption; or to replace the four coal-burning boilers with two large oil-burning ones similar to those at the fore end of No. 1 boiler room. The first alternative was rejected because it would add 96 tons, and the ships were already over the 10,000-ton limit. The second was expected to save 9 tons, while increasing oil capacity by 196 tons. Endurance would increase by about 500nm at cruising speed. Further weight would be saved by reducing complement (eliminating stokers). As both ships were badly congested, saving men was welcome. Both ships were still under construction. To avoid delay, Director of Dockyards proposed removing two large boilers from Effingham and place them on board Frobisher, replacing four small boilers. Four new boilers would be ordered for Effingham. DNC observed that, had the ships been designed for all-oil fuel in the first place, they would have been more economical and satisfactory; as it was, structural arrangements did not allow the best advantage to be taken of the coal bunker spaces freed by conversion. At this time (item 1731 of 1923 memoranda) it was estimated that on completion Effingham would displace 10,120 tons; Frobisher would be similar. DNC was trying to shave weights to bring the ships within the limit. Director of Plans argued that, because the ships had been laid down in 1916, they were outside the Treaty regulations, the additions over 10,000 tons being automatic increases due to construction (ships normally gained about 80 tons per year).
37. DNO discussion of Hawkins class armament arrangement in Principal Questions for DNO 1922-23, beginning p 2621, Naval Historical Branch. The initial DNO paper was dated 6 February 1922. It referred to a conversation with Captain R G H Henderson, the ship’s former CO (later Controller). DNC claimed that in conversation with a member of his staff Henderson said that the after gun, in the worst position, could be fought on about 95 per cent of the days the ship was likely to be at sea.
38. The Naval Anti-Aircraft Gunnery Committee was asked to recommend an alternative anti-aircraft battery. On a weight-for-weight basis it could barely substitute two 4in guns for the four 3in HA. However, if all ten 3in guns were replaced, there was enough weight for a far more effective HA battery of six 4in. Ideal positions were one gun each side of the conning tower platform, one gun on the platform between the funnels, one each side abaft No. 4 7.5in gun, and one on the existing HA platform. The Committee emphasised the poor sky arcs of the existing guns. The recommendation was forwarded to DNO in May 1920, but nothing came of it. In April 1921 DGD wrote to ACNS that it was pointless to replace the four 3in, because two of them could fire star shell (that could not be done if there were only two of the heavier 4in). The rigging problem was solved by an alternative ‘fighting’ rig. The 7.5in main battery could not deal effectively with surfaced submarines, but no LA secondary armament was worth fitting: submarines were being given deck and conning tower protection which nothing short of a 4.7in gun could defeat. On the other hand, it did not seem worthwhile either to add four 4in HA and two 4.7in LA guns. It seemed unlikely that the cruiser would really find herself engaging surfaced submarines. The four 3in LA guns were useless, and the 3in HA guns could fire at surface targets.
39. Minute in Principal Questions for DNO 1923-26, beginning p 3113 (all DNO Minutes were sequentially paginated), Naval Historical Branch.
40. Lillicrap Constructor’s Notebook 5, notes dated 17 February 1930.
41. ADM 229/18.
42. ADM 229/22, describing a conference on 3 October 1939.
4. War Experience
1. According to Stephen Roskill, Naval Policy Between the Wars Vol I (Collins, London: 1968), p 214, the rule was formulated in response to a Admiralty query of 12 August 1919 as to what period of immunity from war the Admiralty could expect for reconstruction. On 15 August the Cabinet directed that in preparing the 1920-1 Estimates the services assume no major war for the next ten years. The rule was extended several times during the 1920s and was made self-perpetuating in June 1928 at the behest of Winston Churchill, then Chancellor of the Exchequer. It was abrogated in 1932 after the Japanese invaded Manchuria and directly threatened British interests in China. The rule was generally used not to veto new construction but rather to cut purchases of expendable items, such as ammunition or quartz sonar (Asdic) transducers, which would be needed in wartime. In 1929, however, Churchill directly attacked the central Admiralty assumption that Japan was the most likely future enemy, deploying the Foreign Office to deny any such threat.
2. The first memorandum on oil fuel stocks (21 May 1921) seems to have appeared in the 1921 Admiralty Board Memoranda (ADM 167/64). It emphasised the value of the fleet as a deterrent to Japan and also as a means of maintaining British neutrality in the ‘not unlikely’ event of a war between the United States and Japan. A war fleet to fight Japan would include twenty-six light cruisers (reestimated as thirty-seven in 1922; this figure was mentioned in the 1923 Admiralty memo on the required capacity of the Singapore base, item 1710 of the 1923 Board Memoranda). Fleet passage from the UK to Singapore would take forty days, assuming three days of ‘tension’ before the outbreak of war. The fleet would proceed as a unit at 16kts (capital ships steaming from Suez to Aden at 15kts). The fleet would stop only to refuel. Under the ‘Ten Year Rule’ the specified oil fuel stocks were not built up; they were among the items covered by the deficiency spending of the 1930s.
3. The many explanations in surviving papers are not credible partly because they did not match up over time. It seems clear that the Admiralty did not want to admit that the seventy-cruiser figure was associated with war in the East, probably for fear that the government might disown that scenario (as Winston Churchill induced it to do in 1929).
4. For example, in May 1928 First Sea Lord Admiral of the Fleet Sir Charles Madden laid out the disposition of cruisers he envisaged once the big 10,000-tonners were in service with the main fleet (the five Kents went to China upon completion). Eight 10,000-ton cruisers (London class plus four later ones) would form the scouting line (‘A-K’ line) of the main fleet, presumably operating with its four battle-cruisers. Another ten cruisers (eight ‘D’ class and two ‘E’ class) would work with the fleet’s destroyers. The rest of the cruisers were on foreign stations, where they would contribute to trade protection in wartime: the four Hawkins class on the East Indies station; Norfolk, Dorsetshire and three ‘C’ class on the North America station; York and Exeter on the New Zealand station; and two ‘C’ class at the Cape. Another four ‘C’ class would be in reserve. By this time British cruiser squadrons generally contained four rather than five ships.
5. The fire-control committee report is ADM 116/2068. The Committee presented its interim report on 10 August 1919 and its final report on 9 March 1920. It proved impossible to provide the desired strong representation of fleet officers. Instead the fleet was represented by the Director(ate) of Naval Artillery and Torpedoes (DNA&T). Captains of HMS Excellent and HMS Vernon, the gunnery and torpedo schools, were also consulted. Stanley V Goodall of the DNC Department signed the report. Commanders represented DNO and DTM.
6. The DCT was first proposed by Commander F Elliott (HMS Benbow). The report traced the DCT idea back to the King George V class battleships, which had a revolving armoured control tower carrying a rangefinder, control officer, and rate keeper. It failed because of inefficient training and hunting (servo transmission) and limited view; arrangements to bring the tower onto the target were, according to the report, very inefficient compared to those available in 1919. The tower was attractive because personnel remained in the same positions whatever the target bearing (otherwise they had to walk around to face the target). ‘Now that director firing is universal, and inclinometers are also being introduced, it follows that considerable duplication of “aids to spotters” is avoided if all the important control personnel can be contained in the same revolving structure, able to align their optical instruments parallel at any moment.’ For the Committee, the issue was whether gyro training control could eliminate own-ship movements in azimuth to allow effective spotting and tracking of the target. This type of gyro had recently been fitted to turrets in HMS Valiant. Existing capital ships and cruisers had a fore top for fire control with a separate control tower lower down plus turret positions. The report cited HMS Hood and recent light cruisers as examples of unacceptably-complicated aloft fire control arrangements.
7. Paper from DNO (G.0775/1920 in ADM 116/4041). The same file contains a longer paper on the armaments of future capital ships and light cruisers, a memorandum of 4 March 1920 by DNO.
8. RN practice was to fire alternating salvoes for quicker results from spotting; a ship could fire two half-salvoes in the time it would take to fire a single full salvo. The US Navy preferred full salvoes, and its large broadsides made a considerable impression on the Grand Fleet in 1918 (but did not change British practice).
9. DNC suggested that the Committee wanted to counter the new US Omaha class then being built, only four of whose eight 6in could bear on each broadside. He did not say, probably because no one in the Royal Navy knew, that they were being modified with a twin 6in mount at each end, i.e. with the same eight-gun broadside the Committee wanted.
10. Given its enthusiasm for aircraft on board cruisers, on 19 November 1919 the Committee proposed that the Hawkins class all be capable of carrying aircraft. HMS Vindictive was to be converted back into a cruiser. The Committee proposed that, like Vindictive (as a cruiser), the other ships should all have a hangar forward in place of No. 2 gun and a light steel deck abaft the after funnel to stow two two-seat seaplanes. The hangar would accommodate four Sopwith Baby seaplanes or four fighters. Nothing came of the proposal.
11. Lillicrap Notebook 4, p 6. Lillicrap was one of two cruiser preliminary designers active in the 1930s; the existence of the other designer is evident because Lillicrap was not involved in a few major designs, such as that of the Southampton class. Unfortunately no Notebooks for the other designer seem to have surfaced.
12. ADM 1/8586/70 is the summary report (27 March 1920). DNC referred to this report in commenting on the Fire Control Requirements report. The Committee, headed by Vice Admiral Richard F Phillimore, was formed in response to a Admiralty letter of 16 June 1919. The Committee was to take into account the increased effectiveness of shells against armour and also the increased effectiveness of anti-torpedo protection. It was to report on what ships should form the main fleet upon mobilisation. Evidence taken is in ADM 116/2060.
13. It seems neither DNO nor DNC was aware of the role of poor British magazine practices in the Jutland disasters.
14. ADM 7/943 is the Minutes of the committee, April 1919–June 1921.
15. The ideal system, as described by the 1921 report, would employ a stabilised director, set to the required elevation by a fuse range sight, using target vertical and horizontal angular velocities to allow for target motion. The initial sensor would be a height finder, the expectation being that the fire-control officer could estimate its course and speed when the aircraft was first sighted, at the same time that its height was measured. The computer would calculate vertical and horizontal deflections and fuse length (i.e. slant range) from height, angle of sight and relative course and speed of the target and of the wind. Fire control could be made more effective by measuring rather than guessing target vertical and horizontal angular speeds. The British army employed such a rate-measuring (tachymetric) system, but the naval problem was substantially more complicated than that ashore due to ship motion. DNO argued that no system of measurements could properly replace estimated Angle of Presentation (aircraft altering course would frustrate any tachymetric system). Unknown to the Admiralty, the US Navy employed tachymetric fire control, and the 1932 anti-aircraft committee strongly endorsed that approach.
16. Ships had rangefinders and fuse indicators. The rangefinder measured slant range to the aircraft, which was converted into height, and the vertical deflection estimated. The vertical deflection was doubled to allow time for loading and fuse-setting, and the corresponding fuse delay set. Guns fired a quick burst, whose duration was about the time of flight, then stopped so that aim could be corrected for the next burst (much as ships fired salvoes at surface targets and spotted the splashes). A ladder of fuse settings was used against a crossing target. All of the manual steps – such as reading off range, reading off fuse setting, setting the fuse – slowed the process and introduced errors. They made sense only for slow air targets not far from a ship.
17. Because anti-aircraft engagements lasted a very short time, no fire-control system could deduce aircraft motion from successive ranges, the way surface systems deduced a target ship’s motion. Instead, a measured range was combined with some instant measure of aircraft motion. The British made what turned out to be an unfortunate choice: the fire-control officer estimated speed (actually, the rate at which the direction to the aircraft changed). The US Navy of the time relied heavily on gyros, so it used a gyro (which defines a line in space) to measure aircraft motion, a method called tachymetric (literally, rate-measuring). The Royal Navy of the 1920s probably regarded any such technique as insufficiently sailor-proof. However, as early as 1919 it certainly recognised that the director sights had to be stabilised (the Committee discussed a simple two-man HA director with stabilised layer’s and trainer’s sights). On the eve of the Second World War DNO (in the last pre-war issue of the DNO publication Progress in Gunnery) declared tachymetric methods superior, and announced that they would be adopted in a few years – but the war intervened. During the war the Royal Navy tried to obtain an approximation of tachymetric operation by adding gyros to its HADT, but that apparently was not enough. There may also have been problems with data transmission from HACP to guns, the gun crews moving their weapons in using follow-the-pointer methods.
18. The four-gun requirement apparently first appeared in a submission to the Committee by Captain E Altham RN on 25 April 1919, who described his experiences under German air attack off the Belgian coast. In his view (and the Committee’s) the guns had to be centrally (director) controlled. Altham’s views were supported by Colonel Simon RE (Royal Engineers), who had been involved in the wartime anti-aircraft defence of London. To Simon what mattered was the moral effect of barrages, which would cause pilots to veer away. Moral effect depended on the intensity of the barrage, and it would be of little effect with fewer than four guns of at least 3in calibre. ‘By day a majority of pilots, unless they happen to have considerable experience, will turn from a well-placed barrage seen in front of them and at the same level or just below.’ Simon preferred fire to hit, if the target could be seen.
19. Board Minute 1537; the rearmament programme is described in Principal Questions for DNO 1922-23, Minute 451 (p 2804), Naval Historical Branch.
20. Torpedo bombers typically attacked at 1,000yds or closer in. In February 1920 the Committee remarked that that longer launching ranges probably would not be adopted in future, because extra torpedo weight, which a higher-performing aircraft could support, would best be used for more explosive. Fire would probably be opened at 3,000yds. The defence would rely on cumulative hits. Unlike a 4in gun, the weapons would not have flat trajectories, and would depend on simple sights. The Committee also considered the ‘wall of water’ defence that several navies tried, in which LA guns fired into the water at ranges between 1,000 and 3,000yds. The contemporary British tactic was to attack in a group of six aircraft, which worked into a position ahead of the target, then split into threes to attack from both bows (so that evasion became difficult or impossible).
21. The idea for the multiple pompom came from Captain C V Usborne RN, president of the committee, in 1919. This ‘multiple Pom-Pom mounting’ (later called the Mk M) should carry the maximum number of 2pdrs on a total of guns and mounting not to exceed 2 tons 12 cwt (the weight of a 3in HA gun), to achieve not less than sixty rounds per minute per barrel. It was to elevate and train manually at 15°/sec while firing, with maximum elevation of 45° and maximum depression of 10° – consistent with the emphasis on torpedo bombers and attacking boats. Control would be either by hand or by power, in the latter case using a continuously-running electric motor. The mount would be locally- or director-controlled (director control could be waived if it proved too difficult). If no more than six barrels could be accommodated, a further design was wanted with double the weight (5 tons 4 cwt). The lighter weight was intended to fit a destroyer, which was typically armed with a single 3in HA gun. The specification is in the minutes of the Naval Anti-Aircraft Gunnery Committee. In November 1920 trials were held to decide whether the mounting should be arranged so that fire from its guns converged at some set range between 1,000 and 3,000yds. At this time the mounting was expected to carry six to ten pompoms. The weapon as built used a new type of pompom gun (but with the existing ammunition and the same 1,900ft/sec muzzle velocity) and its octuple mount weighed far more than expected: 15 tons. In 1932 the experimental quadruple mount was expected to weigh 6¼ tons. Both versions were clearly unsuited to destroyers and to the surviving First World War cruisers. Rate of fire was better than required (ninety rounds per gun per minute) and maximum elevation was 80° rather than 45°. In 1928 a Mk M Pom Pom Committee recommended a further increase to 120 rounds per gun per minute. The specification required at least two minutes of continuous fire without reloading. The design selected offered 1½ minutes, but ammunition trays could be reloaded while it fired, for about 2½ minutes of continuous fire at ninety rounds per gun per minute.
22. ADM 1/8685/151.
23. Admiralty Board Memoranda for 1926. The Royal Navy seems to have been unique in planning to fly fighters and even bombers from its surface combatants.
24. Performance: E.I.H (nine made), 8,000lbs at 50kts; E.II.H (eight made) 8,000lbs at 50kts; E.III.H (ten made): 8,000lbs at 57kts; E.IV.H (three made). Of the latter, one was installed on board HMS Effingham but a second, intended for HMAS Perth, was never installed. There was also an S.I.H: 8,000lbs at 49kts or 6,000lbs at 54kts. Light catapult: S.II.L, 5,500lbs at 50kts. Data from Geoffrey Cooper, Farnborough and the Fleet Air Arm (Midland Publishing, Hersham: 2008). Because launch acceleration was limited to 2.5G, end (launching) speed depended on the length of the catapult. British catapults telescoped to greater length to achieve higher end speeds: 50kts for 43ft, 52kts for 46ft, 54kts for 49ft, 57kts (actually 56.5kts) for 53ft. The heavy vs. light distinction applied more to the energy available, hence the weight of the aircraft. In February 1931 it was estimated that the light catapult (46ft stowed length) weighed 38 tons, compared to 54 tons for the Leander type heavy catapult (53ft stowed length), in each case plus 19 tons of structure.
25. Performance: D.I.H: 8,000lbs at 56kts, upgraded to 12,000lbs at 56kts. D.IV.H: 12,000lbs at 60kts, upgraded to 15,000lbs at 60kts.
26. Dates of catapult removal (by classes) prior to Fiji class: Emerald (March–April 1944), Enterprise (February 1944); Australia (date not known), Berwick (May–August 1942 refit), Cumberland (date of removal not known), Devonshire (May 1943–March 1944 refit), Kent (July–November 1942, may already have had catapult removed, leaving a seaplane and a crane), London (December 1942–May 1943), Norfolk (March–May 1943), Shropshire (November 1942–June 1943), Suffolk (December 1942–April 1943), Sussex (by December 1943); Ajax (mid-1941), Achilles (April 1943–May 1944 refit), Leander (June 1941, replaced late 1941, removed late 1943), Orion (July–August 1941); Hobart (completed October 1942), Perth (received catapult previously aboard Ajax in July 1941, sunk with it on board); Birmingham (April–August 1943 refit), Glasgow (June 1944–May 1945, when ‘X’ turret was removed), Liverpool (August 1942–July 1945, when ‘X’ turret was removed), Newcastle (October–November 1942), Sheffield (March–June 1943); Belfast (August 1944–May 1945 refit). Ships not listed were sunk with their aircraft on board. These data are from Raven and Roberts, Appendix 2.
27. ADM 1/8653/266, printed paper on Empire naval defence (February 1921), p 15.
28. ADM 1/8653/266: Light Cruiser Construction Programme, PD 01914/1923, a docket including 1921–3 material. The paper, dated February 1921, was titled Empire Naval Policy and Co-Operation.
29. It would be impossible to provide against both; the best the Royal Navy could do would be to match the strongest other naval power. ‘The worst situation with which the British Empire could be faced, from a naval point of view, would occur if Japan seized the opportunity of aggressive action in the Pacific at a time when the situation at home was threatened from another quarter, and reinforcements capable of dealing with the whole of Japan’s main force could not immediately be spared … Happily this extreme case is improbable …’ (p 11). This would be exactly the case if Japan seized the opportunity presented by a war with the United States. Presumably it was assumed that war with the United States was grossly improbable (at this time France and Italy were considered friendly and Germany had been disarmed).
30. PD 01620/21 of April 1921. Dreyer’s remarks were dated 7 May 1921.
31. These big cruisers were apparently inspired by the Hawkins class (the US standard gun calibre closest to 7.5in was 8in). The US wartime naval staff in London, led by Admiral Sims, accepted the Admiralty view that the Hawkins class were freaks, and that something like an ‘E’ class cruiser was ideal. The General Board war planners, focussed on the Pacific after the Armistice, preferred the larger cruiser. The London staff came home hoping that their organisation, the Office of the Chief of Naval Operations, would take over programme planning from the General Board. US Navy preference for large cruisers was a major consequence of their defeat.
32. A thickness of 2in was chosen on the basis of recent test firings against the German light cruiser Nürnberg. There should be increased thickness over magazines.
33. Dreyer associated the new DCT with medium ranges, hence wanted a spotting top and director tower carried as high as possible. This idea went nowhere.
34. Remote-controlled explosive motor boats had been a prominent feature of the German defences of the Flanders coast, but it is not clear why the British thought the US Navy was so interested in such weapons. A few years later the Royal Navy worked on a sort of ultimate distance-controlled boat, a semi-submersible controlled from an aircraft, intended to penetrate enemy harbours.
35. Surviving Constructors’ Notebooks do not appear to include any more detailed basis for these conclusions. No relevant designs appear in the d’Eyncourt Notebook. DNC’s comments were dated 30 June 1921.
5. Treaties and Heavy Cruisers
1. This clause was included in the proposal examined by the US Navy’s General Board beginning on 12 September 1921. General Board papers on the Washington Treaty in William V Pratt Papers, Naval War College. Admiral Pratt was the US naval adviser at the Washington Conference, and he was Chief of Naval Operations during the 1930 London Conference; in effect he was the US Navy’s arms-control expert (and exponent). In the pre-Conference proposal cruisers were included in a more general category of auxiliary surface combatants, and not considered as a category on their own. As listed in 1921, the US Navy had 213,550 tons of cruisers built and building and the Royal Navy had 358,620 tons, but the US figure was dominated by obsolete armoured cruisers (only 75,000 tons of new Omaha class cruisers were underage). Because the auxiliary category also included destroyers, the US Navy in theory balanced the tonnage of British cruisers with the enormous tonnage of new US destroyers, so the total of cruiser and destroyer tonnage was 592,940 tons for the US Navy and 510,390 tons for the Royal Navy. The General Board proposed totals of 500,000 tons for each navy (and 300,000 tons each for Japan, France and Italy). It also proposed that cruisers (like capital ships, whose replacement age was later increased to twenty years) become overage at fifteen years and destroyers and flotilla leaders at twelve. Replacements for overage ships were exempted from the ban on new construction, and each navy was allowed to build new ships to provide an underage fleet of the proposed size. Overage ships could be replaced on a ton-for-ton basis, but no keels could be laid until the older cruisers were fifteen years old. The proposed treaty included a ban on cruiser guns of more than 8in calibre (initially the General Board rejected any limit on battleship tonnage, but the limit on cruiser guns was intended to prevent capital ships from being built in the guise of cruisers). The limit on cruiser guns does not occur in the General Board paper on the treaty, but it does occur in a copy of the draft treaty in a volume of pre-Conference papers in the Pratt papers.
2. British positions are taken from ADM 1/8630/142, the collected Minutes of the British Empire Delegation, 13 November 1921 through 31 January 1922. These were meetings 48 to 73 inclusive.
3. Capped shells were most effective when they hit nearly perpendicular to the target’s armour. Remarks are in ADM 1/8653/266, the docket on light cruiser design 1921–3. In June 1923 DNO asked for a staff ruling on the requirements to be met; his query is DNO Minute 499 (p 3194) in Principal Questions for DNO 1923-26, Naval Historical Branch. In 1920 the requirement for 7.5in AP shell had been laid down: it had to be in a fit conditions to burst after perforating a 4in plate at 30° to the normal at 1,100ft/sec (equivalent to a range of 15,000yds; DGD minute, 9 December 1920). Trials in 1918–19 showed that this requirement could be met by a 7.5in SAPC shell; but the same shell could not penetrate a 6in C (cemented) plate at 20° at 1,729ft/sec (5,900yds). DNO said that he understand these requirements had been kept low because existing light cruisers did not have armour more than 3in thick. He reported that current 7.5in AP trials showed that one round nearly succeeded in penetrating a 5in C plate at 30° at 1,602ft/sec (equivalent to 6,500yds range). DGD asked Naval Intelligence for details of the side armour of the new Japanese 10,000-ton cruisers, but nothing was available (data were available for other foreign 10,000-ton cruisers). DNC estimated that unless armour was concentrated over a very small area, leaving most of the ship unprotected, maximum thicknesses were 4in side and 2in deck. The maximum given for any foreign ship was 4in side and 2½in deck. DGD suspected that, compared to British practice, the Japanese would add deck protection at the expense of side protection. They actually used a 3.9in belt and a 1.4in deck in their new Myoko class. DGD concluded from the 7.5in shell tests that any 8in APC (i.e. capped) or SAPC shell could penetrate 4in side armour at the required ranges, so effort should concentrate on deck penetration. If (as incorrectly expected) the Japanese sacrificed side for deck armour, the latter could be as much as 4in thick over the vitals, probably non-cemented (NC). He therefore proposed as a Staff Requirement that the new 8in shell penetrate 4in NC plate at about 20,000 to 25,000yds (i.e. at a 50° angle at 1,100ft/sec). Recent trials showed that the latest British APC and SAPC shells were at least as good as specially-designed ones at oblique angles (60° to normal). He therefore proposed that the shells be capable of penetrating (in a fit state to burst) a 6in C plate at 30° angle at 1,450ft/sec (equivalent to 7,000yds). DNC argued that a 4in plate could indeed defeat an 8in shell arriving at an angle of descent of 40° (i.e. striking a deck at 50° to the normal). However, it was most unlikely that so thick a deck could be provided over both magazines and machinery; it would probably be limited to magazines. Even 3in armour would suffice. Moreover, the 1,650ft/sec shell would strike side armour at an angle of descent of about 7° (i.e. about 83° to the normal) and would be unlikely to penetrate even a 1in deck at such a range. There was a real possibility that proof-testing shells at very oblique angles would cause ricochets which would ruin the tests. DNO generally agreed with DNC, and therefore suggested reducing the target thickness to 3in, but keeping 4in as a goal. As more information was received, the situation became less difficult. In November 1924 DGD remarked that data on foreign cruisers showed that all of them sacrificed protection either for speed or for armament; semi-official reports of US and French designs showed no armour at all. The Japanese reportedly carried twelve (actually ten) 8in guns, for which they had sacrificed protection, and the Italians had sacrificed protection for speed. Some earlier 6in cruisers were actually better armoured. DGD therefore returned to his earlier proposal, that shells be designed to penetrate a 4in side and a 2½in deck. He thought that side penetration would be relatively easy at all ranges out to those at which the shell struck at 40° to the normal. However, a 2½in deck would be a more difficult proposition. He wanted penetration at 20,000yds (30° angle of descent, 1,150ft/sec). DNO pointed out that the velocity actually required would be 1,450ft/sec and the actual critical velocity 1,325ft/sec, though at the angles envisaged the shell might fail to ‘bite’ the armour. Failure against 2½in deck plate was particularly disturbing in view of reports (which were wrong) that the later Japanese cruisers had such a deck. One solution was to reduce 8in muzzle velocity so that the shell would fall at a steeper angle, for better deck penetration. Discussion continued through 1926; manufacture of shells for the Kent class had to begin in June 1926. DNO recommended a high-capacity (5 per cent) SAPC shell for all later 8in cruisers. They would be the only anti-ship shells carried, replacing the earlier CP and HE shells. An additional consideration was blast and splinter effect when the shells hit the large unarmoured parts of foreign cruisers. Undue attention to penetrating relatively thick armour might reduce general damaging effect (this applied to a choice between 5 and 6 per cent shells). In the end, ACNS decided to hedge against a possible change of policy by foreign navies (which certainly occurred in the United States, France and Italy) towards better protection by adopting the 5 per cent shell. CNS (First Sea Lord) approved this decision on 9 June 1926. All British heavy cruisers received the 5 per cent shell.
4. In July, DNO wrote that sketch designs of twin turrets would probably cost about £2,000 each unless there was an immediate prospect of orders. A detailed version of the preferred design would cost about £15,000. The budget contained no surplus money, and it was ‘not safe’ to ask for more. He therefore proposed a preliminary step, to develop a general specification so that turret designs could be requested in November or December if money was available (by then there would be a sketch design of the gun). Money for a prototype gun could be included in the 1923/4 Estimates.
5. At Jutland uncovered charges had been piled up in the turrets and in the working chamber half-way down the barbettes, creating a powder train from turret to magazine.
6. The deck thickness was based on built-up construction; presumably it could have been reduced had the deck been in a single thickness. The character of the side armour (face-hardened or homogeneous) would depend on whether an enemy used capped or uncapped shells.
7. Lillicrap Constructor’s Notebook 4.
8. For some reason this reference to Admiralty Memo 285-B (which seems not to have survived) appeared only in September 1922 remarks by Director of Plans. Pound drew up tables showing ships due for replacement under the two standards, with two ships due in 1923 and ten in 1924 under the eight-year standard. If the existing total was to be maintained, a total of thirty-three ships would have to be built in 1923–32: one in 1925, seven in 1926, six in 1928, two in 1930 and five in 1932. Allowing a fifteen-year life, replacement could begin with two in 1926 and two in 1927.
9. Admiralty Board memo from Plans Division, 11 June 1926, on replacement of cruisers, destroyers, submarines, and twin-screw minesweepers, 1926; given the Ten Year Rule, the objective was a fleet at full strength by 1 April 1936, including seventy cruisers. One advantage of choosing this particular age was that replacement construction could be spread out. At this time the Royal Navy hoped for an annual programme of three cruisers, nine destroyers, and six submarines, leaving it in 1936 with five overage cruisers. On this basis one cruiser could be scrapped in 1927 and two in 1928.
10. Director of Dockyards pointed out that the two ‘E’ class cruisers could not be completed using 1923/4 funds, but could only be advanced while Effingham and Frobisher were completed. Completing all four cruisers would delay the carriers Eagle and Hermes as well as leaders, destroyers and submarines left over from the wartime programmes.
11. Lillicrap Constructor’s Notebook 4, instructions dated 2 November 1922.
12. Lillicrap’s notes include a table of length to depth ratios, which were generally about 16.4 to 17.3 in cruisers, but 14.88 for Hawkins, 14.6 for Adventure and 11.9 in the new battleship. The lower the ratio, the greater the hull depth. DNC approved a higher stress, 10–11 tons per square inch. Lillicrap later listed two more advantages of the deep hull: it made the adoption of longitudinal framing under the top deck easier, as height was available for good transverse bracket connections; and it made for good ventilation to the mess decks, important for a ship operating in the tropics.
13. The minelayer was 500ft x 59ft x 14ft (6,480 tons), so he scaled to 10,120 tons (580ft x 68½ft x 16¼ft). He estimated speed on the basis of the resistance curves already developed for the minelayer. Resistance (effective horsepower per ton) was measured on a scale of speed divided by the square root of length, so Lillicrap could estimate that the minelayer speed equivalent to 34kts cruiser speed (at a length of 580ft) would be 31.5kts (30.6kts, equivalent to 33kts for the cruiser). For 30.5kts the minelayer needed 34,800ehp. Hulls of the same shape needed the same effective horsepower per ton at the same scale speed, so Lillicrap scaled up to the cruiser on the basis of tonnage. He then converted effective horsepower into the shaft horsepower the ship actually produced by estimating a propulsive coefficient. Scaled-up figures were 101,200shp for 33kts and 116,800shp for 34kts; Lillicrap modified these figures slightly. This was standard practice for preliminary designers in the DNC Department, as illustrated by various Constructors’ Notebooks.
14. DNC did not mention the Hawkins class lines, which were based on those of HMS Furious, similar to Courageous.
15. DNO pressed for a dual-purpose 8in mounting in a comment on a proposed design for a 6in HA mount: Principal Questions for DNO 1922-23, DNO Minute 454, p 2831, Naval Historical Branch. The first item in the file was dated November 1922. Designs had been proposed by Vickers, by Elswick and by Woolwich; they emphasised HA performance at the expense of LA. DNO observed that there was no current requirement for such a mounting, and asked whether it would not be more far-sighted to abandon such mountings in favour of ones which were suitable for both kinds of fire. Such a design would be very different, because for LA fire it would have to be arranged for a relatively small angle of elevation for loading. The reverse would be true of an HA mounting – in either case the time to come to the required elevation would be minimised. However, providing a mounting with two alternative loading positions was rejected as far too complicated. All-angle loading was also rejected as too complicated. It entailed long swinging loading arms, power ramming on the mounting (to be usable at any elevation), and tilting and swinging trays for shell and powder. True HA operation required a QF gun, partly so that shell and charge could be loaded together and partly because the alternative would entail providing water to clean out the breech after firing (which would be unacceptable for HA firing). A LA gun could use either BL or QF ammunition, but the Royal Navy had elected to use BL. The limit for a fixed round was apparently the new 4.7in destroyer gun; a 6in QF gun would have to use a separate cartridge case. A dual-purpose gun would probably have to be QF because it would have a simple breech for a higher firing rate (and also because the cartridge case would make ramming easier and surer). A QF gun would also be lighter than a BL gun. The Royal Navy did not adopt a 6in QF gun until after the Second World War, and even then it encountered problems of reliability in making such large cartridge cases. For the moment, DNO favoured a dual-purpose 6in mounting in which LA fire would receive priority. DNO’s analysis killed the proposal for a dualpurpose secondary battery for the battleships then being designed (which became the Nelsons). They were accordingly armed with 6in LA secondary guns in twin mounts and with separate 4.7in antiaircraft guns. DNO recognised that, given the limitations in the Washington Treaty, future light cruisers would be armed with 8in rather than 6in guns, and that it was most unlikely that they would add 6in anti-aircraft guns. He did think that if a dual-purpose 6in mount could be developed (for battleships) ‘it would form the armament of a very attractive class that would serve the Fleet well, both against surface attack and against air attack’. DGD considered it unlikely that a heavy anti-aircraft mount could follow the motion of an aircraft nearly right overhead, when the bearing of the aircraft changed most rapidly. Nor was fire control likely to be effective in that case. Most anti-aircraft firing occurred at angles below 50°, so total time under fire at elevations over 70° would be very small. DGD therefore suggested that a future dual-purpose gun should elevate only to about 75°. Allowing for vertical deflection, that would suffice for elevation angles up to 70°. A few guns would be needed to fill the blind spot immediately overhead. To this end a new gimbal mounting was to be developed. DGD therefore proposed that the 8in guns of the new light cruisers be dual-purpose weapons elevating to 75°. DNO concurred (20 December 1922), adding only that in the 8in mount LA fire should be given first consideration, and that nothing should be done to reduce the rate of fire at low angles. For this reason DNO asked for triple or twin mountings with a fixed (relatively low) loading elevation. He originally called for 50° maximum elevation; firms were asked to report whether it would be much more difficult to arrange for a greater maximum elevation.
16. Machinery weight was based on E-in-C’s estimate of space and weight required. In this case it was based on ‘E’ class machinery. E-in-C proposed three boiler rooms to house his ten 10,000shp boilers (two each 42ft long, one 26ft long) and two 54ft engine rooms, for a total machinery length of 226ft, compared to 221ft for the much less powerful machinery in a Hawkins. Typically powerplant weight was divided into an ‘E-in-C weight’ proportional to power (in this case, 1,826 tons) and an auxiliary portion whose weight depended on factors such as generator power, for turrets and searchlights, in this case 241 tons; Lillicrap rounded the total of 2,067 tons up to 2,100 tons. To take into account DNC’s acceptance of higher stresses, he divided hull weight for a known ship into strength and other elements, then handled them separately. In this case he used the most recent cruiser design, the cruiser-minelayer Adventure. Hull weight estimation was apparently difficult; Lillicrap’s Notebooks show several quite different techniques. Typically the next stage was for junior constructors to develop hull weight in detail.
17. This Design 2 (there is no record of Design 1, but it was probably X/Y/Z) was approved on 15 October 1923 (presumably by DNC, for submission to the Board).
18. As submitted in October 1923, the Legend offered 1,025 tons of protection, achieved mainly by shaving total machinery weight (including auxiliaries) to 1,850 tons. Machinery had 1in bulkheads at the ends, plus the deck and side. Main magazines and handing room sides had 4in armour, with a 3in deck overhead (2½in of NCD and ½in of D steel). Shell rooms, which were adjacent, had 1in decks and sides and ends, because they were far less likely to explode if hit. The midships (anti-aircraft) magazine had 3in sides, 2in ends, and a 2in deck. Steering gear was protected by a 1½in turtle deck and a 1in bulkhead.
By this time it was clear that turret weight would grow, so DNC allowed 1,050 tons. Equipment was 675 tons, and hull weight was given as 5,400 tons. No margin was allowed. Page 75 of d’Eyncourt Notebook, National Maritime Museum. When the Legend was submitted, E-in-C machinery weight (i.e. without auxiliaries) was 1,700 tons, but E-in-C hoped to reduce that to 1,570 tons, using faster-revving propellers (300rpm), which DNC considered acceptable.
19. HMAS Australia and Canberra were completed with the raised funnels
20. A sketch dated 19 October 1923 showed two boiler rooms (each 45ft long) and two engine rooms (each 48ft); there was some hope that the boiler rooms could be shortened to 45ft each. It appeared that 80,000shp could be accommodated on the same 186ft length. Estimated weight was 1,850 tons, compared to 1,550 tons for the same output in the lightweight ‘E’ class plant. However, in December E-in-C thought he could cut machinery weight (wet) to 1,700 tons. This figure did not include new compressors wanted for new-generation torpedoes, which would add another 4½ tons.
21. Lillicrap Constructors’ Notebook 7, in text apparently written much later for part of a paper. He dated the decision to 1925, but it was probably considerably earlier. The Italian navy was considered the leader in ultra-lightweight machinery, designed for high overloads (in the large cruisers Zara and Fiume, the design figure was reported as 76,000shp, but a designed-in overload of 95,000shp was reported). Naval Intelligence (NID) reported a 1930 Italian statement on lightweight machinery: ‘There is a vast difference between a cruiser which works near its base in the Mediterranean and a ship that has to steam to the other end of the world and to be more or less self-supporting as is the case with British ships, which ships must naturally therefore be of a more robust type and of less speed.’ The previous year the first Italian treaty cruiser, Trento, reportedly experienced trouble due to excessive forcing of her boilers, i.e. of trying to get too much power out of them. On the other hand, her 150,000shp turbines seemed to be much smaller than the 80,000shp turbines of HMS London. In 1931 NID reported that the standard practice of paying one million lire for every knot above contract speed had recently been abandoned because to get this bonus firms forced the engines too much on trials. In 1932 NID reported failure of the machinery of the two heavy cruisers built in Italy for Argentina.
22. The main items in the action load were two 50-ton pumps (for the bulge), one 5-ton feedwater pump, two 4in submersible pumps, eight 25in fans for the engine room, and seventy ship-ventilating fans. Turrets used hydraulic power.
23. The ships were completed without transducers or other gear. As of April 1926, with the Kent class under construction, tests had been conducted in the Atlantic and in home waters, and it was considered desirable to gather operating data in the China Sea and in the Pacific. No Asdic-equipped destroyers were operating in those areas, but all the new cruisers were to go to the China station. The problem was considered particularly urgent because underwater sound (as in Asdic) was also being developed for communication, particularly between submarines, which would have an important wartime role there. HMS Suffolk was therefore ordered to be equipped with Asdic. Installation had been deferred pending a determination as to whether sufficient weight was available within the 10,000-ton limit.
24. In mid-1924 DNO pointed out that the ships would probably operate on trade routes in wartime, their opportunities to replenish ammunition being few and far between; they therefore needed maximum ammunition stowage. He wanted 300 rounds per gun (outfit and reserve), the reserve presumably being held ashore. DGD pointed out that the new 8in gun would fire about as fast as existing 6in guns – and 6in cruisers typically found their 200 to 230 rounds per gun inadequate in wartime. ACNS approved the 100 rounds per gun peace allowance, another 50 rounds per gun being held on each station to be embarked under ‘abnormal’ conditions. Controller cited the terms of the Washington Treaty. If it were found that the 10,000 tons would be exceeded if the ships carried more than 100 rounds per gun. The outfit should be 100 rounds and 200 rounds reserve, and that this action should be kept most secret (his emphasis), the instructions for embarking the extra 50 rounds ‘should be in the form of sealed orders to the Captain only to be opened on receipt of the [war] warning telegram’. The bays for the extra 50 rounds should be removed and retained with the reserve ammunition until needed. Any decision was deferred until it became clear how critical weight would be. DNC (now Berry) wrote that he doubted there would be a problem; ships would be designed to take all 150 rounds, with reductions to be made only if they proved overweight. The episode suggests just how difficult it was to design a ship within the 10,000-ton limit. Similarly, the peace outfit for 4in anti-aircraft ammunition was 150 rounds per gun, the war outfit 200.
25. According to the 1927 report of HMS Vernon (torpedo school), ideally the cruiser torpedo should have the same 750lb warhead as the battleship torpedo, with long range (but not as long as the battleships’ 20,000yds) at 35kts, with the ability to be angled and with deep running and ‘W’ (pattern-running) gear. Mk VII was a single-speed torpedo, originally oxygen-enriched (57 per cent oxygen), but converted to air operation early in the Second World War. It was replaced by the Mk IX as the opportunity arose, only nine being fired during the war according to John Campbell, Naval Weapons of World War II (Conway Maritime Press, London: 1985), p 84. Range with oxygen was 16,000yds at 33kts (with natural air, 7,800yds at 35kts). The natural air figure is from a note in the London Cover; Campbell, p 84, gives a range of 5,700yds. In 1927, when heavier torpedo batteries were being considered, Mk VII was rated at 16,750yds at 35kts with a 750lb warhead. The Mk V of the Kents was a conventional-air heater torpedo with 25kt and 35kt settings, with nominal ranges of 8,000 and 14,000yds (as given in test specifications in the 1929 Torpedo Manual). In 1939 the means of gaining greater range was higher-pressure air, called HA air. The Mk IX** HA torpedo offered 10,500yds at 40kts and 16,000yds at 35kts and a duplex (magnetic) pistol which Mk VII lacked. Many British cruisers carried Mk IX torpedoes converted to natural air.
26. PD 01912/23 of 27 October 1923 in ADM 1/8672/227, docket on Emergency Cruiser Programme. This paper referred to justification of the seventy-cruiser goal on other papers, not included. As of 1924 the Admiralty goal was to build up the seventy-cruiser force by 1929, presumably in accordance with the Ten Year Rule.
27. As described in Admiralty minutes and memoranda for 1923, the peace fleet comprised three battlecruisers, a large aircraft carrier, three Hawkins class cruisers, four ‘D’ class cruisers, four other cruisers (probably also ‘D’ class), eighteen ‘V’ class destroyers, twenty-one ‘L’ class submarines, two destroyer depot ships, three submarine depot ships, five sloops, fifteen gunboats, and twelve auxiliaries. The war fleet to operate in the East would comprise twelve battleships, three battlecruisers, four carriers, thirty-seven light cruisers, eighty-one destroyers and leaders, forty-two submarines, and numerous subsidiary ships including armed merchant cruisers. Presumably the war fleet included the peace fleet (the Royal Navy had only four battle-cruisers, for example).
28. Roskill, Naval Policy Between the Wars, I, p 428.
29. Roskill, Naval Policy Between the Wars, I, pp 419ff. Roskill does not mention the cruiser connection, but it is evident in DNC’s explanation of the 6in cruiser design he presented to the Board in 1926. He referred to the 1924 arms-control project. The Admiralty docket is ADM 1/8683/131. It includes the usual comment that the Admiralty would welcome a reduction in cruiser numbers as long as it retained enough for the scouting line of the main fleet and for trade protection. It stated that the attempt at cruiser limitation at Washington was abandoned ‘because the other signatories were not prepared to recognise our special needs in cruisers’. ACNS offered to cut cruisers to, say, 7,000 tons and 6in guns (presumably he had the ‘E’ class in mind). DNC (W J Berry) wrote on 25 April 1924 that a new 6in cruiser design should carry eight 6in guns in pairs, have a speed of 32-33kts, and an oil capacity of 1,800 tons, with a standard displacement of 7,500 to 8,000 tons, with nothing but the lightest protection, and destroyer leader machinery, with something more than half the endurance of the 10,000-ton cruisers. DCNS pointed out that the 8in limit had been adopted at the request of the United States, and that it was unlikely to be changed.
30. Lillicrap Constructor’s Notebook 4, entry dated 23 September 1924: 70 tons could buy a 3in deck over the 8in magazines (23 tons for 1in). The plating had to extend all the way to the ship’s side, so that no shell could pass through an unprotected deck to hit the magazine side. That increased weight to 50 tons per inch of armour, so only 2in could be provided.
31. The Legend for ‘Light Cruiser A (light cruiser carrying aircraft)’ was sent to the Board on 29 October 1923. It may be the ship involved. The Legend in the Notebook uses HMS Hermes, a carrier (sometimes described as an aircraft-carrying cruiser) for comparison. The ship was 530ft (pp), 581ft (overall) x 76ft (extreme, at flight deck) x 15ft (fwd), 18ft (aft), 9,750 tons standard, with 40,000shp machinery (26.5kts at load draft) and a fuel capacity of 2,000 tons and a complement of 720. Armament was given as seven 4.7in HA (240 rounds each) and eight single pompoms (800 rounds each). Protection would have been limited to side armour: 100lbs (2½in) over machinery, 80lbs (2in) over magazines and bomb rooms, and 60lbs (1½in) over steering gear. In contrast to a cruiser, the ship had very limited armament (275 tons), but her equipment weight (855 tons) clearly included aircraft. Machinery was somewhat heavy for a cruiser (1,060 tons) with limited power, and the hull weighed 7,000 tons because it included hangar and flight decks. Armour amounted to only 460 tons. D’Eyncourt Notebook p 74, National Maritime Museum.
32. Lillicrap shaved the 320 tons of deck protection to 230 tons, the 96 tons of side armour to 24 tons, and end bulkheads from 17 tons to 14 tons, presumably because machinery protection was the least important. Weakening machinery protection required some increase in the armour over the midships magazine, which otherwise benefit-ted from that armour (84 tons rather than 63 tons), but Lillicrap still saved 245 tons on machinery protection. The torpedo battery cost a total of 58 tons, including compressors and spare torpedoes. The turtle deck over the steering gear weighed another 40 tons. Eliminating both gave a total saving of 343 tons, enough to make the triple-turret ship feasible in weight terms.
33. The testing tank at Haslar used models to estimate the power needed to drive a hull of a given displacement at a given speed. This was effective horsepower (ehp). Only part of the power a ship’s engines produce (shaft horsepower for a turbine ship) actually goes into propelling the ship; the ratio of ehp to shp is propulsive efficiency (typically about 50 per cent). Thus, given a Haslar ehp figure, a designer would double it to estimate the power output he needed, although in practice he might need a bit more or less.
34. Protecting magazines with 3in deck, 4in side and 3in bulkheads would cost 182 tons; similar protection to handing rooms (which shared bulkheads with the magazines) would cost another 310 tons, and providing the same protection to the 4in magazine would cost another 66 tons. This combination would be 70 tons over the 10,000-ton limit, counting the increased machinery weight. However, it might be possible to shave 50 tons from the machinery weight (to 2,300 tons plus 140 tons for machinery) and to save 22 tons by shaving gunhouses from 1in to ¾in. At least on a weight basis, the design was feasible.
35. Lillicrap Constructor’s Notebook 4, 10 November 1924.
36. Lillicrap’s 8,500-tonner was 535ft x 58ft x 16½ft, carrying about the same protective weight as a Kent (1,090 tons). She was expected to make 31.8kts. She needed 2,400 tons of oil to make the desired 8,000nm at 12kts, but could stow only 2,200 tons below the platform deck. However, ‘peace tanks’ could hold another 400 tons.
37. Lillicrap’s 7 April 1925 notes show a Legend adding up to 6,795 tons after discussing these designs with DNC. However, the next day he received the bad news from E-in-C that a more realistic machinery weight would be 1,550 tons. He compensated partly by reducing protection from 1,090 tons to 990 tons at DNC’s behest. He estimated that the ship would be 530ft pp (560ft oa) x 54ft x 16ft (depth 31ft). Fuel stowage below the platform deck would be 1,635 tons, compared to 1,500 tons in the 6,795-tonner, for an endurance of 6,000nm rather than 5,000nm at 12kts.
38. E-in-C offered 72,000shp on 1,450 tons (E-in-C weight alone). This scaled up to 1,612 tons for 80,000shp, which was short of the Kent figure (as designed; Kent had not yet been completed). E-in-C agreed to 1,610 tons exclusive of auxiliaries such as generators.
39. Estimated dimensions at this point were 565ft x 54½ft x 16½ft.
40. PD 01813, cited in the 1926 analysis of cruiser suitability, ADM 1/9272.
41. DNC found it impossible to provide space amidships for a second catapult. Apparently the British rejected the US practice of mounting two catapults side by side in the waist, preferring the centreline position.
42. Staff comments dated 11 November 1925 on a sketch design offered by DNC on 20 October 1925.
43. Protection shown was 4in sides and 2½in crowns for magazines with 2½in and 1½in ends, plus 3in sides, 1½in deck, and 2½in and 1in bulkheads for the machinery. The 4in side and 2½in crown protection extended over the shell rooms, which in previous cruisers had been more lightly protected (1in sides and crowns). The extra shell room protection had been requested because there was a fear that the ship might be blown up by an enemy shell bursting in a shell room and detonating the shells there. This had not previously been considered a major problem, and it may have been connected with the decision to use relatively high-capacity 8in shells instead of lower-capacity 8in armour-piercing (AP) shells.
44. Weight was saved by simplifying the structure at the bottom of the hoist; the gunhouse was almost identical to that in Kent. Some additional protection was provided to the hoist.
45. 540ft pp, 575ft oa x 57ft x 17ft, compared to 535ft pp, 565ft oa x 57ft x 17ft in the December 1925 Legend. The change was ordered early in March: 5ft more pp, and 5ft more overhang aft. The extra length was intended to maintain speed without any increase in power.
46. Given the extra height, the bridge structure was redesigned, an extra deck being worked in. The sea cabins were placed on the first deck above the shelter deck rather than on the shelter deck as in the past.
47. The load on the turret would be considerable, the aircraft’s wings would block the view from the bridge, and the aircraft and catapult might be affected by blast. One way to solve the muzzle blast problem was to mount the catapult to fire backwards along the turret top – assuming there was sufficient space between turret and bridge (needed in any case to limit blast damage to the bridge). In February 1927 DNC asked that the roof of one turret of the new 10,000-ton cruiser Berwick be loaded to represent a catapult and aircraft, as a test of whether those weights would affect turret training (the tests were actually carried out aboard HMS Suffolk in March 1928). The existing 40hp training engine sufficed when the ship was on an even keel, but two such engines would be needed if she were rolling heavily.
48. The Cover gives no indication of when this decision was reached. In March 1929 the builders (Palmers) requested particulars of aircraft and catapult gear, and a paper was prepared announcing that they would not be needed. By this time installation of a rangefinder in ‘B’ turret was being considered, because there would be no catapult to interfere with it. The ship received her main catapult (an E.II.H*s cordite unit) after delivery.
49. Chatfield was Third Sea Lord (Controller); Field was DCNS; Dreyer was ACNS. As DGD, Dreyer had probably been most instrumental in urging the 8in cruiser.
50. Roskill, Naval Policy Between the Wars, I, p 499.
51. York Cover and also Foreign Cruisers Cover (a better version). Ships reportedly carried sixty rounds per 8in gun in standard condition (110 when fully loaded); they had eight 4.7in dual-purpose secondary guns. Machinery and magazines were protected by a 3in belt extending from 2ft 6in to 3ft 6in below the waterline to a deck above the machinery; the magazines had 3in crowns, but machinery had no deck protection other than structural deck plating. The armour was said to be superior to Krupp. Radius of action was given as 10,000nm at 14kts. The Japanese had saved weight by using high-elasticity Manganese steel. DNI (Rear Admiral W W Fisher) signed the basic paper, dated 11 October 1926. ACNS (Admiral Dreyer) made the various suggestions in a 13 October 1926 comment on the DNI paper. He thought that the ‘E’ class had been designed to carry sixteen torpedo tubes but had been reduced to twelve, which was the opposite of what had happened. The British credited the Japanese with four 10,000-ton cruisers under construction under the 1924 and 1925 programmes (Myoko class), and another four in the new construction programme which was being disputed by the Japanese treasury. They became the Atago class.
52. ADM 1/9272, ‘Gun Armament of Future Cruisers,’ a paper written by ACNS dated 31 August 1926.
53. On the basis of 100 rounds per gun, Controller estimated that a two-gun 8in turret cost 218 tons, a twin 7.5in, 173 tons; and a twin 7in, 139 tons. Thus the most extreme decrease, to 7in calibre, would save 79 tons per turret, or a total of 237 tons in a B Cruiser or 316 tons in an A Cruiser. Ammunition weight was based on an estimate of 211lbs for the 7.5in shell and 172lbs for the 7in, not on the 200lbs of the standard 7.5in shell. DNC pointed out that there would be additional savings due to smaller magazines and shell rooms, small savings in hull weight (smaller turret supports), and perhaps even savings in complement; these would probably amount to 200 tons for an A Cruiser with 7.5in guns and 360 tons for an A Cruiser with 7in guns, or 150 tons and 270 tons respectively for a B Cruiser. To calculate armour penetration, DNO assumed that all three guns had the same muzzle velocity, 2,775ft/sec. At 90° inclination (the worst case) 5in vertical armour would keep out a 256lb 8in shell at about 17,500yds; 4in armour would be effective at slightly over 21,000yds. At a much sharper 50° inclination (which might apply to a chase), 5in armour would be effective at about 11,000yds, 4in at about 15,000yds, and 3in at about 18,000yds. Deck armour defined the outer edge of an immune zone: beyond 25,000yds for 3in, about 23,000 for 2½in, and about 21,300yds for 2in. Against 7.5in shell at 90°, 5in was immune beyond about 11,000yds, 4in beyond about 13,200yds, and 3in beyond about 19,000yds. At 50° inclination both 4in and 5in were effective inside 10,000yds, which was the inner edge of the expected fighting zone; 3in was immune at 13,000yds. However, shells fell more steeply at longer ranges, so deck armour was less effective: the outer edge of the immune zone was about 22,200yds for a 3in deck, about 21,500yds for a 2½in deck, and somewhat over 18,000yds for a 2in deck. As might be imagined, the 7in gun was even worse against vertical armour: 5in was immune inside 10,000yds, 4in at about 10,200yds, and 3in at about 14,500yds. Against deck armour, the shell would penetrate a 3in deck at about 22,500yds, a 2½in deck at about 21,000yds, and a 2in deck at about 18,000yds. Discussing the different critiques of cruiser design, DGD mentioned that ‘war in the Far East [has] been laid down as the basis on which our preparations are to be made’.
54. Based on their analysis, the British thought the Japanese Myokos had 1-2in side armour, and a 120,000shp powerplant (for 35kts). In fact they had 3.9in side armour and developed 130,000shp, these figures being possible because they displaced 10,980 tons rather than 10,000 tons in standard condition.
55. The previous DNO had said that it was more difficult to design a triple 8in turret than the triple 16in, due to restricted space. The output per gun might be reduced by a quarter due to structural arrangements, so in 1924 DNO estimated that the output of a triple turret compared to a twin would be 9:8 rather than 12:8 (DNC disagreed). These estimates were probably made when DNC considered the cruiser with four triple turrets described above.
56. Calculation from Lillicrap Constructor’s Notebook 4. At 16kts a ship burned 1.6lbs/shp/hr; for 20kts the ship would need another quarter-ton per hour for every additional 10,000shp boiler. At 28kts, the ship burned about 1.1lb/shp, and adding steam for full power added ¼ ton per hour for each additional boiler. Consumption at 18kts was 1.5lb/shp/hr, and steam for full power added ¼ ton for every additional boiler. For a Kent, the run-out would require 1,195 tons, the sortie another 170 tons, and the battle another 420 tons, a total of 1,785 tons. Endurance was calculated on the basis of 90 per cent oil usage, hence the 2,000 tons. Dreyer thought his new definition would require about 2,250 tons of oil, 350 tons less than that demanded by the Staff Requirement and 950 tons less than London class capacity.
57. Details from Lillicrap Constructor’s Notebook 4.
58. Notes dated 9 October show 575ft (pp) 610ft (wl) x 60½ft x 18ft (hull depth 34ft) for Design X, compared to 540/575ft x 57ft x 17ft (32ft) for York and 595/630ft x 66ft x 17ft for London. Shortening the hull saved 610 tons compared to London (5,460 tons) but cost 510 tons compared to York (4,340 tons).
59. Yet another version was larger (615ft x 62ft x 18ft), with London-class armament except for six instead of four 4in. This version had increased magazine and shell room protection (5in sides and 4in crowns).
60. Proposed dimensions were 570ft x 54½ft x 16ft (changed to 16ft 6in), hull depth 31ft. Lillicrap estimated 4,000 tons for the hull (confirmed by scaling up the B design hull), 1,640 tons for E-in-C machinery, 130 tons for auxiliary machinery. 900 tons was available for protection. Data dated 4 December 1926. On 1 October, probably as a starting point for designs with more power, E-in-C estimated that a 100,000shp plant would require 52ft boiler and engine rooms (44ft and 48ft respectively for Kent); 90,000shp would require a 48ft boiler room and a 50ft engine room. Lillicrap chose 90,000shp because he had to add only 10ft to the ship’s length.
61. 530ft x 54ft x 16ft; 7,500 tons. In contrast to the 7,800-tonner, armour over the shell rooms was reduced to 1in (side and crown). DGD wanted a thicker (1½in) deck over the machinery. The 125 tons required would be gained partly by reducing side armour over the machinery by an inch (70 tons). Eliminating the box protection saved 320 tons, but the more complete platform deck cost another 125 tons, for a net saving of 195 tons. Lillicrap had to add 1in bulkheads to the magazines, because shells could plunge through the unprotected ends of the ship.
62. As in Surrey as then conceived, half the machinery (the central unit comprising after boiler room and forward engine room) would have heavier armour, in this case 5in on 1in C sides covered by 2¼in on ½in NCD deck. The rest of the machinery had the usual 1in side and 1½in deck. The thick side offered immunity against 8in fire beyond 7,000yds at up to 50° target angle (58° in the A Cruiser) and immunity against deck protection inside 20,000yds. In December 1927 DNC suggested thickening the upper parts of the magazines at the expense of the lower parts protected mainly by being further under water. In this way the ship could have a quarter-inch more on its roof (total 3¾in) and 5in upper sides (2in lower sides) instead of the earlier uniform 4in, without any increase in weight. That offered another 1,000yds of protection to the magazine crown. The 5in side gained 20° in inclination for protection at all ranges against 8in fire; it reduced the minimum range by 3,500yds. First Sea Lord approved ACNS’ suggestion that the magazine crowns be shaved to 3½in, the weight saved being transferred to the 2in parts of the magazine sides.
63. Habitability depended largely on volume available for personnel, but it was very difficult to calculate volume and to relate it to overall ship dimensions. The usual surrogate for crew space (and for oil stowage) was hull length, so a design sacrificed habitability by having its length cut. This was quite sensible; the short York was described as badly congested.
64. DNC provided 3in NCD armour worked on a ⅜in structural deck on the magazine crowns, but he had to take about ¼in off the side armour, both abreast the midships machinery and on the magazine sides. That left 5¾in side armour (NCD) on 1in C armour on the sides. The machinery deck armour was 2¼in on ½in NCD. The desired 2in machinery space bulkheads extended 14ft below the deck (5ft below the lower edge of the side armour).
65. The final version of the design had 1,925 tons of protection. The design provided 2,200 tons of oil rather than the 3,200 tons of the earlier cruisers, but that provided the endurance originally desired (8,000nm at 12kts).
66. The bridge structure had to accommodate considerable equipment for fire control and for night control. Along each side, fore to aft, were a captain’s sight (for target designation), a gyro repeater, a searchlight sight, and a star shell sight (with a star shell calculator nearby). In addition to the DCT for own-ship main battery control, the ship was designed to work with others in concentration fire (as in the battle of the River Plate in December 1939). To that end she had to work with a master ship, receiving her firing instructions and measuring the offset to the master ship so that she could adjust her own fire-control solution accordingly. She therefore had two 9ft PIL (Position in Line) rangefinders. Wholly separate from the PIL system was a pair of 12ft torpedo-control rangefinders used both for that purpose and to support tactical plotting. In earlier cruisers these rangefinders interfered with the view from the standard (magnetic) compass itself, making it difficult for officers to take bearings. In Exeter these rangefinders, mounted on large boxes to limit vibration, hindered access to the wing gyro compass repeaters on the bridge. Along the angled forepart on each side of the bridge was a chart table, and down the centreline of the bridge was the compass platform, a raised section carrying a standard compass with a pelorus forward of it. The torpedo control position was on the lower bridge, behind a prominent slot in the side of the structure. At its fore end was a 12in signal light. Abaft that light were a trainer’s sight, a B sight, a firing pistol, a clear range indicator, and a torpedo order instrument (behind plating). The upper steering position was at the fore end of this level. Blast shutters protected the instruments. The after part of the bridge, abaft the DCT, supported the stalk of an HADT, finally provided forward, as long desired. Dorsetshire and Exeter were both fitted experimentally with roofed bridges (the roof fitted to Exeter was a prototype for one planned for the Leanders). In Exeter the roof covered the compass platform and the fore part of the bridge, its sides angled forward. The area between roof and bridge front was covered by glass windows forming a windshield, which extended beyond the end of the roof to the end of the angled fore part of the bridge. In June 1932 the CO of Exeter wrote to the Admiralty that experience in both ships convinced him that the weather protection involved outweighed any disadvantage due to a more limited view (this would change drastically due to air attacks during the Second World War). Overall, the captain much liked the bridge, which he considered well designed and convenient. He complained only about the wind baffles, which seemed useless, even though (he was told) they had been designed after wind-tunnel experiments.
67. ADM 1/9301, dated 2 July 1929. This was DNC’s submission of the new design.
68. The measures Lillicrap proposed suggest how much could be done. He eliminated the 100 tons of protection added for the 1929 ship. Cutting belt armour from 5in on 1in to 4in on ½in (a total reduction of 1½in) would save another 130 tons; cutting to 3in on ½in would save 215 tons. Cutting the deck from 90lbs on ½in (110lbs) to 80lbs on 1in (100lbs), would save 55 tons. Cutting magazines crowns etc from 3in to 2½in (good to 21,000yds) would save another 25 tons.
69. Lillicrap Constructor’s Notebook 6. Requirement TD 340/35 was dated 2 December 1935.
70. Notes in Lillicrap Constructor’s Notebook 6 refer to TD 126/32, presumably the Staff Requirement, and to D.0618/33, containing instructions from CNS. The originals do not appear in the Kent class Cover.
71. Data from inclining experiments are given in various Constructors’ Notebooks. DNC tried to estimate weight growth due to age. On initial inclining, Berwick was the lightest (13,428 tons deep). In July 1931 Cornwall displaced 13,624 tons deep, having gained 115 tons in three years. Kent was the heaviest (13,520 tons deep). In March 1931 her deep displacement was 13,744 tons, having gained 184 tons. Since her original standard displacement was 9,850 tons, it was now 10,034 tons. In March 1934 Kent displaced 13,742 tons deep.
72. ADM 229/19.
6. The 1930 London Treaty and Its Cruisers
1. ADM 1/8765/313, the Minutes of the Naval Planning Committee for 1928-9 and 1931-2.
2. Each salvo had to contain at least one deliberate miss so that it could be spotted, and cover a large enough area that the gunner would not unnecessarily change ranges. Firing at enemy destroyers required a good volume of fire, the minimum being four 8in or six 6in. The paper on small 8in cruisers included a summary of the expected results of engagements between various kinds of cruisers, assuming each was protected against fire from its own calibre of gun, and on gunnery practices conducted in 1922-7, the results being reduced by 30 per cent to simulate action conditions. On this basis, the small 8in cruiser would put a 6in cruiser out of action in fifty-eight minutes at 15,000yds, suffering 25 per cent loss of efficiency (but not counting the effect of hits in the vitals). At 9,000yds the 8in cruiser would put the 6in ship out of action in eighteen minutes, losing 38 per cent efficiency. By way of comparison, at 15,000yds a cruiser armed with eight 8in guns would put two 6in ships out of action in forty minutes, but at the cost of two-thirds of her efficiency. A third 6in cruiser would turn the scales in favour of the 6in ships. A cruiser with eight 8in would knock out the four-gun cruiser in twenty minutes at the cost of 30 per cent of her efficiency, but would be knocked out by two such ships in thirty-two minutes (one would lose 90 per cent of her efficiency, the other 20 per cent). In reality, most 8in cruisers were hardly immune to 8in fire, so these figures were deceptive. They also probably envisaged an unrealistically high rate of 8in fire.
3. ADM 1/8765/313 contains Chatfield’s memorandum is dated 1.6.28 (i.e. 1 June), but given the dates of other papers in the file it seems more likely to have been written in January 1928. Chatfield proposed a three-type fleet: full 8in cruisers, convoy cruisers and inexpensive 6in cruisers for fleet work, with enough full cruisers to face down the enemy’s 8in cruisers. On this basis he envisaged twenty A Cruisers, twenty-five convoy cruisers and twenty-five fleet cruisers.
4. Appendix III to the Staff Requirements (copy in ADM 1/8765/313) laid out the reasoning for the gun. A single hit must be sufficient to stop a destroyer; a large calibre was desirable to attack aircraft carriers and minelayers. Against destroyers the ship needed volume of fire, which meant both numbers of guns and rate of fire per gun. Mountings should have good ‘A’ arcs. Mountings could be open (with one, two or three guns) or could be turrets (two or three guns). Open mounts with more than one gun had to be power-operated, and were inferior to turrets with similar numbers. Twin 6in turrets existed, and had recently been developed to achieve six to seven rounds per gun per minute. The Germans had adopted triple 6in mounts in their latest light cruisers, but the Royal Navy could not adopt such mountings without trials, so they were not considered. The conference decided that the ship should have 6in directors fore and aft, and that there should be consideration of allowing for divided fire.
5. Policy was that not fewer than two HA guns should bear on any part of the sky, hence the requirement for four such guns, mounted two to a side but capable of firing across the deck. A policy had also been adopted that 6in LA guns in battleships, and the 8in guns of cruisers, should be usable against aircraft. However, at the conference to determine Staff Requirements, Controller rejected a proposal for simple arrangements to allow the main armament (as in the A Cruisers) to fire at aircraft: there was no such thing as a simple HA control system. The ships would have to have the complete HACS 1 system, but it probably would not be worthwhile for a ship carrying only a few HA shells (and changing over from one type of shell to the other would not be quick). This was hardly worthwhile for a ship carrying 4in HA guns. The conference decided to require 50° elevation (more if possible) of the 6in mountings, but no space would be sacrificed to accommodate a third fire-control table (computer) in addition to 6in low angle and 4in high angle. Should some simple 6in HA fire-control arrangement be developed, it would be worth considering, but nothing should be done for the present. The earlier plan to mount four single 4in guns and two quadruple 0.5in machine guns was accepted; there was no interest in the 2pdr multiple pompoms of the larger cruisers.
6. DTM favoured quadruple tubes (work on a quintuple mounting was not promising) with two torpedoes per tube. The conference on the ship’s Staff Requirements agreed on two quadruple tubes. DTM favoured the new Type J destroyer torpedo, which was relatively simple, with a single setting (11,000yds at 35kts) and which did not require enriched air, hence did not require a massive plant unsuited to lively ships. The conference agreed.
7. The basic requirements were to protect magazines so that the ship could not be blown up by similar or smaller ships, and to protect the machinery spaces as well as possible to preserve mobility. At the least, magazines should be protected between 10,000-16,000yds against 5in fire and engine- and boiler-room decks should be immune to 5in fire beyond 15,000yds. Magazines should be immune to 4.7in fire beyond 7,000yds, and engine and boiler rooms beyond 14,000yds. Since the engine- and boiler-room sides were only about a third of the target presented by the deck over them, the sides could be reduced as necessary to provide sufficient deck protection. In these figures the inner edge of the immune zone was set by side armour and the outer by deck armour. The logic of the ranges used was that although 6in splashes could be seen at 18,000yds, hitting was unlikely outside 16,000yds. Because the machinery space was so large, and because hits there could not blow up the ship, it was acceptable for armour to be thinner (this logic had applied to the large cruisers). Hence immunity inside 15,000yds (considered the ‘outer limit of the important fighting range against other cruisers’) was specified. For magazines, immunity at 10,000yds was considered equivalent to immunity at 8,000yds at 60° inclination, and to shorter ranges at greater inclinations. Magazine sides were probably completely protected by water below 7,000yds and between that range and 10,000yds the tops of the magazine boxes presented small targets to 6in fire. It was considered impossible to keep 4.7in shell out of machinery below 14,000yds when fighting beam to beam (0° inclination), but at such ranges the cruiser would probably be trying to close any destroyers attacking her in a day action, and hence would be at a greater inclination, perhaps 60°.
8. The 7,000nm at 16kts equated to a modified version of the endurance formula previously proposed: 200 hours at 16kts (3,200nm) with steam for 20kts; eight hours at 16kts with steam for full speed; ten hours at 24kts with steam for full speed; and twelve hours at 18kts with steam for full speed, all on the basis of being out of dock for four months, with a 25 per cent margin for contingencies. Although calculated on the basis of fleet operation, this was considered sufficient for trade route work (otherwise not analysed). This is from the Staff Requirement for the 6,000-ton light cruiser included in ADM 1/8765/313.
9. Initially the depth charges were to be held at bases and issued only in war, but in 1934 it was decided to keep them on board ships which would have been sent East in the event of war, on the theory that sending them back to base for war equipment would entail too great a delay. The same reasoning applied to splinter mattresses to be used in wartime.
10. The Cover gives no indication of when the change was proposed by E-in-C.
11. Weight increases in Leander were due to increased shaft horsepower (machinery weight 1,395 tons vs. 1,345 tons), to increased side armour over the machinery (but reduced deck armour: total 845 tons vs. 780 tons), and to fitting of a heavy catapult and flagship accommodation (hull weight 3,551 tons vs. 3,570 tons, equipment weight 498 tons vs 485 tons, armament [including the seaplane], 55 tons vs 40 tons). The 1931 ships already had heavier machinery (1,420 tons) and grew due to added equipment and hull weights (845 tons and 3,586 tons, respectively) and to the weight of the heavier seaplanes.
12. In May 1933 DNC tried to check its weight by comparing that figure with the weight of the twin 6in secondary mounts in the battleship Nelson. Initially it was expected that the guns themselves would weigh about as much as in the battleship (total 72.1 tons), the mountings adding up to another 330 tons (compared to 355.1 tons in the battleship). In fact the guns weighed only 55.4 tons, but the mountings weighed 379.5 tons. Had this sort of underestimate applied throughout the ship, she would have come out 1,400 tons heavy.
13. Practice varied between reporting design and actual displacements. Except for the Arethusas, the difference was small, and far smaller than that practised by other navies. After the Second World War, British and US constructors reviewed Japanese practice and stated that overage was accidental rather than deliberate. A draft version of the review notes is in the foreign ship series of US Preliminary Design papers at NARA II. By way of contrast, a German document giving stated, actual design and completion weights was introduced at the Nuremberg war crimes trials as evidence of deliberate cheating, hence an intent to conduct aggressive war. According to its official cruiser history, the Italian navy deliberately designed its Zara class for 12,000 tons rather than 10,000 tons, to secure adequate protection with eight 8in guns. That seems to have been the sole example of Italian cheating. G Giorgerini and A Nani, Gli Incrociatori Italiani 1861-1964 (Ufficio Storico Della Marina Militare, Rome: 1964), pp 496-8.
14. For example, on 4 October Lillicrap discussed with DNC a 470ft (wl) (approximately 450ft between perpendiculars) x 45ft x 13ft 6in ship, for which he was about to do a rough layout. He expected to use four of the new 17,000shp Thornycroft boilers (rather than 13,000shp destroyer leader boilers); the ship would make 35kts on 60,000shp. The ship was about the size of a ‘C’ class cruiser (425ft between perpendiculars, with a length to depth ratio of 17, which probably could not be exceeded, so hull depth would be about 27ft, i.e. freeboard would be 13ft 6in). Adopting destroyer engine-room practice could dramatically shrink machinery length: the leader Codrington needed only 52ft for 39,000shp, whereas Surrey needed 90ft for 60,000shp. Lillicrap thought he could manage on 85ft (less than proportionate to Codrington), with two 44ft boiler rooms.
15. This seems to have been worked out in considerable detail. Dimensions were 470ft (wl) x 45ft x 12½ft (forward) l4½ft (aft); freeboard was 26ft forward, 13½ft amidships, and 13ft aft. Since the ship had open mounts rather than turrets, there was no difficulty in using triple shafts. The ship had six boilers (presumably the 13,000shp destroyer leader type) for 72,000shp at high revs (350rpm) and a speed of 36kts. Oil fuel stowage was 850 tons (5,000nm at 12kts). In addition to the five 6in guns she had two 4in HA guns and one quadruple torpedo tube (no catapult or aircraft). Protection was limited to boxes over magazines (2in crown and side), shell room (1in crown and side), fire-control room (sic, not transmitting station, perhaps because it contained multiple computers: 1in) and steering gear (1in). Protection amounted to 120 tons. E-in-C machinery amounted to 1,168 tons, auxiliaries adding another 55; the hull was expected to weigh 2,250 tons, general equipment 330 tons and armament 260 tons (Lillicrap rounded the total, 4,180 tons, up to 4,200 tons).
16. As reported on 22 January 1930, this ship was 525ft (wl) x 52ft x 14ft 6in (forward) 16ft 6in (aft) with freeboard 28ft forward, 14ft 6in amidships, and 16ft 6in aft. She had the same powerplant as the 5,600-tonner, but the greater length equated to more oil fuel: 1,200 tons, sufficient for 5,500nm at 12kts. Except for the additional turret, she had the same armament as the 5,500-tonner, and the same level of protection (430 tons). Armament (without the aircraft and its catapult etc) weighed 645 tons, compared to 500 tons in the 5,600-tonner and 260 tons in the 4,200-tonner. General equipment increased to cover the larger crew (430 tons), and hull weight increased to 3,150 tons.
17. Dimensions reported to DNC were 430ft (wl) x 42ft x 11ft (fwd) 13ft(aft), with freeboards of 25ft forward, 14ft amidships, and 15ft aft. As in the 4,000-tonner, the ship had a six-boiler 72,000shp pow-erplant with three shafts, running at slightly higher revs (360rpm); expected speed was 38kts. Oil capacity was 750 tons (another estimate was 800 tons). Armament weighed 240 tons, not far from that in the 4,000-tonner. The hull was simply scaled from the 470ft of the 4,000-tonner, and required power was initially calculated as 60,000shp for 36kts. Lillicrap then realised that the same 72,000shp plant he was using for other designs could just provide a speed of 38kts. E-in-C weight would be 1,165 tons. Without power turrets, auxiliary weight was only 55 tons. Endurance was given as 5,500nm at 12kts. DNC asked for a more detailed investigation on 4 April 1930, the ship to armed with four or five single 6in guns, with endurance of at least 5,000nm. The 3,000-ton displacement was not to be considered a rigid limit. Lillicrap thought that he might be able to reduce beam to 41ft. As before, he offered five 6in (200 rounds per gun), two 3in HA (200 rounds per gun), and two quadruple torpedo tubes, but this time he added a pair of quadruple 0.5in machine guns, Lewis guns, and the depth charges standard on British cruisers. He retained the 72,000shp plant, and tried alternative lines based on the flotilla leader Codrington and on the pre-1914 light cruiser Forward, of which Codrington lines were likely to be better (based on comparisons with other hull forms, not new tank tests).
18. An upper-deck 6in mounting weighed 20 tons, but a twin weighed 90 tons, adding 50 tons directly (for the 5.5in Lillicrap estimated 15 tons and 70 tons). However, that was not all. The short-trunk twin mounts also needed twelve men for each ammunition lobby and eighteen for the turret itself, a total of thirty per pair of guns. In Emerald, each single gun was served by nine men (the twin in Enterprise needed fifteen). The single had two men for ammunition supply, so two singles would need a total of twenty-two men. Mounting weight could be cut somewhat by halving the thickness of the shield (total weight would fall to 80 tons per mount). Magazines and shell rooms had to be protected. General equipment weight increased, too. The ship grew to 4,500 tons (480ft x 46ft x 13ft) and speed would be about 35kts on 72,000shp. The longer hull added enough oil tankage to make up for the losses due to the new magazines. This study was reported early in February.
19. As Chancellor of the Exchequer, Winston Churchill urgently wanted to cut defence spending in hopes of helping the British economy recover. In November 1927 he suggested that none of the six 1927/8 and 1928/9 cruisers be laid down. The Board agreed; Roskill, Naval Policy Between the Wars, I, p 555 suggests that was partly due to uncertainty about what sort of cruisers were wanted. At this time current expenditure was running above the Estimates, so any new construction would have required a supplemental appropriation – which, presumably, would not easily have been forthcoming. In the end two of the 1927/8 cruisers were cancelled, leaving HMS Exeter. The British financial situation continued to deteriorate. Churchill demanded further cuts, arguing that they were justified by the favourable political situation. Churchill cited Foreign Office claims that Japan was unlikely to cause any problems over the near future to attack the central justification for British naval planning. Even so, for 1929/30 the Board secured Cabinet agreement for three 10,000-ton cruisers. Labour won the May 1929 British election, with Ramsay MacDonald, who had previously championed arms control, gaining office as a minority Prime Minister. His presence in itself guaranteed that the Admiralty would be pressed to accept any apparently reasonable agreement. For the navy, probably the worst outcome of the London conference was that the Government accepted a drastic reduction in destroyer numbers, which the Admiralty had predicated on abolition of submarines: the destroyers were cut even though the submarines were not. Cruiser cuts were of far less moment, although it did not seem so at the time.
20. There was little hope of building up to seventy cruisers, let alone seventy underage ones, by 1936. On 18 January 1930 the Admiralty sent a memo to the British Empire delegation: it did not retreat from the seventy-cruiser requirement as a minimum, ‘not taking the United States into account’. Not only could fifty be accepted only for a strictly limited period, but it was acceptable only provided other sea powers limited their own programmes, and ‘provided that in our number there is a proper proportion of new construction suitable for extended operations, that is, they must be comparable to the types being built by other Powers, they must be capable of defeating any armed merchantman or raider, they must have sufficient radius of action to carry out their tasks, and must be habitable in all climates’. Statement repeated in connection with the future cruiser programme in ADM 167/86.
21. Stephen Roskill, Naval Policy Between the Wars Vol II (Collins, London: 1967), pp 40–1.
22. On the 339,000 tons allowed to the Royal Navy, 192,200 tons of 6in cruisers could be divided among the three types. To meet the fifty-cruiser requirement the Royal Navy had to build thirty-five ships armed with 6in guns, which could ultimately be divided among the three types: Leander (7,000 tons), intermediate (4,500 tons), and fleet scout (3,000 tons). Without the intermediate ships the tonnage allowed for twenty-one 7,000-tonners and fifteen 3,000-tonners, but ‘the fifteen is too many of such a weak type, and it is more probable that we would want only 6 or 8 at most’. Hence the Royal Navy needed the intermediate type, though actual numbers could not be set. ‘Years ahead may bring many changes but I think the above confirms our view that an intermediate class of 4,500 tons maximum is a probable type. Work done on the 3,000 ton Scout class is however all to the good, through now less pressing than the next size larger.’ Controller memo, 9 August 1930, copied in Lillicrap Constructor’s Notebook 5.
23. As noted in an Appendix to the February 1932 Board Memorandum on the cruiser programme in the light of large foreign cruisers, in ADM 167/86. From 27 May 1930 C-in-C Mediterranean was Admiral Chatfield, who had left office as Controller in 1928.
24. However, in connection with the question of building more large cruisers to deal with the new large foreign cruisers, a Board memorandum of February 1932 on the building programme described these ships as useful only with the fleet, adding that ‘the number of small cruisers that could usefully be included in the Fleet in time of war has been the subject of careful study’. This memo (in ADM 167/86) posited twelve such cruisers ‘but it would be unwise to project beyond 1933, when we shall know more about cruiser types which are being built by other countries.’
25. Lillicrap Constructor’s Notebook 5 shows a June 1930 study, presumably to help the Naval Attaché in Romania estimate what could be done to meet a Romanian request for a cruiser with eight 6in guns. Naval Attaché (Head of the British Naval Commission, Rear Admiral Reginald G H Henderson – later Controller) wrote that the Romanians wanted two or three cruisers; he wondered whether they might be happy with an improved ‘E’ displacing about 800 tons more, with more deck armour and all-oil fuel. Lillicrap sketched an 8,000-ton 32kt cruiser but also a 5,700-tonner – showing how much could be done on a very limited displacement. The sketches resembled a Leander, and they give some idea of what could be provided in an eight-gun ship. The 8,000-tonner was 570ft (wl) x 55ft x 16.5ft, mounting four twin 6in, four 4in HA and two triple torpedo tubes, with an 80,000shp powerplant giving 32kts. Magazine boxes and machinery all had 4in sides and 2in decks, somewhat more than contemporary British light cruiser protection. Estimated cost was £1.9 million. On 5,700 tons (520ft x 52ft x 14ft, 65,000shp for 32kts), Lillicrap offered the same armament but no armour at all over the machinery. Magazines could be protected by 3½in sides and 2in decks, and shell rooms by 1in sides and decks. The ship would cost £1.5 million.
26. Lillicrap Constructor’s Notebook 5 describes the development of this design in detail. Controller’s request for the design was dated 7 August 1930 (copy in Lillicrap Notebook). Lillicrap’s notes written the next day include the cut in 4in guns.
27. Power was 60,000shp. A three-shaft arrangement was rejected because it was difficult to work in the after magazine. E-in-C adamantly opposed destroyer machinery both because he did not want to lose reliability and because he wanted the ship to develop full speed in the tropics.
28. He developed a series of alternative designs 470ft or 490ft long: A, B, X, Y, Z, P and Q, all with the desired three twin 6in guns (with three 4in HA guns in Designs A and Q, two in all the others). All had the same protection, 3in side and 2in crown over magazines, 1in shell room, 3in (total) side and 1in deck over machinery, 1in over transmitting station side and deck, and 1½in over steering gear. Except for Designs Z and Q (490ft long) all were 470ft (wl) x 47ft 6in x 14ft. The other variable was number of boilers, hence machinery box length: four boilers in Design A, six in the others. Design B showed that the extra protection weight required for a longer machinery box (620 tons rather than 560 tons) more than made up for lighter armament weight; on 60,000shp the ship would make 31.8kts rather than 32kts. Design P used 86,000shp cruiser machinery, the heaviest in the series (1,540 tons [E-in-C weight, since auxiliaries were the same in all designs]) to drive a 470ft ship at 34kts, but the extra machinery box length cost so much oil that the ship would make only 4,000nm at 15kts. Design Q was therefore lengthened to 490ft. The longer hull was easier to drive (82,000shp, 1,465 tons) to achieve the desired 34kts, and the shorter machinery box and longer hull accommodated enough oil for 5,000nm at 15kts. Design X used 68,000shp destroyer machinery to make 33kts (machinery weight was the same as in Designs A and B, 1,070 tons). Design Y used 72,000shp cruiser machinery (1,285 tons) and sacrificed endurance (4,500nm at 15kts) but made 33kts. Design Z showed that 64,000shp cruiser machinery (1,140 tons) could drive a longer hull at 33kts, and the extra length accommodated enough oil for 5,300nm at 15kts, the greatest endurance in the series. These ships were about the size of a ‘D’ class cruiser (465ft 6in (wl) x 46ft x 14ft 6in (mean)ft, 4,850 tons, 40,000shp, 5,950nm at 10kts). The 82,000shp Design P would probably displace 5,500 tons. The table of these designs went to DNC on 22 October 1930, for use by the Sea Lords at a meeting on 23 October.
29. Dimensions at this stage were 480ft (wl) x 47.5ft x 14ft, and an estimate based on other cruiser designs showed that the ship should make 32.5kts on 64,000shp.
30. Analysis of eleven possible hits on the machinery showed that the new arrangement was superior; it always retained at least a third of its total power. The analysis did not take into account the additional protection to the after boiler room given by the longitudinal bulkheads.
31. The larger boilers would operate at 350psi and at a superheat of 630° F. This degree of forcing saved 85 tons. At this stage the boilers and turbines were rated at 15,500shp each. It would be somewhat uneconomical to keep one large boiler lit for harbour services, so DNC suggested providing a diesel generator. Slightly later DNC and E-in-C favoured a 16,000shp boiler, which was adopted.
32. As pointed out by D K Brown, Nelson to Vanguard: Warship Design and Development 1923-1945 (Chatham Publishing, London: 2000), p 73. The designers rightly calculated that flooding the wing space would cause only a small list. However, flooding one or two of the big adjacent compartments would dramatically reduce the ship’s stability, so that the small amount of asymmetric flooding would have a much greater effect. A ship so damaged would capsize quickly – as several cruisers did. In a footnote to his edition of the official DNC history of wartime construction, Brown pointed out that this subtlety would not have been obvious before the era of computer calculation, presumably because it was a dynamic rather than a static effect.
33. The design offered slightly better protection: the deck over the transmitting station and the low power supply was 1in NC over ⅜in D.1 plating, equivalent to the 1½in D.1 approved for repeat Leanders. Careful review of weights showed a saving of 50 tons, for a displacement of 5,450 tons. The ship was 480ft long on the waterline.
34. Lillicrap Constructor’s Notebook 5.
35. Probably the ship would need no more than 11,000–12,000shp. That would scale to 210 tons, but instead 400 tons (four boilers, two shafts) was used; machinery spaces would probably be only 70ft long (two 18ft boiler rooms, and 34ft for engines). Lillicrap used the hull weight of the 5,000-tonner and Leander armament to get a total unprotected displacement of 3,540 tons. He took Surrey protection as sufficient to resist 8in fire: the machinery required 5½in side armour (over a ½in skin), 5½in (on ½in) bulkheads, and a 2¾in (over ½in structure) deck. The belt would be about 10ft deep, so machinery protection would come to about 410 tons. Magazines and shell rooms would be similarly protected (5¾in sides, 3in decks), a total of 400 tons if the lower part of the magazines had 2½in armour, as in other cruisers, plus another 60 tons if the whole side were 5¾in thick. Because the machinery space was so short, the belt would not really protect the ship’s stability and buoyancy. Lengthening it to 200ft would add about 800 tons of armour, displacement rising to about 6,000 tons. Protection included 1in over the ammunition hoists (a ring atop the magazine, and a ring above the deck) and over an ammunition lobby. The ammunition lobby set the minimum hull depth at ‘Y’ turret, where it was immediately above the magazine.
36. Lillicrap Constructor’s Notebook 6, pp 9ff.
37. Lillicrap was apparently not assigned to this study; he worked instead on studies of small capital ships which were probably intended to help form British policy at the (abortive) 1932 Geneva Conference. The figure comes from the second Amphion Cover, folio 9, paper dated 5 July 1933. The original call for design studies, of the nine- and ten-gun ships, was made by Controller in a paper dated 8 February 1932. A later Minute on the same paper (which has since been lost) called for ordering one or two triple turrets for installation in 1932/3 programme ships, giving the nine- or ten-gun armament. Later the weight of the triple was fixed at 125 tons. Additional protection was dropped because of the weight limit; DNC thought the Amphion, at 7,250 tons, was the largest acceptable ship. He gave the displacement of the ten-gun ship as 7,550 tons, of the nine-gun as 7,300 tons, and of the eight-gun (three turrets) as 7,200 tons. The seven-gun ship seems to have been added for this 1933 paper. Speeds of the ten- and nine-gun ships were given as 31.8kts and 32.4kts, using the Amphion powerplant.
38. Lillicrap Constructor’s Notebook 6, p 67 includes a DNC request on 16 November 1932 for an Arethusa-size cruiser with twin turrets fore and aft and the other two guns in single weather-deck mountings. That would save weight on magazine protection and handing rooms. Lillicrap answered that such a ship would be unable to carry an aircraft and catapult, which were essential. He also thought it unlikely that three-gun ahead fire would be acceptable, and that it would be difficult to provide sufficient HA guns. Putting the single mounts on the broadside, to give a total of five in some directions, would not be acceptable. The only acceptable arrangement would be single guns in ‘B’ and ‘X’ positions. Much of the weight saved would go into blast screens to protect the crews of the open mounts, but enough would be left over to provide another ½in on the belt (2¾in on ½in structural plating).
39. Leander had four turbo-generators, two on the platform deck immediately forward of the forward boiler room and two in the gearing compartment. With the modified arrangement, the forward generator compartment had to be omitted and the two generators relocated to the wings of the after boiler room. They would be diesel-driven (one set of turbo-generators was placed in each engine room). This was the same arrangement as in the new Arethusa class. Later, answering Controller’s request that extras be ruthlessly cut, DNC commented that British electric generation capacity exceeded that of US ships, but the example he gave (Omaha, four 100kW turbo-generators) was badly out of date, closer to the First World War ‘E’ class.
40. ADM 268/52.
41. Like the 1921 Naval AA Gunnery Committee, the 1932 Committee took into account the only existing experience of air attack, during the First World War, although it was well aware that technology had changed radically since then. According to the 1932 report, British ships operating off the Belgian coast and in the Heligoland Bight were often attacked, but the only success was a hit on the light cruiser HMS Attentive. The Battlecruiser Force and 1st Cruiser Squadron were attacked, without success, on 1 June 1918. A German destroyer had been sunk by bombing at Zeebrugge, and another destroyer severely damaged while underway. British aircraft harassed the German Goeben and Breslau when they sortied in January 1918, and made many bombing attacks on Goeben when she was stranded for six days afterwards. A Turkish shore AA battery forced the attackers to such altitudes that they scored only two hits in 250 tries. Aircraft were far more successful against submarines and merchant ships.
42. The Committee cited rapid growth of aviation in the US, British and Japanese navies; the development of French long-range flying boats for Mediterranean operations; and the reported training of Italian bombers for long-range flights over the sea, plus the recent development of torpedo bombers by the Italian air force. On the other hand, attack over the sea was a specialised art, not quickly mastered by an air force not interested in the problem. Air Ministry representatives and DNAD estimated the threat posed to a fleet transiting 100 miles from enemy air bases ashore. They assumed that three aircraft could be rearmed and refuelled in fifteen minutes, which they said was the usual capacity of a modern air base. In that case the fleet might be subject to an initial wave attack (eighteen aircraft per wave), waves following each other quickly. It would take the enemy force two to four hours to regroup, after which it might attack with thirty-six, twenty-four or eighteen aircraft every three, two or one hours respectively. This analysis led the Committee to conclude that the problem was two or more waves of about eighteen aircraft each in quick succession.
43. The US aircraft approached at 8,000–10,000ft, diving vertically at the target and releasing bombs at 3,500–4,000ft (the aircraft could not level out from a lower altitude). The press reported that US attempts had been disastrous, and the Air Ministry thought it unlikely that an aircraft could be designed that could sustain the stresses involved and also lift a heavy-enough bomb. It preferred to work on a sight useful to a glide bomber, approaching the target at a 45° or smaller angle. The 1932 report discounted American claims that near-vertical dive bombing made possible 40 per cent hits compared with 4 per cent for horizontal bombing.
44. By this time work on radio-controlled targets was well advanced, so the next step to an anti-ship missile seemed obvious. ‘It is difficult to visualise this form of attack being profitably employed against any target other than large areas.’ According to a footnote, ‘The Air Ministry regard this idea as exceptionally secret and would prefer that it is not generally promulgated.’ The report went on to point out that piloted explosive aircraft, which could hit manoeuvring ships, ‘cannot however be ruled out. It is reported that, sooner than accept defeat, ramming other aircraft is a recognised principle among Japanese pilots.’ (Ch 2, para 22) Did anyone remember that when the Kamikazes appeared?
45. According to the report of the 1932 Committee, but not to the Cover or to any other design material. This would tally with the decision to give their main batteries 60° elevation. Plans called for not fewer than thirty HE anti-aircraft shells per gun. The peacetime outfit per gun was 180 CPBC (i.e. anti-ship) plus twenty HE with provision for time fuses, the extra ten HE shells being provided in wartime either in vertical stowage round the foot of the shell bins or in place of the ten practice rounds per gun.
46. It seems to have been clear that many ships would be badly cramped, so the report showed five alternative arrangements for twin 4in guns: (1) three twins on each side amidships; (2) two on each broadside and one on the centreline abaft them, saving one mount; (3) two on the centreline and two in the waist forward of them; one on the centreline, two in the waist, roughly as in the ‘E’ class; (5) two, en echelon so that four might bear together on some arcs, hence better than four single mounts. Note that the standard arrangement of later British cruisers, two mounts on each side, was not included.
47. Controller (Backhouse) memo copied in Lillicrap Constructor’s Notebook 5. Controller’s paper on the triple turret has not been found.
48. This explanation was given in Progress in Gunnery, but Campbell, Naval Weapons of World War II, p 36, attributes it to an attempt to reduce shell interference. In the Mk XXII mounting the cordite hoist was totally separate from the shell supply. The cordite hoist was outside the mounting, linking the handing room in the hold to a small (protected) compartment outside the barbette on the deck immediately below the turret. From here it was passed into the mounting and then handed up into the gunhouse.
49. Board Memoranda for the first half of 1932, ADM 167/86, p 2928 (memo dated 25 February 1932) is Controller’s memorandum for the Board on the problem presented by the big American and Japanese cruisers, which were of 10,000 tons and 8,500 tons respectively.
50. ADM 167/87, memorandum dated 22 December 1932.
51. Unfortunately this study has not been found in the Constructors’ Notebooks, and there is no trace of it in Board Memoranda. Bessant Constructor’s Notebook 3, p 58 contains details: 545ft (pp) 570ft (wl) x 58ft (extreme) x 16ft 6in, 50,000shp for 30kts. These provisional data were dated 24 July 1933. Armour data were not given (Bessant was calculating the resistance of a scaled-up Leander hull). This design was designated KVII, in a series which probably began with the ten-gun Improved Leander: there were two designs each for ten- and nine-gun ships, and two for the eight- and seven-gun Arethusas. However, K designations seem not to have been applied to these designs, at least in the Lillicrap Notebooks, so the missing first six K designs may have been later approaches to developing a more powerful cruiser.
52. Memorandum on cruisers by First Sea Lord, 25 July 1933, in ADM 167/89.
53. The undated Staff Requirement was inserted into the Southampton Cover as loose sheets. A note indicates that it was found, and added, on 17 October 1941. These sheets are marked Folio 121.
54. For each aircraft, DNAD wanted an 18in torpedo plus three 500lb bombs, six 250lb bombs, and twelve 20lb bombs, corresponding to a torpedo plus three aircraft-loads of bombs. In 1936 DNAD wanted the torpedo plus two loads of the heaviest bombs carried, one of the next heaviest, and one of anti-submarine bombs. That meant two loads of 500lb or ‘B’ bombs, one of 250lb bombs, and one of antisubmarine bombs. DNAD defended these loads on the grounds that only 5 per cent hits were expected, and that the ship might be weeks from a fleet base when she fought. The A/S and light bombs were considered useful with the fleet for disturbing patrols; on trade routes the 250lb bomb was more useful. Below a certain point aircraft which could carry heavy bomb loads were no longer worthwhile, so if there as not enough space for good bomb stowage it might be better to supply ships with seaworthy seaplanes rather than high-performance ones.
55. Constructor’s Notebook J L Bessant III.
56. In May 1934 E-in-C suggested a modification in which the boilers in the after boiler room would be placed side by side instead of in tandem, their shapes modified so that the outer shafts could pass under them. DNC rejected the idea on structural grounds. That was unfortunate, because the outboard voids associated with the tandem boilers allowed dangerous flooding.
57. War complement was set at 731. About 610 sleeping billets could be arranged in the normal places. If in addition turret spaces, ammunition lobbies, workshops etc were used, as approved for Leander, another 100 could be provided. However, accommodation would be worse than in later ships of the Leander class. The ship might have to be lengthened to provide the full 7.5 per cent margin over the calculated war complement. The peace complement was 706, including twenty-three officers (twenty-nine in War Complement), nine warrant officers (nine), and six midshipmen (six). Using Leander practices the ship could accommodate the full war complement, but the margin would be twenty-three rather than the desired fifty-two. The authorised margin had not been provided in any 6in cruiser to date. The worst problem was insufficient sleeping space, which was measured by the space between slinging billets (for hammocks). The alternative to providing more space was to sling hammocks closer together (reducing the space between billets from 21in to, say, 18in, the old destroyer scale; capital ships and the Kents had 24in, Leanders 21 in, and Arethusa and recent destroyers 20in; but the Medical Director General strongly opposed anything below 20in and preferred 24in). Another possibility, which was rejected, was to merge several small messes, e.g. to form one for all chief petty officers, including Engine Room Artificers. Various measures, such as reducing washrooms, which had already been accepted for the Leanders, were adopted.
58. The ship could now dock in No. 2 Dock at Gibraltar, in No. 5 at Malta, and in the Admiralty Dock at Hong Kong, all of which had been referred to in a CNS paper.
59. The belt and the deck covering it were extended 46ft forward, for a total protected waterline length of 232ft (40 per cent of waterline length), compared to 167ft (31 per cent) for D. The main part of the belt was reduced from 5in to 4½in and the magazine side protection somewhat reduced. It was now possible to place the transmitting station, main switchboard room, low power supply rooms, and auxiliary W/T offices, telephone exchanges, and lower steering position on the platform deck behind 4½in belt armour. The belt was covered by 1¾in decks and 2½in bulkheads. Magazine boxes had 4½in sides tapering to 3in at lower edge, decks 2in, bulkheads 2½in. Sides, ends, and decks of ‘A’ and ‘Y’ shell rooms were 1in. Turret ring bulkheads and recesses for shell and cordite hoists had 2in sides, 1in on ends, and decks made up to 1in as necessary. Shell and cordite hoists were in 1in tubes, with 3in armour rings where they passed through the armour deck. Protection to steering gear: 1½in sides, 1¼in deck. Bullet-proof protection to the bridge: 20lb side, 15lb roof; this was on the compass platform, plotting office, remote control office, wheelhouse, and after control positions. All boiler room and engine room fans were under main belt protection except after engine-room supply fans.
60. In the October statement 4in guns were credited with 200 rounds per gun, but that was increased to 250 in November, and the outfit per 2pdr barrel was increased from 1,400 rounds to 1,800. The November statement also showed six double Lewis guns, with 2,000 rounds per barrel. The 6in guns were given 200 rounds each.
61. Controller (Rear Admiral Charles M Forbes) pointed out that in the original design, aircraft had accounted for the weight of a 6in turret, but offered an offensive load of only 4,500lbs, compared to 67,300lbs for the turret, and had a scouting value which depended largely on the weather.
62. Dimensions: 572ft (pp) 600ft (lwl) x 61ft x 16ft 6in; 8,835 tons std. 72,000shp = 32kts. Endurance 7,000nm at 16kts. Armament four triple 6in, three twin 4in, three quad 0.5in machine guns, two triple TT, one heavy catapult and five TSR aircraft. Apparently at first plans called for four twin 4in, but reducing to three saved on the required complement. At about this time DNC stated that there was space for only three aircraft. Armour over magazines and bomb rooms: 5in NC on ⅜in D.1 on lower 3ft 6in of sides; 2in NC on ⅜in D.1 crowns, 3in NC on ⅜in D.1 ends. Shell rooms: 1in D.1 sides, crown, ends. Machinery: 5in NC on 1in D.1 sides, 1¼in D.1 crown, 2½in NC on ½in D.1 ends. TS, LP supply, HA calculating position: 1in D.1 sides and crown. Steering gear 1½ to 1¾in D.1. Turrets: 1in NC shields and rings, 1in D.1 supports; 1in D.1 ammunition lobbies, etc. Bridge: ⅜in BP
63. Approved by DNC on 30 October 1933.
64. Alternatively, the second HADT could be added aft; that was rejected to avoid smoke interference. Placing the two directors forward offered a more self-contained arrangement of guns and control positions on each side, as well as better arcs of training on after bearings.
65. The belt and 50lb deck were extended aft over the 4in magazine (a distance of 33ft), which therefore did not need their own 4½in side armour. The existing 2in crown was reduced to 1½in which, with the 1¼in deck immediately over it, was equivalent to the 2in against bombs. The extended belt and deck protected the main W/T office on the platform deck at the waterline. DNC pointed out that 4½in C armour could resist the 6in 112lb shell beyond 7,000yds at 30° to normal (5in C would resist the same shell at 6,000yds). The 1½in D.1 deck would be immune at ranges below 16,000yds (the Staff Requirement). DNC added that it would cost 42 tons to add ¼in to the belt, and 84 tons to add ½in (to reach 6,000yd immunity). Thickening the 1¼in deck to 1½in over the machinery would cost 69 tons. That addition would raise the minimum altitude at which the deck would resist a 250lb bomb only from 2,000ft to 2,800ft; as heavy bombs would presumably be dropped from greater heights, this was not worthwhile. Thus DNC preferred simply to extend the belt and deck aft. To do that while retaining the existing 2in deck over the 4in magazine would cost 42 tons, and the magazine would still be open to 6in gun fire above 16,000yds through the wing portions of the lower deck (which would have to be thickened to 2in, giving a total of 75 tons). Thickening the belt to 4¾in would bring the total to 120 tons, absorbing the entire margin.
66. Ordered on 21 November 1935 by DNC in accordance with verbal orders from Controller. Bessant Constructor’s Notebook 3 includes a report dated 6 November 1935 on moving the forefunnel back by angling the smoke pipe back above the upper deck. At most the funnel could be moved 9ft back in this way, using supports from the forecastle deck. That would have cost weight and upper deck space. Anything more would have required lengthening the ship.
67. She was chosen because she would entail the least disturbance. Sheffield duplicated Newcastle because they were building at the same yard. The modification seems to have been proposed in March 1935.
68. J L Bessant Notebook IV (308/4) dated 1934-6.
69. ADM 1/27412.
70. This total had no binding force. It was adopted to allow for growth beyond the 9,000 tons of the original ships.
71. Surviving Constructors’ Notebooks do not show Designs KXI through KXIII. Presumably the developed Southampton design was KXI. It seems unlikely that KXV was the first attempt at a cruiser with quadruple turrets, so it may have been KXII or KXIII.
72. The original quadruple-turret design was 589ft pp/616ft wl/623ft 6in oa x 63ft 4in x 17ft 3in x depth 33ft 6in; the new one was 579ft/606ft/613ft 6in x 63ft 4in x 17ft 3in x 33ft 6in, 10ft shorter but with the same standard displacement of 10,000 tons. In both cases, speed was 32.5kts at standard displacement and 31kts at deep displacement. The only difference in armament was substitution of triple for quadruple 6in guns, in both cases with 200 rounds per gun. In addition, both designs showed four quadruple pompoms and two quadruple 0.5in machine guns (the change to octuple pompoms was not shown), plus two triple torpedo tubes. The new Legend showed the belt tapered to 3in over magazine sides (forward magazines had box protection). In both versions the belt was closed by 2½in C bulkheads, but in the revised version the deck over the belt was 2in NC instead of 1¼in D. In the earlier version magazines behind the belt had 2in crowns rather than the 1¼in deck elsewhere. Turret protection in both versions was 4in C face, 2in NC sides, and 2in roofs. Turret support rings were 2in to 1in NC. Bullet-proof plating on the bridge was ½in deck, ⅜in sides. Steering gear protection (in both versions) was 1½–1¼in D over steering gear.
73. Magazine crowns, whether in boxes or behind the belt, had 3in rather than 2in armour. Forward magazine sides were 4½in (like the belt), tapered to 3in at their lower parts. Shell rooms were above the magazines, for additional protection. The steering gear was given a 2in deck and 1½in D sides and ends. Low-power cables carrying data from the transmitting station and HA calculating positions to the DCTs and HACS were enclosed in trunks of 20lb bullet-proof plating, apparently an innovation.
74. This amounted to protection for exposed personnel including gun crews; radar, RPC for the pompoms; stiffened upper deck; modified transition plating between the forecastle and the side; a modified forecastle deck forward; a modified sheer strake; a new transverse bulkhead at 104 station aft; and Oerlikons.
75. ADM 1/9355.
76. Lillicrap Constructor’s Notebook 6 describes the design process, beginning with the assignment of Designs P through T by Director of the Tactical Division, 13 April 1934:
Sketch designs for the 4,500-tonners were scaled down from Arethusa, with similar machinery spaces and weights. Based on the ‘M’ class, to protect magazines against 6in fire required 4½in sides (tapering to 3in) and 2in decks. In 1928 the Tactical Section had called for machinery protection against 4.7in (destroyer) fire above 4,000yds. That required a 3in side (effective at 4,000yds at a 30° angle to the normal, and at 5,000yds at any angle) and 1in deck (proof to 15,000yds). Against the 5.1in gun, then being considered for future destroyers (and capital ships, as secondary armament), the 3in side was proof above 9,000yds (7,500yds at a 30° angle). For Design P, Lillicrap added up hull (scaled from Arethusa), armament, general equipment, protection, and machinery weights not proportional to power, and subtracted from the allowed 4,500 tons to get what he could have for the powerplant: 840 tons, which scaled down from Arethusa’s 64,000shp plant to give 47,500shp, which in turn gave a speed of 30.75kts using curves of known hulls and scaling. For Designs R and S, with their required speeds, he went the other way, finding that they needed all of Arethusa’s 64,000shp. Given armament and machinery weights, he subtracted to find what was left for protection. It was not much. Taking Arethusa magazine protection (weaker than required), he was left with enough for either a 1¾in belt (plus ½in of hull steel) and a ⅜in deck or a ¾in belt and 1in deck. Design Q had to exceed 4,500 tons, because adding up the basic hull, machinery as in Design R (to make 33kts), protection as in Design P, and general equipment accounted for 4,450 tons. The heavier ship needed a bit more power, and that made her heavier still, with more protection to cover somewhat larger machinery spaces. Design T was derived from Design Q using triple turrets weights from the new ‘M’ class (Southampton class). Compared to Arethusa, armament weight was about 20 tons less, but protection was 120 tons more, e.g. for the ammunition lobbies. The ship also needed more generator power.
Controller then asked for an additional design, for a small fleet cruiser not to exceed 3,800 tons, with five or six 6in guns and two quadruple torpedo tubes, to make 38kts light and 35kts deep. Endurance was 6,000nm at 16kts. If practicable, the ship should have a 1in belt and ½in or ¾in deck over the machinery, with side protection over the magazines. A six-gun design should have two twins forward and two singles aft, i.e. for maximum chasing firepower. A five-gun design would have three guns forward (one twin and one single or three singles, two en echelon) and two singles aft. Work began on 9 May 1934. Instead of scaling up from the 3,000-tonner sketched in 1930, Lillicrap again scaled down from the much larger Arethusa, presumably because she represented more modern design practice. He found that her hull form was too resistful; he was forced back to the 3,000-tonner, which had been based on destroyer hull form. Machinery had to follow destroyer leader rather than cruiser practice. Even so, it proved impossible to reach the desired 38kts. This sketch design was submitted on 2 July 1934, becoming the basis for a further alternative, Design U.
ADM 1/8828 includes TD Memorandum 126, ‘Memorandum on Small Cruisers,’ dated November 1934, which includes the table of design alternatives, with approximate Legends. Design data were:
77. Magazine sides reduced from 4in to 3in, belt over machinery from 3in to 1¾in and deck from 1in to ⅜in or, alternatively, to ¾in side and 1in deck.
78. C-in-C Mediterranean wanted twelve 4.7in HA guns on 3,500 tons with a speed of 26kts, with maximum possible protection, Asdic and good searchlights. A quick estimate suggested that protection would be the usual 3in sides and 2in crowns for magazines and 2in sides and 1in deck over machinery.
79. According to a paper summarising views on the small cruiser, in ADM 1/8828, C-in-C Mediterranean envisaged the cruiser force in the war zone as five 8in, four Leanders, four Arethusas, and eight new-type AA cruisers. The staff considered this too weak to meet Japanese reconnaissance forces. C-in-C Mediterranean wanted a policy of two AA cruisers to every five Southamptons, until eight were available.
80. ADM 1/9384.
81. This was the largest calibre allowable for destroyers under the 1930 treaty, and it was used aboard French destroyers. The Royal Navy considered it for future destroyers, but dropped it after unsuccessful trials. In 1932, when new battleship designs were being prepared as a basis for the British official position at the 1932 Geneva conference, the twin 5.1in was proposed as their dual-purpose secondary gun. The Staff Requirement for the 5.1in gun was issued in March 1932. It was to fire ten rounds per gun per minute, with elevation limits of +70° and –10°, and with a maximum surface range of 18,000yds. Ceiling was to be 10,000ft at a plan range of 16,000yds. Time of flight to this range should be less than 45 seconds. A battleship should have nine or ten guns on each side. To achieve the desired rate of fire, the gun had to fire fixed ammunition (no separate cartridge case) and the round had to be within the 110lb limit for man-handling. That in turn implied a 70lb shell; the performance requirement set a muzzle velocity of 2,500ft/sec. The 70lb shell implied 5.1in calibre. It is not clear why early versions of the King George V design showed 4.5in rather than 5.1in guns. The 5.1in in turn gave way to the 5.25in selected as the battleship secondary gun.
82. ADM 1/9384 (and also the Dido Cover) includes TD 83/35 of January 1936, ‘Staff Memorandum: The Small Fleet Cruiser and RA(D)’s Flagship: Examination of Design 5.25-B’. The memorandum assumed that the choice lay between five 6in guns and ten 5.25in, declaring the latter clearly superior, the 80lb shell being very well suited to a ship of this size, sufficient to deal with a small cruiser. Greater numbers offered a much better chance of hitting an evading destroyer. The powerful anti-aircraft armament would be a considerable fleet asset. The extra 250 tons, to provide alternating engine and boiler rooms, was well worth while. At this stage combined HA/LA DCTs were envisaged. The small fleet cruiser would work in the screen, for shadowing, for supporting British flotillas against enemy destroyers, and, under cover of more powerful forces, for offensive raiding and patrol operations with light craft. DNC was asked to consider two more possibilities, Designs C and D. Design C was Design B with two triple torpedo tubes, with 3in rather than 2in deck protection over magazines, and with a light catapult. Flag and plotting facilities would be as in a cruiser flagship rather than RA(D) flagship. Design D was as Design C except for a 3¼in deck over magazines. DTD wanted to know the effect on Design C of a heavy catapult, or 400 rather than 300 rounds per gun, and the effect if one set of machinery spaces was given 3Min deck protection rather than the 1in envisaged. It is not clear from the printed version whether the increased deck protection in alternative D was 2Min over the machinery rather than the magazines. The RA(D) ship was compared to two US designs recently reported by NID, A and B (neither of which can easily be identified). A was a 29-30kt 3,500-tonner armed with eight 5in guns, thirty-two 0.5in machine guns and four triple torpedo tubes, a catapult and three aircraft, and 3in belt and 2in deck protection. B was a 29-30kt 5,000-tonner armed with two triple 6in guns, thirty-two AA machine guns and four triple torpedo tubes, the same catapult and aircraft, and 3in to 4in belt and 2in to 3in deck armour. ‘The USA ships have well-protected turrets, somewhat better decks, and include a Torpedo armament and aircraft, but it is recognised that the figures are not conclusive.’ It was estimated that a 29.5kt version of the 4,750-ton RA(D) design would displace about 4,200 tons. The table also included the slightly earlier T design (5,600 tons, 35kts, two triple 6in), a 29.5kt version of which would displace about 4,900 tons. There is no indication that DNC produced Designs C and D, although he did head off calls for 3in deck protection.
83. Lillicrap Constructor’s Notebook 6 shows how the two alternatives and their successors worked out. The requirement for the RA(D) ship, TD 83/35, was dated 1 June 1935. Lillicrap once more began by scaling down the Arethusa and her machinery, and adopting the earlier arrangement of undivided boiler rooms. Taking armament weights into account, he had enough protection weight to provide the 3in sides and 2in deck magazine protection of the larger ship, leaving enough for 3in side and 1in deck over the machinery. On this basis he could achieve 33kts on a 4,500-ton hull 470ft long. With alternating engine and boiler rooms the ship would displace 4,700 tons (at 475ft). Another calculation showed that reducing speed from 33kts to 31.5kts could save 200 tons, enough to increase the HA battery to two twins and to add 1in to the magazine sides. With alternating boiler and engine rooms, the resulting ship would displace 4,500 tons (475ft long). If the ship had the earlier undivided boiler rooms, the weight could go into another inch of belt over the machinery and another quarter-inch of deck over the machinery: 4,500 tons (470ft). Alternatively, with the original armour and the single 4in guns, the ship could be cut to 4,100 tons (455ft). On a 4,700-ton, 475ft hull, with undivided boiler rooms (63,000shp for 33kts), Design B had enough weight to protect machinery but not magazines against the new 4.7in shell at 5,000yds: 3½in side and 1in deck over machinery, 3in side and 2in deck over magazines – whose sides would be protected almost completely by water at that range. Armament weight was based on 250 rounds per gun plus 50 star shell per ship. Particulars of Designs A and B went to DNC on 28 June 1935.
84. On 8 October DNC, presumably at Controller’s request, asked for the cost of a 5,000-ton ship with ten 5.1in HA/LA guns in twin mounts, capable of 32–33kts. Lillicrap took Arethusa as his basis, with about the same hull and protection. He estimated that the ship would cost £1.07 million. On 24 October DNC asked for a 4,500-tonner on the lines of the RA(D) ship armed with five twin 5.25in guns and two quadruple pompoms, to make 32kts. The ship would carry neither torpedoes nor aircraft. A marker in Lillicrap’s Notebook at this point is ‘Dido and other cruiser designs’. The 5.25in gun had not yet been completely designed, and on a ship this size a slight change in armament weight could have considerable effect. Lillicrap began with the 4,500-ton (470ft x 48ft x 13Mft) ship, recalculating power for 32kts (say 54,000shp). Available tonnage for protection gave him 3in sides and 2in crowns and ends for magazines, leaving enough for 2in side and 1in deck over the machinery (a re-estimate added 1in ends to the machinery). Lillicrap noted that the available length was too little, but his Notebook does not include any calculation of what was needed. This ship had adjacent boiler rooms, and greater length would be needed if the more survivable alternating engine and boiler rooms were adopted. the sort of lengthwise arrangement sketch which became common a few years later. Lillicrap discussed the design with DNC on 2 November, and in view of difficulties in providing enough length (probably for the five centreline 5.25in mounts), he proposed an alternative in which extra length was used to provide alternating engine and boiler rooms (which he called the Modified Leander machinery layout). Lillicrap justified adding 15ft to accommodate the new machinery arrangement and recalculated hull weight. He ended up with a 4,750-ton (485ft x 49½ft x 13¾ft) ship requiring 55,000shp to make 32kts. More weight was available for protection, so he could provide 3in side armour (and the earlier 1in deck) over the machinery. Lillicrap reported this design and the 4,500-tonner to DNC on 5 November, and they were forwarded to Controller as Designs A and B. The 485-footer became the basis for the Dido design.
85. At its meeting on 28 November 1935 the Board decided that the 1936/7 cruiser programme would comprise two Southamptons, the RA(D) flagship, and two further cruisers of a type to be decided depending on how the London Naval Conference came out. At that time the conference seemed close to collapse, and it seemed unlikely that any reduction beyond 10,000 tons would be achieved. Nothing smaller than a Southampton should be built until it was clear what foreign powers were doing. It was certainly not clear that the proposed 8,000-ton limit would be ratified. On 25 February 1936, however, First Sea Lord compared the British underage cruiser force with the combined forces of Germany and Japan, the two likely future enemies. The Royal Navy enjoyed parity in 8in cruisers built and building, superiority in large 6in cruisers (8,500 tons and above: ten to six) and in medium 6in cruisers (5,500-8,500 tons: ten to six), but inferiority in small 6in cruisers (nine to fifteen). Chatfield therefore decided that further construction of large cruisers should be suspended in favour of small ones, about 5,000 tons, any of which might be suitable as RA(D) flagship. The two cruisers of undecided type became two more Didos. They offered an additional advantage in that they were less expensive than the ships originally included in the programme, as it was reported to the Defence Requirements Committee. As the international situation worsened, in June the DPR (the Sub-Committee [of the Committee of Imperial Defence] on Defence Policy and Requirements) formally requested ways to accelerate the building programme. It was told that in each of the next three years seven rather than the planned five cruisers could be ordered. The 1936/7 programme was the first affected. ADM 167/95, Board memoranda for 1936. The acceleration memo is DPR.88. The new programme was brought before the Board at its meeting on 24 June 1936. The fourth and fifth Didos were part of the Second Supplementary Estimate of July 1936.
86. The main points agreed at a meeting on 12 February 1936 in the First Sea Lord’s office, with First Lord (CNS), Controller and DCNS present were that they should be prepared to go to 5,000 tons; that the ship should have 300 rounds per gun; that she should have two triple torpedo tubes; that magazine crowns should be thickened from 2in to 3in; that the forward gun might be on the forecastle deck; and that the questions of having octuple rather than quadruple pompoms and a thicker deck over machinery should be considered. The proposed additions (other than a thicker deck over machinery) added up to 152 tons, the largest items being extra ammunition (33 tons for 5.25in, 48 tons for pompom ammunition and pompoms). Lillicrap thought the ship would have to be lengthened 10ft to accommodate larger magazines, adding another 80 tons. That brought her to about 5,000 tons (495ft x 49½ft x 13¾ft). At this point the 4,500-ton Ship A was still in contention. Lengthening her by 10ft would bring her to about 4,750 tons (480ft x 49ft x 13½ft) or even 5,000 tons if she was given 6in more draft. That in turn would require more horsepower and more machinery weight (and length, to be protected). Another possibility was to lengthen the belt 60ft and thicken it from 3in to 4in. On 8 February Lillicrap and another constructor, Jackman (who had investigated hull weights in greater detail) produced a table of alternatives for Controller (submitted 19 February): see below.
Controller did not see any great benefit in a 5,000-tonner, and preferred A1 but with 3in magazine crowns. ACNS preferred B2 (alternating machinery spaces). The 3in belt could be penetrated by the new 6in shell (112lbs) at 16,000yds at 90° or at 13,500yds at 60°. Penetration by the new 4.7in destroyer shell was doubtful at 40,00yds. Comparable figures for a 4in belt were 11,000yds and 9,500yds; it was immune to 4.7in fire. A 3in deck would keep out a 1,000lb bomb dropped from 5,000ft and a 500lb bomb dropped from 9,000ft. There was little to choose from between a 1in and a 1¼in deck. The file on the RA(D) cruiser includes small-scale drawings of cruisers with adjacent and alternating engine and boiler rooms labelled P (495ft x 50ft x 13?ft, 5,150 tons) and Q (506ft x 50½ft x 14¼ft, 5,300 tons), which were supplied to support the decision for or against this arrangement, with BD guns.
87. Machinery was covered by a 3in belt and 1in deck and ends. Magazines had box protection: 3in sides and roof and 2in ends. Other key spaces, including shell rooms and transmitting station, had 1in sides and crowns. These thicknesses in Controller’s memo to the Board do not quite match those in the Legend submitted at the same time. The Legend shows turret protection of ½in bullet-proof plating rather than the 1in planned earlier.
88. Only on 5 May did a meeting in DNO’s office choose a mounting design from among several offered by Elswick. It was an upper-deck mounting with a 1in shield weighing 75 tons including ammunition hoists. At this point the heavier BD mounting was introduced; it seemed that the choice lay between four BD and five upper-deck mountings, the former saving 6 tons by eliminating one set of magazines and their protection. Expected revolving weights were 56 tons and 70 tons. Making all five mountings the heavier BD type would add 170 tons. It was pointed out that the extra weight could be balanced by reverting to adjacent engine and boiler rooms, but that was not acceptable; instead, the ship was allowed to grow to 5,300 tons. An investigation into the stability of a ship with the BD mounting showed the need for increased beam, so that was made 50ft 6in. Figures balanced well enough for a 506ft x 50½ft, 5,300-ton ship, and powering estimates suggested that it needed 58,000shp (32,000ehp with propulsive coefficients typical of small light cruisers). The modifications planned were reported to bidders on 3 February 1937 (the extra beam had not yet been chosen), and all the modifications were approved by Controller on 8 March. ACNS was gratified that the magazines and shell rooms were now protected against 6in plunging fire at 19,000yds and below, and against 250lb dive bombs. The Board formally approved these changes on 15 April 1937.
89. When the decision to proceed was made in April, 1936, it was £1.3 million. When the Board approved the 5,300-ton design, it was £1,340,000. The 5,450-ton design was expected to cost over £1.6 million.
90. Cruisers and larger ships were provided with a closed lower bridge below the compass platform. In August 1938 Director of Navigation (D of N) attacked this practice, arguing that the compass platform was and should be the primary conn, and that simply having a closed bridge reduced the effort to make the compass platform entirely efficient. ‘After many years of handling H.M. Ships, D of N would not permit a closed bridge to be used for this purpose in any vessel he may command. The “lookout” is inevitably more restricted and subject to “blind” spots. Sound signals are less likely to be heard and the difficulty of estimating their direction would be much greater. The risk of the bridge personnel becoming drowsy at night is immeasurably greater. During dark hours, the faintest light, e.g., the glow from the pelorus repeater, would make the windows opaque from within, except for external objects having their own illumination … the use of a boxed cubby-hole from which to control the ship entails a degree of risk which in D of N’s opinion should not be accepted.’ DNE argued that in some (unspecified) situations a ship should be conned from a closed position. It was decided that the existing arrangement, in Manchester and later large cruisers, was too elaborate; the upper steering position (two decks below the upper bridge) should be combined with the existing closed bridge. No change would be made in Manchester, Liverpool and Gloucester.
91. The 5,300-ton design showed only one DCT on the bridge (both DCTs were HA/LA at this stage). Initially the design showed only a single transmitting station (and two HA control positions). ACNS pointed out that there was a design requirement that the armament be split for HA fire; it might as well be split for LA, too.
92. Controller announced this change at a meeting on 3 September 1939; he hoped to accelerate completion not only of these two ships but also of the two following them. They were to be adapted as RA(D) flagships because they would be the only two of this subtype, because the lack of gunpower was more acceptable in an RA(D) flagship, and because RA(D) would now be able to fly his flag in a more up-to-date ship. Aurora and Galatea would be released from RA(D) flagship roles to form a homogeneous squadron with their sisters Arethusa and Penelope. DSD agreed, subject to the ships being arranged so that ultimately they could be rearmed with 5.25in guns (which was never done). Projected completion dates for Scylla and Charybdis were June and May 1941; actual dates were June 1942 and December 1941. The armament statement on 20 November 1939 showed 400 rounds per gun. The ships retained their pair of quadruple pompoms and their torpedo tubes. In October 1941 plans called for them also to have a single 4in star shell gun, fifteen depth charges (rather than the original six), two single power-operated pompoms (S.B. Mk VIII), and six Oerlikons (rather than the four planned earlier). Endurance was given as 5,500nm at 16kts. On full-power trials on 3 December 1941 Charybdis made 30.774kts at 62,440shp at 6,800 tons.