THE SLIDE TOWARD WAR
The 1936 Treaty
The seventy-cruiser policy collided with the demands of the 1930 treaty that many older cruisers be scrapped by 31 December 1936. On 6 October 1934 First Sea Lord Chatfield signed a Board memo advocating retention of many of these ships. He estimated that on 31 December 1936 the British Commonwealth cruiser force would include thirty-seven underage ships built and building plus twenty-two overage ships eligible for scrapping and replacement, including three ships earmarked for scrapping in 1935 (Brisbane, Constance and Castor). Ten of the latter group, including the four Hawkins class, would be scrapped under treaty terms. Given the grave world situation, Chatfield considered the shortage of cruisers a critical problem. The government had now accepted that seventy was the minimum acceptable, but there seemed to be no hope of paying for more than three or four a year. Chatfield hoped to gain more tonnage at the coming London Conference (1935) and also to change the scrapping programme. He also needed more personnel, having only enough for forty-four ships, Australia contributing another four and New Zealand two. The United Kingdom would have to provide the remaining twenty ships, of which eleven would be in full commission and nine in reserve. He estimated that he could meet manning requirements by 1942 without opening a new Boys’ Training Establishment, so he therefore made 1942 the target year for filling out the seventy ships. Instead of scrapping the four Hawkins (which exceeded the limit on cruisers with guns of more than 6.1in calibre), he planned to rearm them with 6in guns, on the grounds that their high endurance and good sea-keeping made them more valuable than ‘C’ or ‘D’ class cruisers.
Curacoa typified the production version of the anti-aircraft conversion, armed with twin 4in guns and one quadruple (rather than octuple) pompom in place of her original five 6in guns. Two quadruple 0.5in machine guns replaced the two 3in anti-aircraft guns originally mounted abeam the forefunnel; by the time of this photograph they had been replaced by Oerlikons (Curacoa had a total of five, plus two single pompoms, which are probably the objects abaft No. 3 twin 4in gun). Conversion originally entailed erection of a pole mainmast carrying a DF coil, but when ships received air-warning radar (Type 279 in this case, during conversion, in January 1940) the DF coil was moved to between the funnels. The Type 271 surface-search radar was installed some time before September 1942 (the ship was lost that October). Type 285 HA control radar was installed in June 1941.
All of these factors suggest that this photograph was taken shortly before the ship was lost in a collision with the liner Queen Mary.
The Naval Staff hoped to ‘isolate’ the big new 6in gun cruisers at the coming conference, in which case the Hawkins class would be among the limited number of such ships the United Kingdom could retain, although they would clearly be inferior to newer foreign counterparts. If no such isolation proved possible, they should be rearmed and retained, ‘C’ class cruisers being scrapped to make up for any tonnage ceiling. Frobisher was already used for cadets’ sea training, and no other suitable cruiser was available, so if the class was retained, she would continue in this role. If not, she would be demilitarised and retained. In the end, Chatfield got both ‘isolation’ of the biggest cruisers and, when the British government invoked the escalator clause of the 1930 treaty in July 1936, the right to retain all of the ships which would otherwise have been scrapped, the latter.
The capital ship building ‘holiday’ of the Washington Conference and the London Naval Treaty of 1930 expired on 31 December 1936. Despite a dramatically worsening international situation, neither the British nor the US government could easily abandon naval arms control, because in both countries much of the population saw arms control as a guarantee that nothing as brutal as the First World War could happen again. Few wanted to acknowledge that several major governments seemed to welcome another war; surely no one could be that insane, particularly in Europe. Rearmament was therefore a reluctant process. During preparatory talks for the next naval conference, in March 1934, it became clear that neither the United States nor the United Kingdom would accept the Japanese demand for parity, although both hoped that some deal could be arranged.
The Admiralty Board revived its demand for seventy cruisers: sixty modern and ten overage ships (twenty overage if necessary). First Sea Lord Chatfield said that he would prefer no agreement at all on cruisers to abandoning the seventy-cruiser goal. The British also wanted to do away with the new large 6in cruisers, Chatfield proposing a 7,000-ton limit on future cruisers. The US Navy was willing to accept a limitation on the size of future cruisers, once it had completed the nine large Brooklyns and it had its eighteen 8in cruisers. US negotiators pointed out that the American public expected arms reductions, not the increase envisaged by the British; in June 1934 President Franklin D Roosevelt called for a ten-year treaty which would, over that period, cut strength by 20 per cent. The British answered that they now faced both European and Far Eastern threats, and that they could not depend on the United States to help them against the Japanese, and they therefore demanded substantial modification to the 1930 London Naval Treaty. Complicating the situation was escalating rivalry between the two Washington Treaty powers which had not signed the 1930 treaty, France and Italy. Given the Japanese insistence on parity, which was utterly against British and US interests, by October 1934 Chatfield doubted that any treaty could be negotiated. British attempts to negotiate with Japan were fruitless, but it proved relatively easy to work with the Americans. Through late 1934 and early 1935 British also talked with the French, and with the Germans, who wanted an agreement allowing them a fixed percentage of British naval strength.
By July 1935 the British favoured qualitative limitation, such as a cap on cruiser tonnage (at this stage, 7,600 tons). Overall tonnage limits were not as important, though still desirable. The British came to value a requirement that each country announce its building programme in advance. Any power deciding to accelerate construction would risk other powers reacting before its planned ships could be completed. US reactions were mixed; the US Navy’s General Board correctly predicted that the British would support qualitative limits and the announcement provision. Despite considerable pessimism, on 23 October 1935 the Cabinet decided to issue invitations to a naval conference. President Roosevelt told his delegates to continue to press for limits on overall tonnage, but admitted that once the Japanese demands were rejected that would become unlikely. He therefore hoped for a tripartite (with Japan) qualitative limit backed by an escape clause which would deter non-contracting parties. This was the evolving British position. The British and American delegates met informally in London the day before the conference opened, agreeing to resist the Japanese demand for a common upper tonnage limit (i.e. parity). In January 1936 the Japanese withdrew from the conference altogether, having been as disruptive as possible up to that point.
Only Britain, France and the United States ratified the resulting treaty, which eliminated all overall tonnage limits. It incorporated an Italian proposal that programmes had to be published annually within four months of 1 January, that at least four months had to elapse between announcement and keel-laying, and that programmes could not then be changed (Italy did not ratify the treaty). An 8,000-ton limit was placed on cruiser tonnage, and the ban on new 8in cruisers repeated. Without any limit on total battleship tonnage, any power might build super-cruisers it called battleships. To make that difficult, the treaty placed a lower limit on battleships, 17,000 tons and 10in guns (35,000 tons and 14in guns was the upper limit). To encourage Japan not to exceed that limit, the treaty contained an escalator clause allowing for increases in the event a non-signatory refused to accept these limits; the signatories were to consult together to minimise their departure from the agreed terms. On this basis battleship gun calibre was soon increased to 16in and two years later maximum battleship tonnage was increased to 45,000 tons. A special clause for cruisers allowed any party to announce that it was building cruisers of up to 10,000 tons, which could be armed with guns of greater than 6.1in calibre. This would release other signatories from their obligation not to build such ships. British constructors’ notes suggest that on two occasions the British government considered activating the escalator clause for cruisers (but they did not do so). The treaty was signed on 25 March 1936, coming in effect on 31 December 1936. It was intended to run until 31 December 1942.
Anti-Aircraft Rearmament
Rearmament funds made it possible to carry out some of what the 1932 air defence committee had suggested. The multiple HACS (and pompom directors) envisaged by the 1932 committee made it possible to split anti-aircraft fire in various ways, so the Royal Navy devised a centralised system controlled by an Air Defence Officer (ADO) in an Air Defence Position (ADP), communicating both with the captain and other control officers, and with the guns.1 Ideally the ADP was a central position high enough in the ship to be clear of funnel haze and cordite smoke, with an all-round overhead view, close to, but independent of, the compass platform. In 8in cruisers that meant a position at the front of the DCT barbette; in light cruisers, the roof of the captain’s shelter. The ADO maintained an air plot which, in theory, allowed him to allocate his ship’s anti-aircraft resources. That plot in turn was adapted to radar inputs during the Second World War. ADO was linked both to air-defence lookouts (who sent target bearings mechanically from their sights [target bearing transmitters in US parlance]) and to the weapons via their directors. Short-range weapons were grouped (one Mk M pompom, which had a director, and one quadruple 0.5in machine gun). Particularly when aircraft were attacking at short range, voice control was impossible due to the noise, so mechanical connections between ADP and other elements of the system were essential.
Above and below: Clearly connected with the anti-aircraft rearmament programme was a pair of prototype anti-aircraft cruisers, HMS Coventry and Curlew. HMS Coventry is shown as completed. She had two items in short supply at the time, a pair of HA directors (with their associated computers) and an octuple pompom in ‘B’ position. The main battery was ten 4in guns; conversion was simplified by using the existing structure: ‘A’ and ‘B’ positions for 4in and pompom, the positions for two 3in antiaircraft guns abeam the forefunnel, the platform for No. 3 6in gun abaft the after funnel for two more 4in guns, the after 6in positions for 4in guns, the after 3in antiaircraft position for the pompom, and an additional position on the quarterdeck. The after pompom was removed from both ships in 1938-9 and replaced by a pair of quadruple 0.5in guns. About 1939 the stub mainmast was replaced by a pole carrying a DF loop. (James Fahey collection of US Naval Institute).
When the air threat was re-evaluated in January 1937, the emphasis was still on long-range barrage fire to break up attacking formations, including high barrages by guns without specific anti-aircraft control. However, the dive-bombing threat was fully accepted (and was expected to increase in importance). Future aircraft would fly faster at higher altitudes, so fire would have to be opened at greater ranges (8,000-12,000yds) to sustain fire long enough to inflict damage, and to keep enemy aircraft out of ranges from which they could make dive-bombing attacks. Attention was (and had been) paid not only to the needs of the fleet at sea, but also to fleet air defence at anchor, because the fleet would often have to operate from unprepared anchorages, even places like Alexandria, the main Eastern Mediterranean fleet anchorage.
A key point entirely missed by the Royal Navy and by all other navies at this time was that anti-aircraft fire tended to ruin pilots’ aim rather than actually shoot aircraft down. Thus the fact that light weapons fired tracer was essential, because pilots saw the tracer flashing past them, but for the 1937 report, tracer was valuable only to the extent that it showed large errors and indicated (to those on the ship) which target was being engaged.
The upgrades, class by class, proposed by the 1932 anti-aircraft committee in effect formed the basis of the fleet rearmament programme. Each ship should have two HA control systems, at the least offering simultaneous control on both sides. Given foreign improvements, ‘it is clear that an increase in the strength of AA armaments throughout the British Navy is overdue’ (this passage in the report was underlined). Due to lack of space, the maximum battery possible in cruisers and capital ships was four guns on a side in twin mounts. Unfortunately that was not enough (given existing standards of control) to bring down one aircraft in a group of three.
Coventry shows early-war modifications in a 29 June 1940 photograph, including the addition of splinter shields around her 4in anti-aircraft guns. She had been fitted with topmasts (to take air-warning radars), so her forward HA director was relocated to the fore end of her foretop. The two 4in guns abeam her forefunnel had been landed, presumably to reduce topweight. The ship did not receive Oerlikons until a refit at Alexandria in May 1942. She was sunk on 14 September 1942.
In the committee’s view the Leanders had the maximum battery they could accommodate, and they could not take any Mk M pompoms. Given the HA capability of the 6in guns, it was essential that they be given some form of fire control. There was no space for a second HADT. The positions selected for the four 0.5in machine guns were the best possible.
In York and Exeter a second HADT could be bracketed to the bridge structure, above and abaft the main DCT (the sided positions suggested for larger 8in cruisers were not practicable). The 4in single mounts could be replaced by weather-deck twins. A Mk M could replace the existing single pompoms, but the position was close to ‘B’ turret, and there was too little space. The single pompoms could be replaced by quadruple 0.5in machine guns, but these positions were untenable when the 8in guns fired. The alternative was to place them atop the roofs of ‘B’ and ‘X’ turrets. Ideally two could go on the turret tops and two in place of single pompoms.
Norfolk and Dorsetshire had their single 4in guns amidships, on a boat deck surrounding their funnels, rather than aft, as in earlier large cruisers. This position apparently made it possible to replace each single mount with a BD twin, as there was considerable space under the deck. Two Mk M pompoms could be mounted inboard of the 4in guns, between Nos. 1 and 2 funnels. Four quadruple 0.5in machine guns could be mounted on platforms built out from the deck supporting ‘B’ turret. HADTs could be placed alongside and below the DCT, and directors for the Mk M pompoms on the after corners of the upper bridge.
The Kent and London classes had their single 4in guns on a boat deck somewhat further aft, alongside the second and third rather than the first and second funnels. This platform was built above the torpedo tubes; there was no surplus space underneath for a BD mounting. Instead, the Committee proposed twin mounts for the 4in guns, and positions for the Mk M pompoms alongside the forefunnel. BD mounts could be installed if the torpedo tubes were moved (one in place of an existing 4in gun, one in place of single pompoms), but then the Mk M pompom would be subject to 4in blast. The ships had high bridge towers carrying their DCTs; the Committee envisaged placing the two HADTs alongside.
The two Emeralds already had an awkward anti-aircraft battery arrangement, with two guns on the sides and one at the after end of the after gun deck, abaft the superimposed 6in gun. The Committee suggested replacing all three with twins. The ships already had HADTs between their after control and after funnel, a position considered preferable to the original AA control position on the after control, to keep it clear of smoke. However, the ships were scheduled to receive new tripod masts just forward of their after funnels, which would block after arcs. No forward position was available, so the Committee proposed placing the director in the original AA control position atop the after control, accepting the smoke problem but expecting it to be reduced when the funnels were raised, as planned, 15ft (when the new mast was stepped). No position was available for a second HADT, or for pompoms, but there was space for two quadruple 0.5in machine guns. The Committee made no recommendations either for the ‘C’ and ‘D’ class cruisers or for the Hawkins class.
Above and below: The anti-aircraft cruisers proved so valuable that two more ships, HMS Caledon and Colombo, were converted in 1943. HMS Caledon is shown on 1 February 1944. Improvements included more close-in (self-defence) weapons, on the theory that the ships would be valuable targets in their own right. Instead of pompoms, they had two twin Hazemeyer Bofors guns, visible amidships. Other close-in weapons were three single Bofors (not in Colombo), six twin power Oerlikons and two single Oerlikons.
The Board understood the importance of anti-aircraft modernisation, and planned to upgrade ships when they came in for large repairs. Thus it began to consider AA modernisation of the Kent class heavy cruisers, in 1932, and definite plans were completed in September 1934.2 At this stage the HA improvement was to have been limited to replacing the four single 4in guns with two twin and two single 4in, and there was no proposal to install Mk M pompoms. In May 1935 it was decided to replace the one twin 4in mount on each side in Penelope and Aurora (Arethusa class) with two, and to add a second HACS, which brought the ships into line with the Kents (but the 1936 programme did not include a second HACS for anything but heavy cruisers). Again, there was no mention of Mk M pompoms. At this stage the Naval Deficiency Programme included capital ship but not cruiser modernisation, and it did not include new anti-aircraft weapons for the cruisers.
Emerald and Enterprise were included in the list of prospective improvements, but nothing could be done before the outbreak of war, and then the planned 1940 refits were abandoned. However, both ships were modernised in 1943.3
The crisis between Britain and Italy over the invasion of Abyssinia nearly brought the two countries to war. Italy had a large air force, and the threat of war dramatised the need for heavier anti-aircraft armament in British battleships and cruisers. Coventry and Curlew were rearmed on an emergency basis during the crisis, Mediterranean Fleet anti-aircraft armament being considered too weak. They were presumably conceived as exactly the sort of fleet anti-aircraft ships envisaged by the 1921 antiaircraft committee. The 1935 provisional armament statements for both ships described them as ‘special duties’ ships. Armament was ten single 4in guns (250 rounds each) and two octuple pompoms (later one: 1,800 rounds per barrel), and an HACS was installed, its HADT occupying the top of the foremast. Torpedo tubes were landed. Both ships had two single 4in removed in 1939–40; Coventry later landed a third gun.
Emerald as refitted in 1943, from Portsmouth plans dated 24 March 1943. The dashed box at the foot of the foremast, below decks, was the Second (and Auxiliary) Wireless Office; the main office was the dashed box under the mainmast. The yard shown on the foremast carried the small crosses of Type 86 tactical radio at its ends (one is indicated). The multiple strands of the ship’s ‘flat-top’ antenna are indicated, the line leading down above the forefunnel being the receiving antenna (one on each side). Below it, the ship’s foretop carried a rangefinder. The ‘lantern’ atop the bridge carried the ship’s Type 273 surface-search radar. The masthead carried the receiving antenna of her Type 281 air-warning radar, with the transmitter on the mainmast. Above the forward Type 281 antenna was the ‘hayrake’ of the associated Type 243 interrogator (for IFF). The small rectangles on the main starfish are the FV1 countermeasures antennas. The plan associates the two ‘hourglass’ omni antennas below with the ship’s Type 91 jammer, one for receiving and the one below for transmitting (but the contemporary FV1/Type 91 manual claims that the FV1 antennas were generally used to transmit, which would make sense because they were somewhat directional; the omni may have been an alternative). The only wireless antenna shown aft (of several) is the emergency receiver shown against the after funnel. The quad pompom on this side is not shown, but note its platform, with the pompom director (with Type 282 radar, and Type 282 radar office below it) abaft the second funnel. Abaft it is the after control structure, with its HA director topped by a Type 285 radar antenna (with Type 285 office below). The light gun under ‘B’ mount is an Oerlikon, the only one included in this elevation drawing. Boats shown are a 27ft whaleboat abeam the forefunnel and, aft, a 30ft motor boat forward of a 25ft motor boat. On the opposite side of these boats was a single 36-footer. The object on deck forward of the after funnel is the main (wireless) aerial trunk. Not visible in this drawing, but evident in the original, were 25ft booms on each side under the slight overhang aft, and two booms rigged from each side forward, from the superstructure just forward of where the bridge met the shelter deck level: two sounding booms abaft two lower booms. Note the absence of a main battery fire-control radar (Type 284). (Norman Friedman)
In the wake of the crisis, a two-part fleet anti-aircraft upgrade was proposed in 1936. Class I modernised the main fleet, while Class II provided trade-protection anti-aircraft ships.4 The Treasury approved Class I spending in April 1936, and an AA Rearmament Committee was formed. Plans called for adding four 4in guns (making a total of eight) to all 8in cruisers, as well as a second HACS, and bringing short-range firepower up to two octuple pompoms with directors and two quadruple 0.5in. Ships would receive twin 4in guns at large repair, eight single guns being an acceptable interim battery, the guns becoming available as battleships and 6in cruisers were rearmed with twins. The Kents were already being rebuilt to this standard.
In 6in cruisers (Leanders, Amphions [including Sydney], and Arethusa and Galatea) single 4in guns would be replaced by twins, but there was no mention of pompoms. Ships would not receive a second HACS. Achilles was not refitted because she was in New Zealand waters (the guns were replaced only in 1943–4), and HMAS Sydney (ex-Phaeton) was not refitted because she was already in Australian waters (she was sunk carrying her single 4in guns). Arethusa and Galatea only had their upgrades completed in December 1940 and September 1941.5
Including capital ships and carriers, the programme required fifty twin 4in guns (Mk XIX twins), twenty-five sets of HACS, twenty-two octuple 2pdr with their directors and ammunition, and ten quadruple 0.5in guns. Equipment on hand would be used in 1936–7; deliveries of new twin 4in guns (two per month, rising to four in July 1937) could begin in February 1937. HACS deliveries (one or two per month) could begin in June 1937. The only really new item was the proposal to fit octuple rather than quadruple pompoms in the 8in cruisers. The programme would complete AA upgrades in two to three years; otherwise the programme of large repairs would not have been completed until 1942 or later. Eight cruisers would be rearmed in 1937 and Apollo (HMAS Hobart) in 1938.
Class II included rearming cruisers, ‘V’ and ‘W’ class destroyers, and sloops as convoy AA escorts. Initial funds bought weapons for six ‘C’ class cruisers and HMS Whitley. The programme responded to a shift towards the possibility of a European war during which British shipping would be threatened by shore-based aircraft (that was far less likely in a Far Eastern war). Escort rearmament was initially given priority over smaller fleet cruisers for which definite dates had not yet been settled: the rest of the Leander and Improved Leander classes, Arethusa, Galatea, York and Exeter, and replacement of singles by twins in Cumberland and Suffolk.
In December 1936 the British government assured the Japanese and US governments, the other signatories of the 1930 London Treaty, that five of the ‘C’ class cruisers it was retaining would be converted into anti-aircraft ships, and that they would be scrapped by 1941. As of March 1937 plans called for taking four cruisers in hand during 1938 for completion by that December. The programme applied to six ‘C’ class cruisers and eight ‘D’ class cruisers, in addition to the two earlier conversions. Selected merchant ships would receive the same set of weapons and fire controls as a converted ‘C’ class cruiser.6
Enterprise as refitted 1943, based on a 17 February 1942 drawing of her planned rig, done by Devonport Dockyard. The dashed lines below her bridge structure are cable passages leading down to her Second Wireless Office, below decks, the trunk for its main aerial being visible as a vertical pipe forward of her bridge structure. Both of the more or less vertical lines near the bridge indicated antennas for the Second Wireless Office, terminating on the 32ft signal and wireless yard shown. Below it were two shorter yards (not shown) with fixed manoeuvring lights at their ends. The trunk higher up in her superstructure, nearer her mast, served a secondary Type 30 wireless. The ship’s LF and MF wireless sets used ‘flat top’ antennas, for which wires were rigged between the two masts, with vertical elements connected to them. These wires are shown in part as dashed lines. They served a Type 36S wireless. The two slanting dashed lines abreast the fore funnel were receiving aerials. The raked dashed line is simply the foot of the fore leg of her tripod foremast. Topmasts were the receiver (foremast) and the transmitter (mainmast) of her Type 281 air-warning radar. Not shown is an additional line leading up to an insulator on the fore side of the forefunnel. It is described as part of the Type 281 receiving antenna. The large dashed space below decks aft is the main wireless office. The slanting vertical dashed line leading up to the mast is one of two receiving antennas for this office. The corresponding transmitter emerged from a trunk at the fore side of the after funnel. An outrigger on the side of the after funnel (not shown) carried the upper end of a short-range wireless antenna, which was a nearly vertical wire. The object which appears to be carried by stanchions is the Type 281 transmitting office, serving the transmitter atop the mast. The Type 281 receiving office occupied the after extension of the shelter deck, atop stanchions. This mast carried a 24ft yard. The box abaft the second funnel was the Type 285 office. The Type 284 office occupied the after part of the second bridge level, the one at whose forward end an MF DF coil was bracketed. In the plan view, the positions just abaft the open (12ft) rangefinders in the bridge structure are for pompom directors (the pompom positions were not shown). The two long rectangles flanking the forefunnel were boiler room vents, as were the two rectangular objects athwartships between the first and second funnels. Two more boiler room vents flank the after funnel. Note the two searchlights sided just forward of the HA director on the control structure between the second funnel and the Type 281 transmitting office. No others are shown. The object atop the Type 281 transmitter office is an insulator group. The object below the fore starfish is an electric steaming light above an oil-burning steaming light. Boats (omitted from the plan view) were a 27ft whaler on each side near the break of the forecastle. Abaft this boat on the each side was a 32ft cutter. The boat abeam the after funnel was a 30ft motor boat. (Norman Friedman)
Basic requirements for the ‘C’ class conversion were laid down in around April 1937. The conversion design was adapted to the later ‘C’ class cruisers with superfiring guns in ‘B’ position; converting any of the earlier ships would require a separate design, and hence take too long. Ten such ships had been built, of which Coventry and Curlew had already been converted into anti-aircraft cruisers. The five ships initially earmarked were Cardiff, Ceres, Caledon, Calypso and Caradoc, but Caledon and Caradoc were removed from the list because they were of the earlier non-superfiring design. By July 1937 plans called for taking five ships in hand during 1938: Calcutta, Cardiff, Cairo, Ceres and Colombo. The remaining six ‘C’ class cruisers would follow during 1939: Capetown, Carlisle, Curacoa, Caledon and Calypso (in December 1937 three Caledons [non-superfiring] were omitted from the programme). Given delays in the programme, in June 1938 the Admiralty proposed that instead of modifying five ships soon to be scrapped, they or an equal number of ‘C’ class cruisers of equal or greater tonnage should be scrapped before 1941, and an equal number of ships with longer remaining lives (to 1943 or 1944) substituted as anti-aircraft cruisers. The five named ships would be refitted for continued service as 6in cruisers. Other candidates for scrapping by 1941 might be a ‘D’ class cruiser. The other warships envisaged for AA conversion at this time were thirty-six ‘V’ and ‘W’ class destroyers, two Stork class, six Grimsby class and sixteen Milford class sloops, and one other escort. The list was considered provisional, e.g. more destroyers could be converted at the expense of ‘C’ class cruisers.
Sometime in 1937 Class I received first priority, Class II cruiser and destroyer rearmament moving to the end of the programme, beginning in 1939 (although initially a 1938 start was planned). Main fleet ships (capital ships, carriers and cruisers) were rearmed first, followed by modern escorts (sloops and minesweepers), and then by the AA escort programme. A parallel programme provided existing destroyers with modern anti-submarine armament (Asdic and increased depth charge batteries).7
As of February 1939, two cruisers (Cairo and Calcutta) were in hand at Portsmouth and Chatham respectively. They were completed on 3 June and 6 March 1939 respectively. Calcutta was followed by Curacoa (to be taken in hand June 1939, completed 24 January 1940). The first ship converted at the third yard, Devonport, was Carlisle (taken in hand April 1939, completed 20 January 1940), to be followed by Colombo (to start September 1939 for completion by the end of January 1940) and Capetown (to be taken in hand in November 1939). The outbreak of war precluded conversion of Capetown and Colombo, so only four ships were completed.
Because she was so much larger than the ‘C’ and ‘D’ class, Emerald (and her sister, Enterprise) was much more extensively modified during the Second World War. She was refitted at Portsmouth between August 1942 and 6 April 1943. These photographs were taken on 14 March 1943. Her bridge was rebuilt. The old pair of single pompoms and the pair of quadruple 0.5in machine guns were landed. Instead she was given two quadruple pompoms and six twin 20mm guns. Two of the four quadruple torpedo tubes were landed. She retained her three single 4in anti-aircraft guns. Radars installed at this time were Types 281, 273, 285 and 282 (the latter for the pompom directors). The catapult was removed during her 31 March – 5 April 1944 Rosyth refit.
Coventry as converted into an anti-aircraft cruiser in 1936. (John R Dominy)
Coventry as an anti-aircraft cruiser in 1940. She shows topmasts and a tall mainmast for her Type 279 radar, plus splinter bulkheads. (Henry R Dominy)
Initial plans called for all existing armament to be replaced by four twin 4in HA guns, two quadruple pompoms, and two quadruple 0.5in machine guns.8 The HADT was bracketed on the tripod foremast, on the ADO platform. DNO preferred a fifth twin 4in gun to the pompoms, self-defence weapons being of little or no value for area convoy defence. The two 0.5in machine guns would offer some defence against point-blank attacks ‘which an AA cruiser, if properly fought, was more likely to encounter than close range air attack’, i.e. dive bombing. If required, a second HADT could be mounted on the boiler casing aft, as there was space aft for a second HA computer. However, the second HADT could not be delivered until 1942–3, by which time (as seen in 1937) many of the ships involved would already have been scrapped. It would not be capable of LA fire control. It turned out that weight precluded substitution of an octuple for a quadruple pompom. Meanwhile DNC was producing a sketch design. It turned out that topweight precluded mounting a twin 4in gun in B’ position, so the sketch showed a pompom there, with a twin 4in abaft the funnels and two in ‘X’ and ‘Y’ positions. Formal Staff Requirements were developed at a meeting on 26 August. To the basic role of helping to protect a convoy against persistent attacks by land-based aircraft, it added as secondary roles the protection of advanced naval bases and, ‘in emergency, supplementary protection to United Kingdom seaports’.
The Staff did want the second HACS, because it wanted to be able to split the guns into two groups of four guns each. For self-defence the ship should have one quadruple pompom, plus the usual pair of 0.5in machine guns. Space would be left for a second HACS, but no wiring would be run. However, the second HACS Mk III featured in a summary of the ships’ fire-control systems produced in August 1940. The fire-control computer space forward was a combined transmitting station (for surface fire) and HA computing position; the analogous space aft contained only an HA computer. A provisional armament statement (1940) for the first four ships (Cairo, Calcutta, Carlisle and Colombo) showed eight twin 4in (270 rounds per gun), one quadruple pompom (1,800 rounds per barrel), and two quadruple 0.5in machine guns, plus six Lewis guns. In common with other British cruisers, they carried depth charges: two chutes with six First World War Type D depth charges.
Interest in conversions revived in 1942, a Staff Requirement being issued that April. Now it was recognised that the ships would be important targets for enemy aircraft, hence that they needed more close-range anti-aircraft weapons, which could be fitted only at the expense of one of the four twin 4in. Plans thus called for three twin 4in (‘A’, ‘B’ and ‘Y’ positions, all with RPC), two twin (Hazemeyer) Bofors in Q’ position and two more on each side in ‘X’ position; and eight Oerlikons (four on the bridge structure, with good forward arcs, two abreast the forefunnel and two aft). They would have one HACS director above an ADP, directly abaft the compass platform, the corresponding computer (table), and a barrage director aft. They would be fitted for fighter control, with the necessary automatic plot. Colombo and Caledon were converted to meet this requirement. A provisional armament statement showed four twin and two single Oerlikons, with the note that twin Oerlikons could replace twin Bofors if the latter were not yet available. Neither ship was complete until early in 1944.
Above and below: HMS Dauntless shows typical mid-war modifications in these 1 May 1942 photographs. At this time she had three 4in anti-aircraft guns, two single pompoms (the weapons at shelter deck level abeam the bridge), and eight Oerlikons. She had a Type 290 air-search radar on her foremast and the ‘lantern’ of a Type 271 aft. The radars and the guns were probably fitted during a Portsmouth refit (23 February – 8 March 1942). Durban showed similar features in September 1942. Some ships were never really modified: as late as April 1943 Diomede and Durban had the pre-war battery of two single pompoms and two quadruple 0.5in machine guns. They also had eight single Oerlikons. Despatch lacked the machine guns, and Durban had no 2pdrs. Diomede was converted into a training ship at Rosyth between July 1942 and September 1943.
HMS Danae received an elaborate ‘D’ class upgrade, as shown in these August 1943 photographs. She was the prototype for a projected modernisation of the ‘C’ and ‘D’ classes. Her No. 3 6in gun was replaced by a pair of quadruple pompoms, each with its radar-equipped director (using Type 282 radar). All the single 4in guns were landed, but a twin 4in was mounted to superfire above the aftermost 6in gun. No separate 4in director was fitted. All torpedo tubes were landed. The ship was unusual in that she had the FV1/Type 91 intercept/jamming system on her foremast, the FV1 antennas being distributed around the spotting top. Danae was given to the Polish Navy in September 1944 as ORP Conrad, replacing ORP Dragon, which had been damaged by a Marder human torpedo at Normandy and then expended as a blockship. Dragon had been loaned to the Polish Navy in 1943. Modernisation of Danae was so extensive (and expensive) that it led to reconsideration of the future of the ‘C’ and ‘D’ classes. In mid-1943 plans called for modernising the next three cruisers, Dauntless, Capetown and Ceres, beginning in June, August and September 1943. Not modernising them would release valuable shipyard labour and would also release ships for subsidiary tasks such as aircraft targets and training. On 27 April 1943 DOD(H) (Director of Home Dockyards) therefore proposed cancelling the planned modernisations. For his part, C-in-C Eastern Fleet valued the old cruisers as convoy escorts for the Indian Ocean. However, it appeared that he actually preferred armed merchant cruisers for this work. The situation was further complicated because existing armed merchant cruisers were being converted to troop transports, depot ships and the like. An additional factor was the appearance of modern French cruisers released when deadlocks over the French fleet in Alexandria were resolved. Caradoc had already been earmarked by Home Fleet as a gunnery training ship because her full speed was too low for fleet operations. On 27 May 1943 Deputy Controller circulated a decision: ships would be refitted but not modernised. Alterations to gun armament would be limited to substitution of twin Oerlikons for singles and also for single pompoms. ‘D’ class cruisers would have their after (centreline) 4in guns removed and as many Oerlikons as possible substituted. The two sided 3in in the ‘C’ class and 4in in the ‘D’ class would be retained to fire star shells. Dragon, which was a Polish ship, was modernised similarly to Danae during a Chatham refit (April – 16 May 1944). She had the same twin 4in gun aft, the two quadruple pompoms replacing No. 3 6in gun, the torpedo tubes removed, and eight twin power Oerlikons. Prior to this refit she had no pompoms at all, but No. 5 gun had already been landed, and in April 1944 she was credited with five twin power Oerlikons and four single Oerlikons. She had been turned over to the Poles the previous year after a refit at Cammell Laird. Initially the intention was to use her in the Eastern Fleet, which would have justified a limited antiaircraft battery. When they found that she would serve instead in home waters, the Poles asked for the increased anti-aircraft battery. Initially they were offered a single quadruple pompom, but the Polish naval commander pressed for two. First Sea Lord personally ordered two pompoms to be provided even though they were in short supply (they were earmarked for new destroyers), in view of the valuable service the Poles were providing. Durban also apparently lost a 6in gun (but that is not reflected in the April 1944 armament list), and by early 1944 only Despatch and Durban retained torpedo tubes. Dauntless was refitted at Chatham for training. It seems unlikely that she ever received any additional close-range armament. In April 1943 she and Dragon were both credited with two quadruple pompoms (she was also credited with eight Oerlikons), but that probably reflected plans rather than reality, particularly since neither ship was listed as having surrendered a 6in gun to provide the necessary space and weight. Despatch lost all of her 6in guns (she retained one 4in anti-aircraft gun) and gained sixteen single Bofors in an October 1943 – 6 May 1944 Portsmouth refit. This was apparently to prepare her for service as a gunnery school firing ship, possibly in combination with Durban (as was being considered by C-in-C Portsmouth in February 1944). However, it was soon evident that her hull was in too poor a condition for such employment, and in March 1944 C-in-C Portsmouth asked that she and Durban be made available as accommodation ships. She was recommissioned on 5 April 1944, and her radars were made operational. Off Normandy she was used as a headquarters for the northern Mulberry harbour, berthed in Gooseberry III, and then as a shipping control vessel until relieved by HMS Waveney. It was pointed out in August that full value was not being gained from the heavy close-range armament because it lacked fire controls. No such equipment could be made available, and by late August 1944 the ship had been relieved from her ‘Neptune’ (Normandy invasion) duties for reduction to reserve. Durban was expended as part of the Arromanches breakwater.
Top and above: Delhi rearmed, as in April 1942 after alterations (such as fitting radar) were completed in the United Kingdom. The US Navy supported this rearmament because it wanted to see how the Mk 37/5in combination performed in combat, although, ironically, the US was itself at war by the time the ship was ready. (A D Baker III)
Delhi rearmed (inboard profile). (A D Baker III)
1. Officers’ heads (bathrooms)
2. Passageway with officers’ cabins port and starboard (P&S)
3. Steering machinery room
4. Fuel oil tankage
5. Wardroom stores (P); paymaster’s and marines’ stores (S)
6. Gun crew shelter
7. Centreline passageway with 5in ammunition passing trays; wardroom to port, officers’ cabins and offices to starboard
8. Provision room (P); captain’s stores (S)
9. Warrant officers’ stores
10. Fuel oil tankage
11. Captain’s day (in port) and dining cabin
12. Passageway with officers’ cabins (P) and warrant officers’ mess and pantry, captains’ pay office and engineer’s office (S)
13. Trunk
14. 5in projectile stowage and handling room
15. 20mm ammunition stowage
16. 5in powder magazine (centreline); fuel oil tankage (P&S)
17. Main radio office (S); gunnery office and mechanics and chief stokers’ messes (P)
18. No. 2 steam turbine engine room
19. Centreline passageway with engineer’s workshop (P) and miscellaneous workshop (S)
20. Vegetable locker
21. No. 1 steam turbine engine room
22. Passageway (P&S); 2pdr pompom equipment (P): magazine cooling and ice-making machinery (S)
23. 2pdr pompom magazine
24. Fuel oil tankage (P&S); 5in projectile magazine and handling room (centreline); 5in powder room (P)
25. After funnel trunk
26. No. 2 boiler room (passageways P&S at 1 platform level)
27. Lower portion of No. 2 boiler room
28. 2pdr pompom director (P&S)
29. Forward funnel trunk
30. Boiler room vent space
31. No. 1 boiler room (passageways P&S at 1 platform level)
32. Lower portion of No. 1 boiler room
33. Type 281 radar receiving office
34. Crew’s galley
35. Degaussing machinery room centreline, flanked by coal and wood stowage
36. Lower gunnery plotting room
37. Fuel oil tankage
38. Fuel oil tankage
39. Fuel oil tankage (pitometer trunk at centreline on after bulkhead)
40. Passageway and crew’s recreation space
41. Passageway centreline; chief petty officers’ mess (P); sickbay (S)
42. Passageway centreline; stoker petty officers’ mess and dry canteen (P); petty officers’ mess and issue room (S)
43. Lower conning tower (P); canteen stores (S); flanked by gunner’s stores (P) and central storeroom (S)
44. Fuel oil tankage
45. 5in ammunition hoist
46. Fuel oil tankage (P&S) flanking 5in projectile stowage and handling room
47. Crew berthing (with ammunition hoist space centreline at forward end
48. Crew berthing with artificers’ mess to starboard
49. Cold and cool provisions storerooms centreline; refrigeration machinery room (P); dry stores room (S)
50. 5in powder room (flanked by fuel oil tankage)
51. 5in powder room (flanked by fuel oil tankage)
52. Crew berthing space (mess)
53. Central stores, with 5in ammunition hoist centreline
54. 5in projectile stowage and handling room centreline, flanked by fuel oil tankage
55. Fuel oil tankage
56. Crew berthing space (mess)
57. Crew berthing space (mess)
58. Flour and biscuit stores
59. Stores
60. Fresh water tanks (P&S)
61. Canvas stowage
62. Inflammable stores
63. Chain locker
64. No. 1 storeroom
65. Upper bridge with lookout positions (P&S)
66. Remote 5in gunnery control room
67. Plotting office
68. Chartroom
69. Wheelhouse (pilothouse)
70. Radio direction finder (RDF) office (P); gun crew shelter (S)
71. Radio signal (message) distribution office (P); head (S)
72. Captain’s (sea) cabin (P); navigator’s cabin (S)
73. Vent shield
74. Passageway, with mail office (P)
75. No. 1 storeroom
76. Lower chain locker
77. Paint room (boatswain’s stores)
78. Watertight compartment
79. Watertight compartment
Plans called for refitting the eight ‘D’ class cruisers with four twin 4.5in mounts, one quadruple pompom and two quadruple 0.5in. Orders for the 4.5in mounts were let in 1938 to keep gun and mounting production running at a steady pace, even before the full Class II programme had been approved, since they could also be used in new ships and in the Class I programme. Large repairs and conversion of a ‘D’ class cruiser were expected to take twelve months. Three ‘D’ class would be taken in hand on 1 January 1940 for completion by 1 January 1941, one was to complete by 1 May 1941, and the others were to be rearmed beginning in 1941. By mid-1939, however, the British war plan included cancelling the ‘D’ class conversions. Their 4.5in guns became available for other purposes, including arming the Dido-class cruisers Scylla and Charybdis.
In the autumn of 1940 the Admiralty became interested in buying US 5in/38 guns to arm new destroyers, and there was naturally parallel interest in using these guns for a new anti-aircraft cruiser conversion, which, given the size of the weapons, had to be one of the ‘D’ class. A Staff Requirement envisaged two directors and either four twin or six single mounts. On 9 November DNC wrote to Controller that he had little information on the weapons, but that it might be necessary to accept fewer mounts; by mid-November he had settled on five singles and two fire-control systems. The next day the British mission in Ottawa provided full details. By late December Controller wanted a feasibility study (arrangements to buy the guns and fire controls were still being settled), and in January DNC calculated that the ship could carry five single 5in/38 and two Mk 37 fire-control systems, which at 113 tons would weigh slightly less than the four twin 4.5in and two HACS (130 tons) planned for the ‘D’ class conversion. Another alternative for a ‘D’ class cruiser was apparently five twin 4in with two HACS (82 tons). A design conference met in Bath on 8 January; this was an urgent project. The five guns would occupy positions formerly occupied by 6in guns, so that they could use the same magazines. On 10 January DNC drafted a letter to the senior British naval officer (Admiral Evans) in Ottawa, explaining that the British yards were too overloaded for the job, so ‘the best, in fact the only, way’ to get the conversion done quickly would be send the ship to North America. Ideally the whole job would be done in an American yard, but ‘apart from the political aspect’ American yards were undoubtedly full, ‘and they might not like to convert a British ship that is to be returned to the British service’. The next best, and likeliest, possibility was conversion in a Canadian yard under American supervision. Conversion was on standard British antiaircraft cruiser lines, with two pompoms (with directors) and four quadruple 0.5in machine guns. Because the US guns used AC electric power, whereas the British used DC (and had limited generating capacity), the ship had to be, in effect, rewired.9
The passage of the Lend-Lease Act (11 March 1941) transformed the situation, since the ship could be converted in a US yard using US equipment. HMS Delhi was chosen (she had been about to join the East Indies squadron at Singapore). Like the smaller anti-aircraft cruisers, she became a two-target ship (a US destroyer armed with the same five guns had only a single fire-control system). The two Mk 1 computers of the Mk 37 systems were installed in a single plotting room. British radars were of course installed: Type 285 for the 5in directors, Type 282 for the pompom directors, and one Type 281 for air warning. A D/F office and coil were installed between the funnels. Like other modern British cruisers, Delhi was fitted with Asdic on conversion (in her case, Type 128A). While Delhi was being rearmed, plans were made to provide her with Oerlikon guns.
HMS Delhi was rearmed at Norfolk Navy Yard as a possible prototype for further ‘D’ class conversions, but with the entry of the United States into the war so elaborate a conversion was no longer practicable. Delhi had the same main battery as many US destroyers, but she had two (rather than one) Mk 37 fire-control systems, hence could engage two targets at the same time. Her British radars were installed only after she returned to the UK. She is shown in April 1943.
In May 1941 word was received that the US authorities might be willing to rearm a second ‘D’ class cruiser along the same lines. Although he had only observers’ reports of the efficiency of the US long-range anti-aircraft guns, DTSD urged further conversions, asking Operations Division to release either a second ‘D’ or even an unconverted ‘C’ class cruiser. A sketch design was drawn up for ‘C’ class conversion.10 If that proved impossible, he suggested asking instead for a merchant ship conversion similar to the merchant ships already converted into anti-aircraft escorts. Director of Plans welcomed the idea ‘since it seems almost axiomatic that a 5in HA/LA equipment already in supply to the US Navy and with tachymetric control must be more effective than the British 4in HA/LA controlled by HACS. In addition the heavier calibre renders it just possible to retain the re-armed ship within the cruiser category instead of declassing her to AA duties only.’ Availability depended entirely on the refit programme. All the ‘D’ class cruisers other than Dauntless (which completed a large repair at Portsmouth in July 1939) were due for large repairs when war broke out, but no such refits had been scheduled, because ships were so badly needed in service. Dragon was to follow Delhi for a US refit, but not for rearmament (2–3 months rather than 5–6 months’ work). Despite boiler defects which reduced her speed, Dunedin was being kept in service. Due to her defects, she was the best candidate for conversion, but cruiser strength could not be reduced before the autumn of 1941, when new ships would enter service. She and Dragon were currently on the South Atlantic station.
The position of the ‘C’ class was similar. Ships not scheduled for AA conversion were to have been refitted to extend their lives to December 1941. As Gunnery Firing Cruiser, Cardiff was not available, while Capetown (which had been earmarked for AA conversion) was being repaired at Bombay following torpedo damage. Caledon and Ceres were to be refitted at Colombo. Caradoc was to have her turbines rebladed.
In July Dunedin was approved as the second ship, but only subject to successful trials of Delhi, experience with US-made 14in guns in the First World War having been less than happy. By the time Delhi had been completed, the United States was at war, and the opportunity for a second conversion was gone.
The Fiji class
With the 1936 London Naval Treaty the British finally stopped construction of big cruisers. What could DNC design within the new 8,000-ton limit? Apparently the first attempt was to replace the four triple turrets with three quadruple ones; it came out to 8,500 tons (this design may have been K20).11 A successor K21 (Design P) was developed on 23 January 1936 to answer DNC’s question about a ship with ten guns (twins in ‘A’ and ‘Y’ positions) with anti-aircraft armament, speed and protection as in the 1935 Southamptons (the Gloucester class).12 An initial estimate showed that 15ft could be saved on length (total 569ft), and that standard displacement would be about 8,900 tons. That assumed that the turrets were the short-trunk type as in Southampton and Leander. K22 (Design Q) was a nine-gun ship with three triple turrets. In that case 36ft would be saved, and displacement might be about 8,500 tons. K23 (Design R) showed how much protection Design Q could retain if she were limited to 8,000 tons. All three designs retained Southampton-class machinery producing 82,500shp. For Design R length was based on that of the three-turret ship, beam and draft being adjusted to give the desired displacement. Subtracting the estimated hull, machinery, armament and equipment weights (adjusted for the saving in personnel) left about 1,210 tons for protection. The protection provided for HMS Liverpool, a Gloucester-class cruiser, appropriately scaled down for size and number of guns, required 1,415 tons. Reverting to the reduced protection of the Southampton class would save 155 tons, leaving another 55 tons above the desired tonnage. Taking ½in off the belt would save another 75 tons. Alternatively, the belt could be cut to 2½in rather than 4in and turret protection be reduced (from that of Liverpool) to a 2in roof, 1½in sides and a 3½in face.
Below and above: The Fiji class was an attempt to package the armament of a ‘Town’ class cruiser in a hull about a thousand tons smaller. HMS Kenya, newly completed, is shown in 1941. Kenya had the usual air-warning and gunnery radars, but not a surface-search set. The view from aft shows one of the methods used to compress the ships. Instead of an after DCT, they were given a fixed rangefinder and director atop ‘X’ turret. In the stern quarter view, note the surviving quadruple 0.5in machine guns and also the Oerlikon mounted right aft on the quarterdeck.
Now that DNC knew what he could have on 8,000 tons, the question was whether he could get more by making further reductions. Design S (K21 modified) was the ten-gun ship with reduced protection. Length could be reduced to 566ft. She could make 32.5kts on 77,000shp (330rpm machinery). However, at 300rpm it would take only 74,000shp to drive her at the desired speed, so it might be possible to use Amphion class machinery (slightly pressed). Machinery weight would be reduced from 1,545 tons to 1,475 tons, saving 70 tons. The ship would have to sacrifice two of the three aircraft of the Southampton class, reverting to Leander-class arrangements. She would have a 3½in belt, but otherwise would be protected like a Southampton (armour weight would be 1,083 tons, compared to 1,529 tons in Liverpool). Speed would be the 32.5kts of a Southampton.
All of this was encouraging, so on 3 February DNC asked what displacement the 8,500-ton ship (with protection like Liverpool) with three quadruple mounts would have if speed were cut back to 32 or 31.5kts. At 32kts the ship would probably displace 8,330 tons (77,000shp); she would need 71,000shp for 31.5kts (8,170 tons) and 66,000shp for 31kts (8,000 tons). This was presumably design K20A. Controller then returned to versions of the three-turret K23. One possibility (K23A) was to cut back side protection to that of the Leander (3½in rather than 4½in), but to leave 4½in magazine sides. Transmitting and other control positions had the belt protection instead of the 1in in a Leander. The alternative (K23B) was to trade off speed and length (530ft rather than 550ft, both on the waterline) rather than protection for the three-turret main battery (30kts). Both designs would clearly be badly cramped. K26 was a further development of K23B in which the catapult and hangars were eliminated, to open some midships space (a light reconnaissance aircraft could be stowed on deck and lowered over the side to be launched in calm weather, but its weight was not included in the ship’s standard displacement). The space opened up could accommodate two more twin 4in guns (total of six, as in Belfast), and it also allowed for enough additional power to boost the ship to 30.7kts. To improve 4in ammunition supply, the magazine was placed amidships, between the forward engine room and the after boiler room (no such magazine could be fitted into K23A or K23B). Trading the weights of the two extra twin 4in for machinery would bring speed up to 31kts, making K26 fairly attractive. The three designs with three turrets were reported on 9 July.
Controller saw the 5.25in gun as an alternative, so on 11 February he asked how many he could get into an 8,000-ton ship with the same other requirements (speed, protection, aircraft) as a Southampton.13 A first cut was discouraging: if the Southampton hull were simply scaled down to the smaller displacement (562ft long) and protection and other weights were left virtually unchanged, very little weight would be available. This was probably K24, and it was so discouraging that it was not reported. Controller went back to DNC to ask what could be gained by reducing protection. This was the K25 series. DNC found that with machinery for 32.5kts and with minimum protection (3in belt and 1¼in deck over machinery, 3½in side and 2in deck over magazines, and 1in side and deck over shell rooms), there was enough weight for seven twin 5.25in with 300 rounds per gun (but space for 400). Four turrets were superimposed forward (but with two turrets side by side on ‘B’ level, so that there were three levels forward) and three aft. Shell rooms had to be placed above magazines, because there was not enough space for them to be side by side, as in a Dido. Unlike a Dido, these ships had protected turrets, with 3in faces and 1in sides and roofs. K25A was the basic version. In K25B through K25F speed was reduced to 32kts and the weight saved used in various ways: to substitute a complete 3in belt for the belt/box arrangement (K25B); to increase deck protection (in K25A) to 2in over the machinery (K25C); to increase machinery side protection to 3 ½in and give control spaces 1in box protection (K25D). In K25E, speed was cut to 31kts and a complete 3½in belt and uniform 2in deck provided over both machinery and magazines, 3½in thick. K25F added an eighth twin 5.25in turret to K25A at the cost of reducing speed to 31.7kts. This design showed four turrets superimposed both forward and aft. K25H showed that the cost of providing a 4½in belt over machinery and magazines (and 4in turret faces and 2in roofs and sides) was a reduction to 29kts, which was unacceptable. These alternatives were reported on 25 July 1936.
The K25 design for an 8,000-ton cruiser armed with 5.25in guns. (Norman Friedman)
The K25A-F design for an 8,000-ton cruiser armed with 5.25in guns. All had the same configuration, but they differed in speed and protection (the highest was K25A, 32.5kts). Protection in a typical version, K25E (31kts), amounted to a 3½in belt over machinery and magazines and shell rooms with a 2in deck atop it. The 5.25in turrets had 3in faces and 1in sides and crowns. Handling spaces under the turrets had 1in protection. Controller, Rear Admiral Reginald H H Henderson, much favoured the 5.25in cruiser, but the Board rejected his views. Henderson was responsible for several innovative ships, including the ‘Tribals,’ the Didos, the armoured carriers, and HMS Unicorn, as well as for the King George V class. (Norman Friedman)
The K26 design for an 8,000-ton cruiser armed with three triple 6in turrets and six twin 4in, as in Belfast. Speed 30.7kts, length 530ft. There was no catapult, but there was deck space for a single long-range seaplane. Machinery protection would have been 4½in side and 2in deck, with similar magazine and shell room protection (belt and deck rather than box). K23A was similar but had four rather than six 4in mountings. She would have been 550ft long and would have made 32.5kts. Machinery would have been protected by a 3in belt and a 1¼in deck; magazines and shell rooms would have had box protection (4½in side, 2in deck). (Norman Friedman)
HMS Mauritius is shown during the September 1943 Italian landings, about to shell German batteries near Naples. She displays the standard initial surface-search radar arrangement in this class, a Type 273 forward of the bridge – an unpopular arrangement, because it blocked the view from the bridge and the helmsman’s position.
On 16 July DNC offered two more alternatives, K27 and K28, which reverted to 6in batteries. K27 (545ft) had three quadruple turrets and the lighter protection (3½in side, 2in deck); speed would probably be limited to 30kts (figures were preliminary because the size and weight of the quadruple turret had not been settled). The ship would be very cramped, which was what probably killed the design. K28 (575ft) was another attempt at a ten-gun ship, two twins superfiring over two triples. With the same protection as K27, the hull had to be longer and heavier, but not so much longer that the ship benefitted hydrodynami-cally from the longer hull. Speed would drop to 29.5kts, an even more depressing result. Again there were uncertainties, because no long-trunk twin 6in turret had been designed. Two more 6in designs were apparently prepared in July; they must have been K29 and K30 but details have not survived. The Sea Lords chose a 6in design, probably with twelve guns, because when they approved the final design in October 1936, it was described as an elaboration of a design approved in July. It seems clear that Controller was unhappy with this choice.
In mid-September Controller wrote to ask for a Sea Lords’ conference on the 8,000-ton cruiser, as a decision was urgently needed. He offered what he considered the three best design alternatives, K27, K23A* and K25G. An entirely new quadruple turret design would impose at least a six-month delay, which he considered sufficient grounds to reject the idea. Moreover, recent tests had shown that guns had to be at least 6ft 6in apart not to interfere with each other. That sort of spacing would make a four-gun turret impractical. Spacing could be tighter if the turret used a delay coil (0.01 sec) to fire alternating pairs of guns together, but if the ship were rolling heavily, the delay would throw off her salvoes, and it would have to be cut out. The ship would become, in effect, a six-gun cruiser. Henderson considered the nine-gun ship underarmed for her size (she should have ten guns). To bring her speed up to that of the ‘Towns’, ½in had been taken off her belt. For enemy inclinations of 60° and 45°, the 3in belt would keep out shells at 10,500yds and 8,600yds respectively. The 2in deck was chosen to keep out bombs, and more than sufficed to keep out 6in shell. However, Henderson was inclined to consider the eight 4in guns of both K27 and K23A* insufficient. He much preferred the all-5.25in K25G, which was armoured against 6in shellfire. K25G showed no fewer than four superfiring mounts forward and three aft. Henderson liked the 5.25in design because it offered much more antiaircraft firepower, all of it protected against bombs (and their splinters), and because at cruising stations and in harbour half the anti-aircraft battery could be kept manned. Training would be simplified. He was also ‘inclined to think that to strike out on a “novel” design would be good generally for our prestige as leaders of design’.
On 10 October, with a decision even more urgently required, Controller (Rear Admiral Henderson) had two more designs, K31 and K25G*, prepared, both 550ft long. K31 had four triple 6in turrets, K25G* seven twin 5.25in.14 Both had 3/2in side and 2in deck over both machinery and magazines. Unlike a Dido, K25G* had serious turret protection, a 3in face and 1in sides and roof, compared to a 3½in face, 2in side and 1in roof in K31. There was no question that the 6in armament of K31 was superior to the 5.25in battery of K25G*, but K31 had too little HA armament (but the same four twin 4in, two quadruple 2pdr and two quadruple 0.5in as the Southamptons) and it was 0.75kt slower than K25G*. Henderson also noted that the problem of blast interference among the 4in guns had not been solved. On balance, he considered the 5.25in ship a better choice, at least for the first five 8,000-ton cruisers. However, the Board approved K31, the Minute describing it as a worked-out version of the sketch design approved the previous July.
Overhanging the choice was the fact that the Japanese had not signed the 1936 treaty; no one knew what sort of cruisers they were building. Henderson argued that nothing could affect the British requirement for numbers of cruisers, hence for a relatively small affordable ship. ‘When we know what the Japanese design is we can build to meet it.’ In October Vickers offered its own 8,000-tonner armed with 5.25in guns.15
K31 traded off power rather than armour for armament, with 66,000shp rather than 74,000shp. On a thousand tons less than a Southampton, it offered the same armament: four triple 6in (but in long-rather than short-trunk mountings, with the usual 200 rounds per gun), four twin 4in HA (250 rounds per gun, and 200 star shells per ship), two quadruple pompoms and two quadruple 0.5in machine guns, two triple torpedo tubes, and ten Lewis guns in the usual twin portable mounts, plus depth charges (fifteen on ships with Asdic, six on those without). K31 also offered the same aircraft arrangements, with two hangars and the athwartships catapult. That must have seemed miraculous, but it came at a high price: the design was extremely tight, just when naval technology was changing with the advent of radar and the demand (in three years) for much more anti-aircraft firepower. Obviously the ship would be much more crowded than a Southampton, so complement had to be held to a minimum. For example, the 4in magazines were placed immediately below the guns, saving the four men who, in a Southampton, passed it along the deck under the guns. War complement was cut (as of late 1936) from the 46 officers and 734 ratings of a Southampton to 48 officers (two more for additional fire control) and 662 ratings (the original pair of 4in hoists was replaced by four hoists, one per mounting, in January 1938).
Bermuda as completed, September 1942. The ship was completed with large Vokes sound-reduction filters over the fixed air intakes for the forward boiler spaces; these had vertical intake openings and a semi-circular plan form, and they were soon removed. Only one aircraft/boat crane was fitted (aircraft facilities were removed during an April to May 1944 refit). The big Type 273 ‘lantern’ forward of the bridge was replaced by Type 272 (in the foremast) about April 1943. As with all Royal Navy cruisers that served after the Second World War, Bermuda had additional portholes added to her hull sides to improve internal light and ventilation. As built, she appears to have had a ‘non-latex’ covering over the exposed decks abaft the spray deflector, but post-war at least the quarterdeck had wooden planking. An additional six single Oerlikons were added during September 1943; in April 1944 she had eight twin 20mm. During a major refit (June 1944 to April 1945) ‘X’ turret was removed and replaced by two quadruple pompoms. A third was mounted on her centreline forward of her mainmast and four single pompoms were added. During August 1945, in preparation for planned Pacific deployment, the close-range armament was further altered, with two twin and two single 20mm removed and replaced by two single 40mm Mk III (hand-worked) and two Boffins. At this time the two-antenna Type 281 was replaced by a single-antenna Type 281B to clear the foremast head for a Type 293 target-indi-cation radar antenna; the surface-search ‘lantern’ was replaced by the dish of Type 277. Post-war the 40mm and 20mm mounts were removed but the pompoms survived. In the late 1950s the pompoms were replaced by seven twin Bofors and fire controls were updated, the two forward 4in mounts being given on-mount radars of the US-supplied Mk 63 system. After 1954 the ship was given an enclosed conning position at the fore end of the bridge and Type 274 replaced Type 284 on the 6in director. She never received new lattice masts or barrage directors. After 1955 the ship carried NATO pennant number C52. (A D Baker III)
The endurance requirement (7,000nm at 16kts when six months out of dock) conflicted with the demand for at least 30kts at deep (full-oil) displacement; at the required tonnage, K31 could not make 30kts. The solution was twofold. First, E-in-C was asked to provide for a 10 per cent overload. Expected overload speeds were 31.6kts in standard condition and 30.24kts deep. W G John, the constructor responsible, suggested an alternative. The ship’s nominal deep fuel load could be limited (she could make 30kts at 9,770 tons).16 It would not be enough to make the required endurance, so additional reserve oil tanks would be provided. Alternatively, the ship could be required to make her speed after she had burned off enough oil. Four- and two-shaft machinery arrangements were offered, the former (with alternating boiler and engine rooms) being chosen.17 The 4in magazine were placed between the after boiler room and the after engine room.
The design was on the edge of being feasible within the 8,000-ton treaty limit. In November 1936 estimated standard displacement was 8,360 tons. To get down even to this figure, the ring bulkheads and turret deck rings (i.e. barbettes) had to be reduced from the 2in and 1in of earlier designs to 1in. Further weight might be saved by taking into account the shorter trunks of the 6in triples, as compared to those in Belfast. On the other hand, recent figures for the 6in triple were 20 tons greater than had been estimated, a total weight growth of 80 tons. Estimates suggested inadequate stability, and the preliminary designer suggested 6in more beam – which would add a few more tons. It seemed that standard displacement might be 8,412 tons – 5 per cent over the limit. British builders had managed to shave considerable weight from earlier cruisers, but that was a lot to expect. DNC asked that displacement be held to 8,300 tons, and that metacentric height be held to 2.5ft.18 By this time beam was 61ft 3in or, later, 60ft 6in (it had been 59ft 6in). Possible ways of holding displacement down were omission of torpedoes and shortening the 4in HA magazine. It might seem that another way to save weight was to revert to box magazine protection outside the belt, but that actually added weight (about 20 tons). The 8,000-ton limit was artificial, but the stability problem suggested that any additional topside weight would cause unacceptable loss of stability – which was the problem throughout the Second World War.
Above and below: Photographed in 1943, HMS Jamaica shows mid-war adaptation of the Fiji class. The view from the bridge was cleared when a lightweight surface-search set (Type 272) suitable for mounting on the tripod foremast, became available. The heavy director/rangefinder aft was eliminated in favour of more light antiaircraft guns (Oerlikons, in this case) atop ‘X’ turret, to help deal with aircraft attacking from aft, along the centreline of the ship. The view from aft was taken at Portsmouth, 18 September 1943; the view from ahead was taken at Scapa Flow in 1943.
Clearing the front of the bridge made space for the two barrage directors, superfiring, visible in this November 1943 photograph of HMS Gambia.
To bring down displacement enough for submission to the Board in December 1936, DNC had to reduce armour: the 3½in NC turret faces of K31 were cut to 2in NC, and the trunks were reduced to a uniform 1in instead of 2in tapering to 1in. The crowns of the magazines and shell rooms had to be cut from 3in NC (which would keep out a 500lb dive bomb) to 2in.
Policy was to provide a 10 per cent margin in boiler power to meet tropical conditions, so that the ship would develop her trial power anywhere in the world. The nominal performance of the ship could be improved by taking the 10 per cent into account in calculating trial performance, which would be valid for temperate waters.19 Thus the 66,000shp of the new cruiser was 72,500shp in temperate waters, offering 31.75kts at 8,000 tons and 31.5kts at 8,300 tons, rather than 31kts at 8,300 tons.
Meanwhile (late in November) an entirely different arrangement was suggested. Three of the four turrets could be placed forward (No. 3 slightly lower than No. 2), and the hangar placed where No. 4 turret normally was, the new No. 4 superfiring over it.20 The arrangement moved the 4in HA battery clear of 6in blast, and it was well-adapted to autogyros, helicopter-like aircraft which were considered very promising. The ship would have been 570ft long (waterline). The hull would weigh more; as in the new 8,000-ton cruiser, the machinery would be reduced (to about 50,000shp) to make up for that. However, speed in standard condition would fall to about 29.9kts, which was too low. An alternative was to revert to a ten-gun battery (two twin, two triple), saving about 137 tons on armament, which could go into more power (57,400shp, for about 30.8kts).
When the Italian heavy cruiser Gorizia docked at Gibraltar in September 1936, analysis suggested that she was about 10 per cent heavier than her stated 10,000-ton displacement (actually it was 20 per cent). In December DNC asked what 10 per cent overage would buy for the 8,000-ton cruiser: half a knot (32kts in standard condition) and an inch more belt armour; the ship would also be lengthened for higher speed.21 This idea was not followed up. However, on 10 October Controller applied the same inverted margin policy he had used for the Belfasts. As submitted to the Board, the design included a 200-ton margin above the 8,000-ton limit.
DNC submitted a sketch design dated 22 December 1936. Shell rooms were above the magazines, to save length and to provide additional protection. As in the earlier design, the two pompoms were mounted above the hangars; the two quadruple 0.5in machine guns were on either side of the forward superstructure. As in a later Southampton, there were two DCTs (fore and aft), three HADTs (one each side of the bridge, one aft), and two pompom directors (on the bridge wings). To hold down standard displacement, ammunition capacity in that condition was held to 150 rounds per gun of 6in and 4in, and 1,200 per 2pdr barrel, but there was sufficient space for the usual 200 rounds per 6in and 4in, and 1,800 per 2pdr barrel. The ship had the two triple torpedo tubes of the Southampton design, and the same aircraft facilities. However, instead of the three strike aircraft (TSR) of a Southampton or a Belfast, the Legend showed three Walrus flying-boats better adapted to trade protection. In his covering memo to the Board, DNC mentioned that alternative aircraft arrangements, including a catapult aft, had been considered. That was probably a reference to the design with three turrets forward.
The greater temperate climate performance was made explicit, as was the special deep load condition excluding 500 tons of reserve fuel oil. Thus at deep load the ship would carry 1,200 tons, but at extra deep load she would carry her full 1,700 tons, speed being 0.5kt less that the rated deep load figure. Speed in a temperate climate was 31.5kts at standard displacement and 30.5kts at a modified deep load, without fuel in reserve tanks. Corresponding figures for all climates were 31kts and 30kts.
The 3½in belt (1in thinner than that of Southampton) covered not only the machinery but also the magazines and shell-rooms, as well as the fire-control computer spaces (transmitting stations and HACPs), switchboard room, LP switch rooms (for data transmission), and wireless transmitting and receiving offices, thus covering more of the waterline than in a Southampton. The belt was closed by 2in bulkheads and a 2in deck, the latter superior to that of Southampton over machinery and control spaces but not as good over magazines and shell-rooms. Turret and steering-gear protection matched that of Southampton. A request to protect magazine sides to the level in Belfast could not, apparently, be met.
Reviewing the drawings, ACNS argued that masts and funnels should be vertical instead of raked, as in a Southampton. Controller agreed that raked masts and funnels made it easier to judge a ship’s course, but they also made it more difficult to take her range (using a British coincidence rangefinder, not the stereo type other navies were adopting). Appearance could not be ignored, however; raked masts and funnels were more attractive. But First Sea Lord found upright funnels more modern-looking, and rejected the argument that sloping kept smoke away from the bridge (it was effective only if the funnel were radically sloped, as in Japanese cruisers). Nearly all foreign navies had adopted either straight or very slightly sloped masts and funnels, and a very slight slope gave no advantage at all. The design was therefore modified with vertical masts and funnels. The issue of clearing arcs of anti-aircraft fire, which has often been used to explain the switch to vertical masts and funnels, seems not to have arisen.
First Sea Lord saw only one problem: accommodation. Something had to be done, particularly for ratings, to overcome the trend towards more and more crowded ships. When the Board approved the design on 14 January 1937, it asked Controller to look into improving ratings’ accommodation.
Building drawings were circulated to departments on 28 February, but the story was not yet over. In March DNC observed that the bottom of the armour belt might be uncovered by the ship’s wave pattern when she steamed at high speed. That required 1ft 6in of extra belt depth. The ship might also be vulnerable to a diving shell hitting short; that seemed to require 3ft below the bottom of the belt when at standard condition and not at high speed. The Japanese were interested in exactly such shells, and intelligence to that effect may have become available in 1936.22 He suggested adding 40lb (1in) D plating below the belt, or perhaps 60lb plating over a smaller area, enclosed by 60lb bulkheads. W G John, in charge of the design, pointed out that the waterline in action condition was 2ft above the waterline at standard displacement. On this basis the protective strips in way of the forward and after magazine blocks could be omitted, but a 2ft 6in belt should be worked below the main belt in way of the 4in magazines, because they were uncovered by the hollow in the ship’s wave pattern. John pointed out that this did not take into account extra exposure as the ship rolled; there was an even chance of a hit against the protected or unprotected area. It is not clear that this protection was provided, as DNC did not mention it when submitting the modified design to the Board a few months later.
Just how tight the design was became obvious when, in April 1937 Controller (Admiral Henderson) asked how much boiler rooms had to be shortened to save 70 tons (having learned that a foot less boiler room saved 3.1 tons on armour). Nothing could be done, because the space between ‘B’ and ‘X’ turrets was set by the demands of topside arrangement: bridge, hangars, catapult, 4in HA guns and boats. No matter how much speed was reduced, the ship could not be materially shorter. Controller then asked what speed the ship could make with 30 per cent more power. The power-speed curve was so steep that only about 2kts would be gained, and that would require a new design (which would take six months to produce).
Controller wanted to know what he could get by deleting the after DCT (14 tons) and perhaps an after 6in turret. Other items which could be deleted to save weight were the aircraft (143 tons) and the torpedoes (40 tons). Every ½in off the upper or lower deck (except over magazines) would save 35 tons. A ½in off the belt abreast the magazines would save 14 tons, and ½in off the belt abreast machinery and control positions would save 30 tons. More powerful machinery typically added something over 200 tons.
A series of designs was produced in which extra power was obtained at the expense of armour and, in many cases, aircraft and torpedo tubes.23 Boiler design kept improving, so that it appeared that on much the same weight 83,000 to 86,000shp might be achieved (total machinery weight would increase about 180 tons, the machinery box being 2ft longer).24 DNC offered two series of sketch designs, A and B, the latter lengthened from 526ft to 538ft (pp). In modified versions (A’ and B’) boosted horsepower was cut to 80,000shp, which gave a knot higher speed but reduced the necessary sacrifice to ½in of side armour, ½in of deck armour over the forward engine and boiler rooms, and two triple torpedo tubes. It might be possible to save enough weight to install the torpedo tubes upon completion; at the least, space could be provided so that they could be installed on the outbreak of any war. Controller chose Design B’. The delay in tendering due to redesign (to October 1937) was acceptable because it would not delay the programme; the crucial items were guns and mountings. Laying-down dates could be put back and the second armoured carrier (HMS Victorious) brought forward. First Sea Lord approved the idea on 28 May, but DCNS wanted a staff meeting called to decide whether greater speed was really worth the sacrifice of protection and the torpedo tubes. The Staff was particularly unhappy with the steady loss of torpedo power in the fleet. The British fleet considered that it enjoyed a particular advantage in night battle, in which torpedoes would be important. The meeting ruled in favour of retaining the original level of protection except that the belt abreast the machinery was reduced to 3½in, the torpedoes were sacrificed, and a speed of 32.25kts (standard) accepted.25 New towing-tank results showed that speed had been over-estimated by a quarter-knot. Controller decided that the side abreast the machinery would be 3½in and the side over magazines 3½in. All decks would be 2in thick. E-in-C would calculate machinery box length for 1,440-ton machinery (rather than 1,390 tons in the most recent B2 design).26
About this time DNC adopted a transom stern, in effect a longer stern chopped off, which offered the wave pattern, and hence the resistance, of a longer hull. The idea had been tested, somewhat unsuccessfully, in the minelayer Adventure.27 The towing tank (Haslar) estimated that effective horsepower (hence shaft horsepower) would be reduced about 4 per cent at 32kts and 3 per cent at 25kts. DNC expected about a quarter-knot more at full power, so that with 80,000shp he expected 32.5kts at standard displacement and 31.25kts at deep load (1,200 tons of oil).
K34, the 1939 alternative to continuing Fiji class construction, had the main battery reduced to three triple turrets but with four quadruple pompoms. Note also the sided HA directors abeam the second funnel. (Norman Friedman)
The Board approved the amended design on 8 July 1937. The after DCT was omitted and no weight was allowed for the torpedo tubes but space was provided for their installation. Elimination of the after DCT was balanced to some extent by placing a control position atop ‘X’ turret.28 All three HACS were retained, with two (not three) computer positions below deck. Adopting an 80,000shp plant (at the new high forcing rate) made it possible to reduce boiler size slightly and thus to reduce citadel length slightly. However, if the earlier forcing rate were accepted, power would be only 72,500shp. The higher maximum power required somewhat heavier turbines. DNC thought the added 90 tons well worth while, but in February 1940 E-in-C refused to sanction operation at full 80,000shp power because destroyers and fast minelayers had not yet forced their boilers even to the rate corresponding to 72,500shp. DNC was impatient with E-in-C’s proposal for full-power trials at both 72,500shp and 80,000shp: he had paid for the power, and he meant to get it. Trials reports show that ships typically developed something very close to the full 80,000shp. Bermuda made 31.091kts on 79,340shp at 10,900 tons
Shorter machinery spaces freed somewhat more space for accommodation, and DNC freed more by squeezing store-rooms. Space was also freed by using only turbo-generators, all in the main engine rooms (there were no separate diesel generator rooms) and by enlarging the superstructure somewhat. Beam was increased slightly, to 62ft (61ft 5½in exclusive of armour).
The Legend and drawings received the Board Stamp on 4 November 1937. Calculated standard displacement was 8,170 tons, within the margin Controller had offered a year earlier. Five ships were included in the 1937/8 programme, ordered in December 1937: Fiji, Kenya, Mauritius, Nigeria and Trinidad. Another four were included in the 1938/9 programme (ordered March 1939): Ceylon, Jamaica, Gambia and Uganda.
Pressure grew to install torpedo tubes; First Sea Lord was willing to have two on each side if he could not have three, and ACNS considered the torpedo ‘the decisive weapon at night or in low visibility’. It would be most desirable to have a three-torpedo salvo to deal with a modern fast target. The triple tubes earmarked for the ships were already being manufactured against the possibility that they would be installed.
British shipbuilders and designers had already wrung all they could from the design. Weight grew as minor improvements were ordered, such as a D.IV.H catapult instead of the earlier D.I.H. and a revised generator arrangement. In January 1939 the ship’s calculated standard displacement was 8,268 tons (8,298 tons if torpedoes were carried). At her launch on 31 May 1939 it became obvious that Fiji was badly overweight: she completed at 8,631 tons (standard). Somewhat later DNC credited the ship with a design displacement of 8,250 tons. Approved additions before completion added up to 319 tons, although more was clearly involved: growth in electrical equipment (70 tons), turret overweight (52 tons), upper deck stiffening (52 tons), added generating power (45 tons), the new catapult (35 tons), torpedo armament (33 tons), degaussing gear (10 tons), heavier cranes (10 tons), Type 281 radar (7 tons), and RPC for pompoms (5 tons). Another 33 tons was partly compensated for (further radars and Oerlikon guns). About 1941 DNC wrote that he had successfully resisted (or compensated for) another 432 tons, including 85 tons for jibs for the cranes and Mk V rather than Mk IV HADTs. The ships kept gaining weight, however. Inclined on 2 January 1942, Gambia displaced 8,846 tons in standard condition (average action condition was 10,167 tons).
By this time there was some question as to just how good a bargain the Fijis were. First Sea Lord wanted something less expensive (requiring a smaller crew), so Controller had DNC design a nine-gun ship, K.34. The initial version had 3½in side armour, but the Staff wanted more (4½in, as in a Southampton).29 On 19 May 1939 the Sea Lords met to decide whether the cruisers of the 1939 programme should be Fijis or K.34s. They were uncomfortably aware that foreign navies might disregard the 8,000-ton limit altogether (First Sea Lord referred to a 15,000-ton ‘monstrosity’ that Germany might build). As long as Britain continued to obey the treaty, at the least she should build to the treaty limit. Something midway between Fiji and Dido (presumably the original intent with K.34) was rejected. Second Sea Lord bitterly remembered the previous attempt to inspire foreign powers to limit themselves, the Leanders. ‘We had had to reply by increasing gradually up to the limit and the pendulum had swung back and forth. While we had wobbled, other powers were building up to the limit.’
The question was armour vs guns. Fourth Sea Lord and ACNS argued that a nine-gun salvo would be good and perhaps preferable to twelve guns and weaker armour (in the run-up to the Belfast class, twelve rather than fifteen guns had been chosen on gunnery grounds). However, ACNS also said that he saw no point in talking about quarter-inches of armour: particularly after a talk with DNE, he favoured twelve guns. DNC pointed to the delay inherent in any new design, while the Staff pointed out that the nine-gun ship would cost little less than a Fiji. The decisive argument was the basic British policy of closing and defeating the enemy with gunfire.
The Sea Lords refused to give up either aircraft or torpedo tubes (First Sea Lord wanted no more cruisers built without torpedo tubes). When DCNS said that the displacement limit could be disregarded, DNC said that was not the problem: stability was. The two Belfasts had both lost significant stability due to changes during construction. Fiji lacked their margin against such weight growth.
The Sea Lords decided to continue building Fijis. DNC thought it worth recording that Second Sea Lord (Admiral Sir Charles Little) said that in the past the Board had been criticised for switching from design to design ‘as though the Board did not know their minds’. The Fijis were the result of careful consideration, with the understanding that their powerful armament was paid for with protection; ‘on the grounds of consistency alone there was an argument in favour of perpetuating the Fiji class.’ The 1939/40 programme, the last before war broke out, was held to four cruisers (all Fijis) because overall defence spending was reaching the limit of affordability, but once war broke out six Didos were substituted for two of the Fijis because they could be completed sooner. That left Bermuda and Newfoundland, making a total of eleven ships. At this time it appeared that British industry could produce enough armament for ten cruisers a year (four Fijis, six Didos).30
Above and below: Like all other British cruisers of their day, the Fijis were very tightly designed, hence could not easily accommodate all the additions required during the Second World War. Removal of ‘X’ turret was part of the solution; only HMS Nigeria retained her ‘X’ turret beyond 1945. HMS Jamaica is shown on 12 June 1945. Her radars have been modernised. The stub mast forward of her after funnel carries a US YE aircraft homing beacon, important for fighter control (cruisers so equipped were sometimes called ‘pylons’, i.e., reference points for fighters). The Royal Navy was much more interested than the US Navy of that era in using surface ships to control fighters, both carrier- and shore-based. At this time the ship’s close-range anti-aircraft battery amounted to no fewer than five quadruple pompoms, four single power pompoms, four single Bofors, two twin power Oerlikons, and six single Oerlikons. Mauritius, Kenya and Nigeria all still had four triple turrets, but all but Nigeria were refitting.
Jamaica as she appeared from completion (June 1942) to at least the end of Operation ‘Torch’, during which she suffered blast damage (from her own guns, firing at low elevation) to the Type 273 ‘lantern’ forward of the bridge. Note the unique prominent aerial outrigger at the fore end of the compass platform, on the starboard side. Close-range armament at this time was two quadruple pompoms and ten single Oerlikons. She was docked on returning to the United Kingdom from the Mediterranean, and she may have been fitted at this time with the lightweight ‘lantern’ of a Type 272 atop her forward starfish. The big Type 273 ‘lantern’ remained, but probably was not repaired, and thus probably was no longer operating. Certainly a photograph dated 24 April 1943 shows both ‘lanterns’. However, note that an account of the use of radar in the Barents Sea battle (31 December 1942), in which Jamaica and Sheffield proved very successful, refers only to Type 273. It makes special mention of the excellent plotting work done on board HMS Sheffield, which may mean that Jamaica had inferior surface radar. The account of radar in the battle of the Barents Sea circulated by Director of the Signal Division (DSD) specially commended the plotting party on board HMS Sheffield, but warned against using the surface-search radar (at that time, Type 273) for rangefinding, despite its excellent ranging panel. DSD wrote that ‘Type 273 enabled the Cruiser Force to detect and plot two enemy units, probably Lutzow and Hipper, and make a surprise attack on one of them. At this time the enemy were undoubtedly concentrating their attention on the targets to the southward (the destroyer escorts [i.e., the close escort of the convoy]) and were thus caught unaware. The Cruiser Force was similarly surprised by an enemy destroyer later on.’ That referred to the use, by Sheffield, of her Type 273 for rangefinding, which made it possible for her plotters to miss the approaching destroyer. (Alan Raven)
After the ships had been in service for two years, 10th Cruiser Squadron answered a standard questionnaire. The ships were seaworthy, but very wet at moderate and high speed, and speed did not fall off excessively in sea and wind – 20kts could be maintained in rougher weather than might have been expected, actual loss of speed depending largely on the length of swell. Once the ship started to bump her speed dropped quickly. Steaming into a full gale with a moderate swell, the ship not bumping badly, she made about 9.5kts on revs for 12kts. Ships turned well, but could be much affected by the wind, given their large sail area. Thus a wind on the beam made them turn faster to windward and slower to leeward. Trying to turn while at rest into a strong wind caused the ship to drift bodily to leeward. This behavior was considered more marked than might have been expected. The bridge was well-arranged but cramped. Habitability seemed satisfactory given the ships’ size, but they were badly overcrowded, and became very uncomfortable when lower deck accommodations were closed up at night. The suggested solution was ‘restaurant messing’, also referred to as the American system. Boats were inadequate and inaccessible.
After war broke out in September 1939, the torpedo tubes were ordered installed despite overweight, but stability was so critical that weight had to be surrendered.31 More was wanted: in May 1940 the captain of the newly-completed Fiji complained that he had too little ahead anti-aircraft fire. He wanted UP (rocket) projectors, which the Royal Navy was then installing on board some ships, and additional light anti-aircraft guns at the fore end of the hangar-room deck (outboard of the navigational rangefinders). The Admiralty approved replacing the navigational rangefinders with two additional quadruple 0.5in machine guns. Typically a quadruple 0.5in could be replaced by a single 20mm Oerlikon machine cannon.
In September 1941, painfully aware that there was no margin for growth, the captain of HMS Nigeria suggested simply keeping some oil in his tanks at all time (he misinterpreted DNC’s comment that stability was satisfactory with 150 tons of oil on board).32 Nigeria already had two approved additional light weapons (Oerlikons) and had added two more (she had retained her original pair of 0.5in machine guns). DGD wanted the four-Oerlikon battery but DNC was unenthusiastic. ‘Stability owing to additions already accepted is below what it should be in a fighting ship intended to stand up to severe damage.’ Free surface made liquid ballast dangerous: ‘stability cannot be left to chance as, except very rarely, a ship can only be sunk once.’ Too many additions had already been made without compensation.
In February 1942 it was decided that all wood decking would be removed from the forecastle and upper decks, including the quarterdeck, as compensation for fire-control radars (Types 284, 285 and 282). Compensation for the Type 273 surface-search set was to remove two searchlights and their platforms from abreast the after funnel, the two searchlights with the best arcs being retained. A second pair of quadruple 0.5in machine guns was added at the cost of the tactical rangefinder and the balloon-filling equipment. A pair of Oerlikons on the quarterdeck cost the wood there. DNC suggested removing the centre gun of ‘X’ turret to compensate for two more Oerlikons on the quarterdeck, while also removing the cradle, shell loading tray, recoil cylinders, run-out and elevating gear would buy a third Oerlikon on the quarterdeck or a total of two on the hangar roof. By also removing the shell and cordite hoists to the gun, the ship could have four Oerlikons on the quarterdeck and two on her hangar roof. Single power 2pdrs could replace the hangar-roof Oerlikons. This desperate step was never taken, but it shows just how tight the Fijis were.
Wartime building practices ended the use of light alloys, which added more weight. Additions included RPC for 4in guns, two barrage directors for 6in guns, splinter protection for the HA armament (the largest new item), Arcticisation, spare 4in barrels, another two Oerlikons, and close-range predictors Mk I instead of pompom directors Mk IV.33 Only partial compensation had been obtained for the addition of radar and a pair of Oerlikons (net addition 33 tons). The only remaining candidates for compensation were removal of the port aircraft crane and boat stowage lowered between the 4in guns. Aircraft could then be recovered only from the starboard side. By February 1942 this step had already been accepted, reluctantly, by DNAD for HMS Southampton. Free access between the port and starboard sides of the 4in gun deck would be lost (the intention had been that either side could be supplied from either set of hoists). Boats would be free of blast, although they would need protection from ejected 4in cases. One searchlight could be carried just abaft the after funnel, giving a total of three.
One way out was to remove ‘X’ turret, as in the Ugandas (see below) and work on drawings began in August 1941. Depending on the availability of materials, it would probably take six to eight months after working drawings had been prepared. Alternatively, centre guns could be removed from all four turrets, but in April 1942 DGD refused to sacrifice any more 6in guns because the Japanese had twelve- and fifteen-gun ships (the Tone and Mogami classes); he did not know that all of these ships had had their triple 6.1in turrets replaced by twin 8in turrets. It was considered undesirable to land the catapult and aircraft until more carriers were at sea. Alternatively, some desired items could be foregone. In order of priority, in the spring of 1942 these were spare 4in barrels, splinter protection (including over 2pdr ready-use stowage) and the two extra Oerlikons, which would be traded for 6in barrage directors and 4in RPC.
The required compensation was so large not because the desired items were all very heavy, but because they were so high up in the ship. In topweight terms the two proposed barrage directors were equivalent to all three guns (and their gear) in ‘Y’ turret, the lowest in the ship. DTSD, in charge of Staff Requirements, considered it outrageous that two directors for an important but auxiliary function should require as much topweight as the armament of a whole turret. He argued for a split approach. Cruisers working with the fleet needed guns but not aircraft, because surely carriers would be present, while cruisers on the trade routes needed aircraft but could afford less armament. The crisis eased somewhat when the call for new pompom predictors was dropped, and when DNC found that the after HA/LA director could substitute for a barrage director. It proved possible to trade the crane and lowered boats for a barrage director forward and 4in RPC. Ships also added two Oerlikons on the roof of ‘X’ turret and two on the after superstructure.
Problems returned as improved radars were demanded. In September 1942 Type 272 (in place of the lighter Type 273) and barrage directors were requested for Kenya, a choice strongly supported by DSD.34 The attempt to rewire the after director as a barrage director failed (except for ships coming in for their rare large refits), so two barrage directors had to be installed aft. There was very little left to remove: either the third searchlight and the torpedo tubes or the aircraft and catapult. It was agreed that the latter choice gave a better-balanced ship. Ships involved were four-turret Fijis and repeat Fijis: Nigeria, Kenya, Jamaica, Gambia and Mauritius, of which Kenya was already in hand (hence required an immediate decision) and Nigeria was about to begin large repairs (there was enough time to convert her after director for barrage direction). DTSD approved this solution, and Deputy First Sea Lord agreed that there was no real alternative. First Sea Lord concurred on 2 October 1942.
Photographed on 24 November 1948, and little changed since 1945, Jamaica clearly shows her two quadruple pompoms and two single Bofors aft. The fifth pompom was on the centreline forward of the mainmast.
Removing aircraft freed weight and space, both in short supply. In Kenya one hangar was fitted with a flat at middle height and used for accommodation, the other being left as a chapel or cinema. Six twin Oerlikons could be added (four on the flight deck and two on the quarterdeck). There was sufficient weight for two 80kW diesel generators. Similar modifications were made to Jamaica and Nigeria. If both hangars were given flats, two twin Oerlikons could be mounted on the flight deck, while if both hangars were used for accommodation at flight deck level (without flats), eight twin Oerlikons could be added (six on the flight deck and two on the quarterdeck). HMS Bermuda proposed a more elaborate conversion, the remaining crane being moved to the centreline and the waist closed in with plating (other ships just moved the crane to the centreline). Proposals to use the hangars as workshops were rejected, presumably because of the topweight involved.
Both Cammell Laird and Vickers promoted new-construction Fijis for post-war export. Cammell Laird offered the design to Argentina in 1947, while Vickers offered a new-construction Fiji or Swiftsure, with suitable changes in equipment and electronics, to Chile in 1946, together with Dido. Vickers offered a somewhat modified Dido to Venezuela in 1948, but none of these deals succeeded.
The Heavy Cruiser Problem Revived
Through the late 1930s the Royal Navy faced a terrible possibility. The Fijis symbolised the sacrifice it had made specifically to stop construction of heavy cruisers, but some countries capable of building such ships had signed neither the 1930 nor the 1936 treaty. They or some customer might build a cruiser armed with 8in guns, reviving the idea and making the treaties, and the sacrifices, irrelevant. In 1937 Chile presented exactly that possibility. Chile and Argentina were rivals, and in the 1920s Italy had built two cruisers armed with 7.5in guns for Argentina. Although they were often described as overarmed failures, the other two major South American powers, Brazil and Chile, naturally wanted equivalents and in 1933 Vickers designed an 8in cruiser for Brazil. Even though no British yard would have been allowed to build it, Vickers could have supplied the ship via its Spanish shipyards. For some reason the prospect did not alarm the Admiralty or the Foreign Office, perhaps because in 1933 there was no sense of headlong rearmament around the world. Also, the project may not have aroused much concern because it had little chance of coming to fruition, given the state of Brazilian finances at the time.
But in 1938 the situation was different. The three major South American sea powers all had ambitious building plans which seemed to have a fair chance of materialising. A ship built for any one country might easily be sold to or requisitioned by a European power, as South American battleships building in 1914 had been by the Royal Navy. In 1937 Chile was buying aircraft from Italy and Germany, hence might be moving towards them (the Chilean Navy was equipped with British-built ships). In July 1937 a Chilean naval representative came to the United Kingdom with a firm offer to order two 8,500-ton cruisers armed with six 8in guns in two triple turrets (the Argentine ships had three twin 7.5in turrets).35 Neither the displacement nor the armament was permissible. At first it seemed that only Italy and Japan could build the ships, but the Chileans later also contacted Finland, Germany, the Netherlands and Sweden. DCNS pointed out that the construction of such a ship would give any signatory a perfect excuse to invoke the escalator clause of the treaty in such a way as to kill the limit on 8in ships.36 The Chileans approached three British builders: Vickers, John Brown and Scott’s. Vickers offered two Southamptons (which could not be built under the treaty). In September 1937 the Chileans increased the pressure, offering a guarantee that ‘if we [the British] are in trouble’ they would turn the ships over, as in 1914. The Chilean Ambassador said it was his President’s ‘ardent wish’ to have the two cruisers delivered before he left office.
HMS Bermuda at Melbourne, 9 September 1946: ‘X’ turret was replaced by two quadruple pompoms; two more are on the nearby after superstructure. She also had four single power pompoms, two single Bofors, two Boffins and two single Oerlikons. (Allan C Green via State Library of Victoria)
Given the sheer size of the British rearmament programme, there was little hope of quickly building two cruisers for Chile. Most of the ships under construction or on order were destroyers, the main exceptions being the Argentine cruiser La Argentina and a Soviet order for cruiser/destroyer machinery (which would figure in early British wartime thinking). To Controller the great bottleneck was fire-control equipment, some new British cruisers (presumably Arethusas) having only one instead of two HACS. Equipment had been diverted for the Argentine cruiser, and Vickers was making predictors (AA computers) for the army. The prospective Brazilian destroyer order would absorb capacity freed when the Argentine cruiser was completed. Controller wanted British yards to promote foreign fire-control systems for their export warships (new Greek destroyers had German systems) or have them made abroad.
By early 1938 Japan was outside the treaty system, and DNI cautioned that she might quickly revert to building 8in cruisers (no one knew that the 6.1in-gun Mogamis had been designed for wartime rearmament with 8in guns). When that did not happen, the British began to hope that nothing would happen – unless the Chileans found a builder for the 8in cruiser they wanted. The Chileans asked the British whether invocation of the escalator clause for battleships had freed the country to build 8in cruisers. It had not. The British thought (incorrectly) that it was working effectively to deter the Japanese from breaking out in their cruisers. In September 1938 Vickers was told that it (in combination with John Brown) could offer Chile two repeat Fijis, to ensure speed of delivery. Details of the design, sufficient to allow bidding, were released to the Chileans. Rumours included what turned out to be a false claim that Bofors had already begun work on 8in guns for Chile. Krupp was reportedly willing to supply guns for hulls to be built in the Netherlands. As a backup, in case they were unable to buy 8in cruisers, the Chileans asked Great Britain, the United States, France, Italy, Sweden, Finland, Germany and the Netherlands whether they would be willing to build two 8,000-ton cruisers armed with 6in and 4.7in guns, with a speed of 32kts and an endurance of 8,000nm at 14kts.
The British felt compelled to go to each potential builder to secure an agreement that no 8in cruisers would be built. Negotiations between Chile and Finland broke down. The Germans rejected a British claim that the Anglo-German Naval Agreement of 1935 barred them from building such ships, but they were not yet ready to break with the British, so in September 1938 they agreed not to allow Krupp to build 8in guns for Chile (the ‘10,000-ton’ Admiral Hipper class cruisers they were building were still to be armed with 5.9in guns, the switch to 8in not yet having been announced). The Dutch abandoned negotiations because they were relying on Krupp 8in guns, although they were still interested in building 6in cruisers. The most extraordinary aspect of the affair was that the British offered to recognise the Italian conquest of Abyssinia (which had almost brought the two countries to war) in exchange for an agreement not to build 8in cruisers. In January 1939 the issue seemed that important.37 The Chileans were not amused, and at a dinner in November the British Ambassador noted the bitterness of junior Chilean officers.
HMS Nigeria (laid up) was the last Fiji to retain ‘X’ turret.
By January 1939 the Chileans had largely but not completely accepted they could buy only 6in cruisers. It was rumoured that the Dutch were building two cruisers ‘more or less on the assumption that Chile will take them over, if and when she can pay for them’.38 The Chileans continued to support the project despite the enormous cost imposed by a severe earthquake, the money having been set aside for that purpose. The Foreign Office was particularly concerned to maintain the British position with the Chilean Navy, and pressed for export credits. In May 1939 the Chileans abandoned their search for large cruisers in favour of two light cruisers of about 3,000 tons with 5.5in or 6in guns and a speed of 36kts (they described the ships as something like the French Mogador class super-destroyers), and Vickers tendered for the contract. The whole cycle began to repeat in October 1939, with reports of fresh Chilean approaches in Europe and in the United States, and an Italian offer of an 8in cruiser, but the outbreak of war stopped further developments. This episode shows just how badly the British government wanted to kill off 8in cruisers in the run-up to the Second World War.
Larger Cruisers
Although prohibited by the 1936 treaty, larger cruisers remained interesting, not least because there were continual rumours that the Japanese were building 20,000-ton super-cruisers, and because it was considered entirely possible that the Germans would do much the same thing. In January 1938 DNC asked his preliminary cruiser designer, W G John, to sketch a variety of heavy cruisers, beginning with a 20,000-ton ship (700ft x 70ft x 25ft), to make 33kts at standard displacement, with high endurance. Armament would be three quadruple 9.2in turrets, six twin 4.5in and four octuple pompoms, but there would be neither torpedo tubes nor machine guns (0.661in). The ship would have two aircraft and a deck catapult. Armour would be an 8in belt covered by a 4in deck, with underwater protection against a 750lb charge, as in a capital ship. The designer was to try to ‘scrape down’ to not more than 15,000 tons; cost should be about £5 million. Admiralty Board minutes do not suggest any interest in building such ships, but in January 1938 it was just becoming clear that Japan could not be enticed to abide by the new treaty restrictions (the British thought the Japanese were designing ships only slightly above the allowable tonnage), and the design may indicate British interest in a cruiser escalator agreement. The design was developed in some detail.39
The ship would be about the size of a carrier; DNC suggested starting with 132,000shp. DNO provided data on a quadruple 9.2in turret. The belt armour would resist 10in shells 40° from normal at 16,000yds; the deck was expected to resist 10in shells and also a 1,000lb AP bomb dropped from 8,000ft. DNC suggested a short forecastle design with a hangar and the usual alternating engine and boiler rooms, the machinery box being pushed well aft (as in Belfast). A cross-section showed a 16ft 6in deep belt extending 11ft 6in above the waterline, with an internal deck below the armour deck but well above water, and a side protective system inboard of the belt. Hull depth would have been 35ft 6in. The belt would have been closed by 6in bulkheads. As in recent much smaller cruisers, the belt armour, rather than armour boxes, would have protected the magazines. The 20,000-ton hull actually proved somewhat small, so John tried a 23,700-tonner as well.40 To get back down, John tried triple 9.2in guns, but found he did not have enough machinery weight to make 33kts. He had to reduce armour.41 By early February 1938 this design was being called an armoured cruiser. In mid-February John reported to DNC, who asked about the effect of substituting triple for quadruple 9.2in turrets. The docket and drawings were sent to Controller on 17 February 1938. As reported, the ship with three quadruple turrets would displace 20,750 tons and would cost £5.2 million, while the alternative with three triple turrets would displace 23,500 tons and cost £6 million.
DNC next asked what could be done with nine 8in guns (Design A), or with nine 9.2in (Design B). Speed was set at 33kts and endurance at 10,000nm at 15kts. Other armament was six twin 4.5in, four octuple and two single pompoms, and twelve torpedo tubes. Armour should defeat the ship’s own guns at 90° inclination (worst case) between 8,000 and 25,000yds (with an inch less over machinery). The deck over the magazines should defeat a 500lb SAP bomb dropped from 10,000ft (over machinery, a 500lb dive bomb); and underwater protection should, as before, defeat a 750lb charge. For the 9.2in gun, this was more than had been asked for before: 10½in C over magazines and 9½in over machinery, with a 3½in deck over magazines and 2½in over machinery.42 For 8in shellfire it was quite heavy, too: 9in (8in) belt and 3½in (2½in deck). Deck armour was the same in both cases because it was determined by the bombs, not the shells. This was the sort of protection the US Navy later provided in the Alaska class, not even approached in any conventional cruiser. Cutting the number of guns in Design B had little effect; the ship would still displace about 20,500 tons and would require 182,000shp to make the desired speed, probably using eight 25,000shp boilers. A sketch showed a ship with a waterline beam of 84ft, bulged underwater to 88ft, so that the battleship-style underwater (side) protective system could be 10ft deep. DNC received the report of this pair of designs on 28 February and in turn reported to Controller on 4 March. The ship with nine 8in guns was expected to displace 21,750 tons and to cost £5.5 million.
Based on this work, in December 1938 DNC asked for details of an Alaska-like super-cruiser, at 20,000 tons, armed with six 12in guns, with 7in belt and 3in deck, which was what the Japanese were (incorrectly) reported to be building.43 Unlike the 8in and 9.2in cruisers, it was legal under the 1936 treaty, because it came in above the prohibited cruiser zone (8,000-17,000 tons, 6.1in to 10in guns). The estimate was based on the 9.2in ship reported the previous February (700ft x 84ft x 23ft), the twin 12in turret being comparable to the triple 9.2in. Work on such ships resumed in February 1939, with analysis of an 18,000-tonner armed with six 10in guns and twelve 4in HA/LA guns, protected against 8in fire, with a speed of 32/34kts and a cost of about £5 million. Protection was that calculated the previous year, 9in side and 3in deck, offering immunity (against 8in fire) between 8,000 and 25,000yds. The deck could resist 1,000lb AP bombs dropped from 4,000ft and 500lb SAP from 7,000ft. On 10 February John reported to Lillicrap that a ship with six 10in guns, otherwise armed as Belfast and protected against 8in shell, with a speed of 33kts, could be built on 18,000 tons (weights added up to 18,550 tons). The ship had 9in belt armour over her magazines and 8in over her machinery, covered by 3½in and 2½in decks respectively. References in the Notebook suggest that there was still interest in 9.2in main armament. Presumably 10in guns, which no major navy used, was interesting because it was the smallest calibre permissible for a capital ship (i.e. not a prohibited heavy cruiser) under the 1936 treaty. Estimated shell weight was 500lbs, roughly twice that of the 8in cruiser gun.
In June, DNC asked for studies of a 10,000-ton cruiser armed with 8in guns, in effect a ‘County’ using current technology.44 Presumably Controller (and the Board) envisaged a partial but not complete breakdown of the treaty, so that 10,000-ton 8in cruisers could once more be built under the specific treaty clause allowing a breakout without mutual agreement as to new limits. Hence, as calculations were made, small excess displacements over 10,000 tons were rejected as unacceptable. DNC envisaged a flush-deck ship armed with three triple 8in guns, four twin 4in, four multiple pompoms, and two triple torpedo tubes, plus the usual aircraft and catapult, capable of 32kts, with a 2in deck. Could such a ship have 6in side armour? Variants should be (a) protected against 11in fire (i.e. German ‘pocket battleship’ fire) at fine inclinations; (b) with armament forward; (c) well protected against 8in fire; and (d) without 4in guns and pompoms. The ship would have new heavy 8in shells (290lbs). On DNO advice it was assumed that the 8in turrets would elevate to 60-70° and that the guns would be as far apart as in the twin mounts of the ‘County’ class – which guaranteed that turrets would be large. Analysis suggested that a 32kt ship would displace just over 10,000 tons. Data were worked up for both a 32kt (80,000shp) ship (Design A) and a 30kt (58,000shp) ship (Design B), both with 5in belt and 2in deck, Design B coming closer to 10,000 tons, but by surprisingly little.45 As an alternative, John considered a ship with four twin 8in mounts. That lengthened the citadel and thus required more armour (1,920 tons vs 1,850 tons). The estimated weight of an 8in triple was 328 tons, compared to 218 tons for the twin, so in this analysis adopting twin mounts saved considerable weight (984 tons for triples, 856 tons for twins). Designs C and D were 32kt and 30kt alternatives with the four twin mounts (10,250 tons and 9,995 tons), Design D having ½in more belt armour. The next possibility was eight guns distributed as two triples and one twin: 10,165 tons. All versions used the same hull, presumably to make comparison easier. Different levels of protection were also evaluated.46
DNC preferred the ship with three triple 8in guns, and E-in-C was asked about machinery. It would be the 80,000shp Fiji plant, but possibly with the 4in magazine forward of the machinery instead of between the after boiler and engine rooms, pushing the machinery box aft. By July, DNO was working up a turret design. As sketched on 29 June 1939 the ship looked like a cross between Belfast and Fiji, with a substantial gap between her bridge structure and a hangar built around her forefunnel. Like a Fiji, the bridge carried a DCT on its centreline and HADTs on either side. The forward 4in guns were in this gap. The ship was flush-decked, but with cut-outs in the forecastle deck so that the after 4in guns (between the two funnels) and the torpedo tubes were a deck lower. An after superstructure just forward of ‘X’ turret carried a third HADT on its centreline, with the after DCT below and abaft it. This structure also carried the ship’s pompoms. This version superseded a more Fiji-like ‘first attempt’.47 John ended up with a 580ft x 69ft 3in x 17ft ship (10,576 tons).48
Once war broke out, there was no longer any point in a 10,000-ton limit, and Winston Churchill, back as First Lord of the Admiralty, pushed for a powerful ship. John was assigned to develop a new 14,000-15,000-ton cruiser armed with 9.2in guns and protected against 8in shellfire, with a good radius of action, higher speed than the German 8in cruisers (say 33/34kts), six twin 4.5in anti-aircraft guns, and four quadruple pompoms; she would have no torpedo tubes. The ship would carry the usual two aircraft and one catapult, and she would be protected against aircraft torpedoes.49 Given the earlier studies, John chose a 7in belt and 2in deck (3in if possible), considering a 9in belt excessive. Cruisers would generally fight at something other than 90° inclination. At 90° it took 9in to keep out 8in fire at 8,000yds, but 7in would be enough at 8,800yds at a reasonable 60° inclination. The end of treaty restrictions shows in DNC’s instruction: ‘Let the displacement come to what it will.’50 The ship quickly grew back to what John had been sketching early in 1938, about 21,500 tons and 700ft long.51 The design was reported on 8 September, and it formed the basis for a further armoured cruiser project described in Chapter 9. As an alternative, John was asked for a ship armed with four triple 8in, but otherwise as the 9.2in ship.52 Both the ship armed with nine 9.2in guns and the one with four triple 8in were expected to displace 21,500 tons and to cost £5.5 million.
These designs suggested that the earlier 10,000-ton cruisers had been a mismatch of guns and ships, and that the ideal was something far larger – and far less affordable. Director of Plans concluded that it would be much better to build as many fast capital ships as possible. Big cruisers were clearly poorly suited to work with the fleet: they would rather fight enemy cruisers in minor actions and on the trade routes. Under such circumstances much would depend on effective range, because it might take hours for a cruiser to close from the range at which an enemy was spotted to firing range. The larger gun offered longer range, both absolutely and because its splashes were easier to spot. Against that, although a design for an 8in turret was proceeding, nothing was being done about the 9.2in. The pits used to build 6in turrets could probably be adapted to 8in, but not to 9.2in, so adopting the larger gun would make for further delays. If any ships were to be built, perhaps they should be faster 8in cruisers more likely to close their targets before nightfall. This reasoning led to the decision to design a new fast battleship using four existing twin 15in turrets, HMS Vanguard.53 However, the 8in (but not the 9.2in) heavy cruiser survived in British building plans.
Director of Plans decided that sacrificing speed might be better than sacrificing armour, so John estimated the effect of holding speed to 27 or 29.5kts. By December 1939, John was working on a cruiser with three triple 8in turrets, other characteristics being as in the earlier 21,500-ton ship. This design is best described as part of the wartime design series, in Chapter 9.
The 8in and 9.2in cruisers were conceived in response to reports that the Japanese were building four or five ‘pocket battleship’ equivalents armed with 12in guns. The choice in 1939 (for the 1940 programme) was two such ships or a 40,000-ton fast battleship. The latter was chosen; it became HMS Vanguard.
In the months before war broke out, the Admiralty developed a new fleet plan including 100 (rather than the earlier seventy) cruisers, because it had to take into account three potential enemies rather than one: Germany, Italy and Japan. The 100-cruiser force would have comprised twenty heavy, forty medium, and forty small cruisers. The Royal Navy had fifteen heavy, thirty-six medium (including thirteen building) and thirty-five small (including ten building) cruisers, a total of eighty-six, but four Hawkins and twenty-one ‘C’ and ‘D’ class were either overage or soon would be. Pre-war planning was complicated by a treaty requirement to inform the Germans of near-term building plans, which the Germans in turn might use to justify breaking out of the limits imposed on them by the Anglo-German Naval Agreement of 1935.
HMS Devonshire is shown on 29 June 1944, her catapult having been removed and her surface-search radar (the ‘lantern’) moved aft. She had been refitted on the Tyne (18 May 1943 – 20 March 1944), her ‘X’ turret being removed, together with her catapult. Another two octuple pompoms were added. Her two-antenna Type 281 was replaced by a single-antenna Type 281B. In October 1945, Devonshire was credited with six quadruple pompoms, seventeen twin power Oerlikons, and six single Oerlikons. Of the surviving heavy cruisers, only she, Norfolk, and Sussex had had their ‘X’ turrets removed.