The Devil's Chariots

by John Glanfield

  Central Workshops at Erin had been the first to improve trench-crossing performance by cutting a Mk IV in two and inserting panels to stretch it by 6ft. Metropolitan was instructed by MWD on 3 October 1917 to do the same, working to Rackham’s drawings, but the order was cancelled two days later. Instead, Foster’s was asked to stretch three Mk IVs by 3, 6 and 9ft respectively, and to fit Ricardo 150hp engines. Tritton tackled the job differently. Instead of Erin’s cut-and-shut infill, he extended the track frames rearwards by buttstrapping a pair of mild steel horns to them. The first Mk IV conversion was completed in December. The 9ft ‘tadpole tail’ extension was adopted for Mks IV and V, the kit being interchangeable and easy to fit. (A set was tested the following May on an experimental Mk V Star, boosting its length to an almost unsteerable 41ft.) When Frank Searle said he could convert ten Mk IV per week, 300 sets were ordered. The first reached Erin in May 1918, where rows of extensions soon cluttered up the rail loading ramp area. There is no record of stretched Mk IV tanks or tadpole tails seeing action.

  Churchill remained infatuated with the idea of a supertank. Anticipating a need for fighting machines capable of crossing the canal systems which lay ahead in a general advance, he told a doubtlessly shuddering d’Eyncourt that he wanted 12 250–300-ton ‘Reservoirs’. With characteristic wit his cover name reflected their giant size relative to the ‘Tanks’. He envisaged machines up to 90ft long to carry four 18-pdr guns and 12 machine guns, with a 100-mile radius of action. They could be built near the coast in England before crossing over on special barges and driving crosscountry to the front. Elles and Capper favoured the idea; Butler disagreed. D’Eyncourt passed the assignment to Moore who reported in December that every draughtsman was needed to prepare working drawings for Mk V and the Medium B. All attempts to date to produce tanks over 50ft long had failed because of the great weight of machinery necessary to deliver power, which in turn raised transmission and track problems. He was accordingly preparing sketches for a 50ft machine with a breadth of 13ft and weighing 80 tons for cross-Channel shipment in sections and then movement by canal. Power from two of Harry Ricardo’s 225hp engines, soon to be built, would drive ‘four parallel tracks each side, having a heavy short and a light long track’. It would mount two 12-pdr guns in sponsons, and at least ten Hotchkiss MG.9 The Tank Committee discreetly shelved the scheme.

  1918 opened with a reorganization of the Tank Corps in January in the light of experience at Cambrai. The current establishment was for 18 battalions, nine heavy and nine light, each with 60 tanks. The balance was now adjusted to increase the number of heavy tank battalions to 12, each with 48 tanks, and reduce the medium-equipped to six, each with 65 machines. At three battalions to a brigade the establishment allowed for growth to six brigades. Tanks were additionally demanded to equip two further brigades with heavy and one with medium machines. In all, the requirement for fighting tanks for 1918 to early 1919, including replacements at 20 per cent per month, totalled 1,654 heavies (increase of 406), and 1,120 medium machines (increase of 92). Other requirements comprised 325 tank tenders (216 of which were the converted Mk IV), 18 wireless-equipped tanks (principally Mk IV), 9 cable laying tanks of infantry-carrier type, and 312 infantry carriers. It amounted to a small increase in fighting tanks, from 2,276 in October pre-Cambrai, to 2,774 – a token endorsement of the weapon’s staggering performance.

  Churchill received this information coolly. He was pushing the tank programme hard regardless of military calculations. He told Moore and Stern that he wanted them to gear up to achieve 10,000 tanks per year, 4,500 from Britain and the rest from a new Anglo-American factory in France. It was a tall order for Moore, involving vastly more steel and a monthly allocation of 1,400 6-pdr guns and 5,000 machine guns, against current supply of 320 and 1,600 respectively. All the 6-pdrs would have to come from Armstrong’s, the sole suppliers. America’s vast industrial resources could not be turned around to volume war production of armour plate and ordnance for many months, perhaps a year. In addition, Stern had taken with him some key members of the MWD headquarters team including the production director; the MoM establishments branch blocked Moore’s repeated requests for more management staff; the War Office persisted in ordering in small batches followed by non-productive delays before reordering; and Moore’s design and drawing office departments were heavily over laden. Working drawings for the Mk IV conversion engaged almost the entire staff, delaying a start on Medium B. Rackham’s Mk VIII involved hundreds of drawings and caused draughting work on Mks V, V Star and Medium B to be transferred to Metropolitan in Birmingham. Manufacturers were unable to maintain delivery schedules for supply of gearboxes and epicyclics. By the time Mk V production started in February it was clear that output would be over 100 tanks below estimate by the end of March. Churchill intervened, securing Dudley Docker’s undertaking to build 660 Mk V and V Star by 30 June – in the event Metropolitan managed a respectable 518.

  Moore put in a demand to the Labour Department in February for a further 1,040 men and 260 women for assignment to Metro. If approved, he expected production of the Mk V series to reach 40 per week in April. Churchill directed that labour should be diverted from gun manufacture and that tanks were to have the highest priority for all resources controlled by the Ministry. Later that month he formed ‘Council Committee 57 – Tanks’ to accelerate delivery of the 1918 programme and to plan production for 1919. Duckham and fellow Council members controlling ordnance and materials collaborated intensively for three weeks to organize reinforcement of Moore’s operations. Churchill set them a national production target of 100 tanks per week. He also made a point of advising them to invite Stern’s proposals. Production improved considerably for a few months but could not be sustained in face of competing demands for resources, not least for steel for the Admiralty. Moore’s hand was greatly strengthened, however, by the appointment of the energetic and vastly experienced Percival Perry as his Deputy Controller on 21 March.

  Many production difficulties were quickly overcome with resourceful management and the excellent relations which existed between manufacturers and their local MWD supply officers. Other problems were less straightforward. Medium A production was slowed by a severe shortage of ball bearings for the track rollers, an outsize having been specified in the mistaken belief that the country held large supplies. Production of engines and gearboxes fell sharply from late July through September after the Ministry took over direct control of forging facilities in May. Only 60 per cent of the stampings required for tanks, aircraft and motor transport were met. The drop stampers assigned by the Ministry to Moore’s MWD were unfamiliar with the class of work required by the engine-builders, and both sectors refused the orders. After daily conferences with the Department of Forgings and Castings, the MWD undertook to place orders for the stamping of gearboxes, blanks and engine parts directly with stampers in a series of unsatisfactory compromises, and output was not fully restored until control was lifted in September.

  Moore was caught in the crossfire between the supply side led by the Prime Minister, Churchill, Stern and d’Eyncourt who saw the tank as a decisive weapon to be built in large numbers to an acceptable but not ambitious performance standard for use en masse, and the army – including especially Capper and Elles – which wanted fewer machines with regular performance upgrades for limited tactical objectives. The army was unconcerned that high output was attainable only with long unbroken production runs. It could point out that the Ministry had consistently failed to meet its own production forecasts and any call for higher volumes of a given type would simply lock in obsolescence over an even longer period. The obvious case in point was the interminable production of the 1,015 Mk IV fighting tanks, the largest and longest run of any British tank in the First World War. Deliveries began in April 1917 and Armstrong’s despatched the last machines in October 1918. By then the design was so obsolete that 70 or 80 ex-works machines were dumped in a park on Newbury racecourse where they were embarrassingly visibl
e to passing trains at a time when every tank was known to be wanted at the front. Capper rated long runs of a given type as ‘a distinct evil… Too much standardisation cripples advance in design and must always leave us with numbers of out of date machines.’10

  In pursuance of its policy the army had halved production of Tritton’s Medium A the previous October and instead ordered Wilson’s Medium B, 650 of which were contracted out to builders. In January 1918 the ‘B’ Type in turn was considered inadequate and Elles called for its cancellation in favour of a much superior Medium C for which Tritton had just built a mock-up. It was 1.5 tons heavier than the ‘B’ Type, offering male and female versions and nearly twice the radius of action at 120 miles. Moore told him that the Medium B was too far advanced to stop and the first 120 would be built in May; this was purely a blocking move – the first machine was rolled out in September. D’Eyncourt pointed out that it would be a very serious matter if word got out that the Ministry was throwing labour and scarce materials at an unwanted tank. Moore asked Metropolitan to try to phase out the ‘B’ Type, switching labour instead to boost Mk V output pending a start on the Medium C. Six weeks later, on 24 February, Elles condemned the Medium C in turn as being too slow at 6.5mph, too heavy at 19.5 tons to load two per tank rail wagon, and inadequately armed with the proposed 2-pdr gun which would be useless as a tank destroyer.11 A redesign to provide the 6-pdr and increase speed to over 8mph was followed by an order for 200 in March and a further 400 on 31 May. Faced with these changes of direction it was hard for Moore to plan and place work coherently, or to avoid drops in output while responding to new orders, revisions, cutbacks and cancellations. Whereas Stern the banker had periodically ignored or exceeded War Office demands, Moore the admiral tried reason. He had been Director of Naval Ordnance until 1914 and was more at home than Stern with production processes and disciplines. (It was Moore who gave Tulloch his ‘Tri-nitro Tom’ nickname.)

  The progress of the gun carriers was another object lesson, not least to Churchill. In August 1917 the War Office had doubled its requirement to 100 machines. The 52 in the new batch were to be built as gun transporters rather than firing platforms. The order for these Mk II carriers brought renewed objections from the Ministry’s transport department which demanded the removal of all gun-carrier work from Kitson’s in Leeds to free capacity there for priority locomotive construction. Stern asked Metropolitan to take the work. The War Office then cancelled the order in September, asking if the balance of unbuilt Mk I carriers could nevertheless be constructed as Mk II. It was wholly unrealistic to retool for what amounted to some 34 machines of substantially new design. Instead, and without reference to the WO, Churchill authorized construction of 150 Mk II GC chassis, 100 of which would be built as supply tanks, ten as tank salvage (recovery) vehicles, and the remaining 40 as gun transporters. Five days later, on 15 October, the War Office withdrew its cancellation and ordered 50 Mk II gun carriers plus 20 salvage tanks.

  The saga continued on Stern’s departure that month, the War Office objecting that it had not ordered the remaining 80 supply tanks on GC chassis, preferring to await the purpose-built Mk IX. Churchill again changed the order – to 120 gun carriers and 30 salvage machines. When the War Office later repeated its refusal to accept any GC types in excess of its order for 70, it was given the customary blocking reply that production was too far advanced to halt – a large exaggeration since no Mk II was ever built. A further revision followed in January 1918 when the WO trimmed its order for salvage machines to 17. Designs for the gun carriers and salvage tanks were still not signed off as approved when in March the War Office ordered all work on GC types to cease. The surviving Mk I carriers were fitted with cranes at Central Workshops for salvage duty. Churchill had experienced the same vacillation from the Army High Command which had so enraged Stern. Their reactions were identical, but overriding and second-guessing military requirements was no solution.

  Adm Moore favoured a ‘divide and rule’ policy concerning his main tank contractors. He told Duckham:

  We desire to register the strongest possible protest against relying too implicitly upon the resources of the Metropolitan Company or any other one firm. Past experience indicates that this Department must control the situation and such control is impossible when too much reliance is placed upon one contractor. There seems to be a desire on the part of the Metropolitan Company particularly, to achieve a position of monopoly in Tank Production. If this ambition is fostered then the Department will be subservient to the Contractor.12

  Moore’s policy meant further fragmentation of final assembly, already shared by 12 firms in a chain of plants between Birmingham and the Clyde. They all relied in varying degrees on a network of 4,000 specialist suppliers and services. Metropolitan and its subsidiary Patent Shaft dominated with a total capacity of 210 tanks per month, dwarfing Marshall’s of Gainsborough which was the next highest at 30 (Mk IX supply tank). Coventry Ordnance of Glasgow (Medium B) and Foster’s (Medium A) were geared to build 22 and 20 per month respectively. The remaining assemblers fell below these levels. Total monthly productive capacity was 350 heavy tanks or equivalent tonnage. Moore may also have feared the crippling effect of industrial disputes on such a concentration of output at Metro. The 1915 Munitions of War Act outlawed strikes and lockouts, but strike action persisted intermittently throughout the war. The Darlaston area was almost shut down by strikes in February 1918, cutting all supply of bolts and rivets to Metro and delaying Mk V assembly. Strikes that month on ‘Red Clydeside’ hit Coventry Ordnance, preventing use of new machinery for tank stampings, Beardmore’s, where riveters working on tank assemblies were out for over four weeks, and Foster’s in Lincoln, which consequently ran short of track shoes. The printed letterhead of Crossley Brothers of Openshaw, Manchester, a major supplier of Ricardo tank engines, is indicative of the general state of industrial relations. It cautioned: ‘We do not accept any responsibility for delay in execution of orders due to Strikes, Lock outs etc. etc.’ This was no idle disclaimer. Production of 2,100 150hp units was crippled for several weeks later that year by the refusal of Crossley’s fitters to accept 200 female operatives.

  Moore established a new production centre in the vast workshops of Glasgow’s North British Locomotive Company (NBL). By 1914 the company had built over 20,000 steam engines in Europe’s largest railway engineering complex. Churchill ensured that much of its capacity was turned over to tank production despite the higher priority attaching to locomotives. NBL was appointed lead contractor for British Mk VIII tanks, undertaking in April to build 1,040 machines. It was hoped monthly output there would reach 90 tanks by October. Another Glasgow tank plant was initiated at the same time at the suggestion of Sir William Beardmore whose company managed one of 16 National Projectile Factories for the Ministry. The Cardonald works was producing heavy shell in ten of its 13 bays; he proposed conversion of the remainder for tank assembly. Proximity to Beardmore’s Parkhead steel mills which rolled much of Moore’s plate was an added advantage. Additional tank bays were to be constructed and an order was placed for 335 Mk VIII machines for delivery at ten per week, doubling to 20 on completion of the extension in July/August. In the event no tanks were built there, the order was cancelled in September and the extension remained uncompleted – hope had outrun the reality of machine tools and other shortages at that late stage of the war. Moore also signed up a South Lancashire consortium led by West’s Gas Improvement Co. Ltd of Manchester. The group of some 25 engineering firms became the Manchester Tank Group which was contracted on 7 March to produce all 1,500 Liberty Mk VIII hulls for shipment to France for fitting out there. It was also to build the Ricardo 300hp engines at 320 per month.

  A site for the Anglo-American tank factory had quickly been found 150 miles south of Paris at Chateauroux with good rail links to the capital and the Channel supply ports. A nine-bay plant was to be built. It would assemble 300 tanks per month and be capable of extension to 1,200 monthly. Despite numerous
appeals the French authorities did not grant permission to start work on site until the end of February 1918. Meanwhile, production of armour and other components in Britain and America awaited working drawings from Rackham’s Mk VIII design team in London. Rackham found himself working on two types – a Mk VIII to be built in Britain for Sir Douglas Haig’s forces, and a Chateauroux-built version which varied from the British in numbers of gears and other details. Design was proceeding in parallel with preparation of working drawings, a situation always to be avoided. Additional draughtsmen were engaged and funded by America, and Moore’s offer to turn his entire drawing office over to Mk VIII work was gratefully accepted. The Allied design was approved on 7 March by Gen Sir John Capper, Gen Elles and Col Samuel D. Rockenbach, Chief of the American Tank Corps. It was hoped to assemble 100 tanks in France in July and to reach 300 monthly by September. To ensure full interchangeability of parts, all jigs for British and American use were made by a single firm, the Leeds Forge Company. It was discovered in March that the limits specified on the jig drawings were incorrect, leading to further delay.

  Stern’s relations were further deteriorating inside the Ministry and with the army. He continued to criticize Moore’s operations, driving Duckham to recommend his replacement. Duckham told Churchill in late February that he regretted ever suggesting Stern’s appointment as Commissioner. At the same time the War Office contemplated urging Churchill to place Chateauroux under American control as ‘any joint control must depend for smooth working on the personality of the Controllers, and is more liable to lead to friction and loss of efficiency than a single control’.13 Counsel’s opinion was sought on the risk of breaching the Anglo-American Agreement if Stern was dismissed. He confirmed there was no problem. Lord Derby chaired a meeting at the War Office on 8 March attended by Capper, Duckham and others at which it was decided ‘to give Colonel Stern another chance – if he again interferes the Ministry of Munitions will remove him’. In the meantime, Stern was to be stripped of his rank and uniform through demobilization ‘if possible’, and Capper and Duckham were to find means to limit his powers.14 The ensuing series of minutes on a War Office file would be hilarious if the matter were not so serious. Loss of uniform would unsettle Stern’s Allied partners and undermine his authority. Most of his senior staff were commissioned officers – as a civilian Stern would have no authority over them and his position would at once become untenable.