‘After Amiens’: Technology and Tactics in the B.E.F during the Advance to Victory, August–November 1918 Part I


On 8 August 1918, the British Expeditionary Force (BEF), together with their French allies, launched a major offensive which finally accomplished an overwhelming physical and psychological victory over the German forces on the Western Front. In this, the Battle of Amiens, the BEF successfully combined infantry, artillery, aircraft and armour in an orchestrated operation, applying to good effect the lessons of four years’ fighting. The BEF’s attack broke its opponent’s defences and penetrated to a depth of up to eight miles, inflicting heavy casualties and causing significant loss of matériel that the Germans could ill-afford.

Experience had taught the high command that if casualties in any offensive were to be minimized, it was essential for attacks to be made on a wide front by infantry with good artillery support and close co-operation by tanks. By the end of the fourth day of the battle, these conditions no longer prevailed at Amiens, and the attack was rightly closed down.

In switching their attention to other parts of the front, the British commanders then faced the necessity of maintaining operational tempo while ensuring that the lessons of Amiens were disseminated and applied in subsequent operations. This was especially true in connection with the tactical employment of armour. During the following months and until the end of the war, and against an increasingly popular perception among all ranks that tanks offered a means to reduce casualties in the offensive, and consequent demands to use them in operations, senior commanders endeavoured to manage the available assets as effectively as possible. Their responses combined innovation and pragmatism but were frequently conditioned by the realities of logistical and operational considerations as well as environmental factors such as terrain and the weather.

This is how the BEF’s assimilation of tanks into the so-called all-arms battle held up under the pressure of events during the last months of the war. In doing so, it stresses certain key factors that defined the environment in which this task was undertaken and reviews the Tank Corps’ own analyses of its performance after Amiens.

Operational tempo: A two-edged sword

The generals’ desire for maintenance of operational tempo was significant in its effect on tank use on the Western Front in the latter half of 1918. While striving to keep the Germans on the back foot after Amiens, the high command created an environment that was counter-productive to the reduction of tank-infantry co-operation to an effective drill honed by practice before battle. The almost continuous nature of operations worked to preclude extended periods of training. After the Cambrai Battle in late 1917 which had ‘defined the standard of training and planning required for truly successful co-operation between tanks and other arms’, any opportunity for infantry and armour to rehearse together was recognized as a significant factor in successful tank-infantry co-operation. The publication of S.S.204 Infantry and Tank Co-Operation and Training in March 1918 had represented a concomitant attempt by General Headquarters (GHQ) to offer practical direction on how this training might be carried out. However, in reality, few such opportunities arose after Amiens.

Nevertheless, the high command continued to recommend the value of training in its doctrinal publications. Yet S.S. 214 Tanks and their Employment in Co-operation with Other Arms, published in August 1918, could offer little more than suggestions for army and corps school instructors to be attached to tank units to train junior officers and NCOs in the hope of gradually disseminating knowledge and ideas. This was surprising in what was, in other respects, a valuable document full of practical and sound advice based on combat experience.

The significance of this lack of training opportunities and the value placed by tank unit commanders on training in co-operation is illustrated by the remarks of Lieutenant-Colonel Harry Johnson of 1st Tank Battalion in reviewing operations with 30th American Division at Bellicourt on 29 September 1918: ‘All the old lessons have been again learned [sic] . . . One suggestion only will cover all the points I desire to make; greater facilities for training in the attack are essential to secure greater efficiency.’

Operations involving armour in late 1918 were also constrained by the actual numbers of tanks available to the BEF for use in the fighting which, in turn, were complicated by the closely-linked issues of tank supply and salvage. These matters have been given more detailed consideration elsewhere but, essentially, because tank wastage in mid- to late 1918 was many times greater than was anticipated, stocks of spares were severely depleted and supplies of replacements were inadequate. As a consequence, men usually engaged on repairs were instead salvaging parts from abandoned tanks. Meanwhile, tanks that might otherwise have returned to action after suffering battle damage were cannibalized to compensate for the spares shortage. In addition, the desire for a rapid tempo in operations involved frequent switching of attacks from one sector of the front to another. This presented insurmountable logistical problems (given the available transport and technologies) when assembling tanks for use en masse in each of these widely-separated attacks.

The approach adopted by GHQ and the Tank Corps to the deployment of available armour was entirely sensible, given the logistical and operational circumstances. They allotted a tank brigade (consisting of two or more battalions) to each of the three armies undertaking the major offensive operations in France, that is, First, Third and Fourth. This meant that the somewhat rigidly established logistical support arrangements needed to maintain the tanks of each unit in the field could be built on and expanded as necessary and not severely disrupted as they would surely otherwise have been.

Logistics were especially important in defining the circumstances for tank use in the final advance. It was only in early 1918 that a sophisticated and relatively robust infrastructure for the equipment, supply and reinforcement of the Tank Corps in France and Belgium was fully established. There was now a vast array of workshops, supply and equipment dumps and training bases along the Ternoise valley north-west of Arras. In addition, the introduction in February of the first roll-on, roll-off cross-Channel train ferry service capable of carrying numbers of tanks loaded on rail flat cars, gave significant advantages to tank units at the front. Yet, despite these important developments, the very nature of the fighting in late 1918 with an advance continuously away from the established bases and supply dumps presented a host of logistical challenges that undoubtedly influenced the numbers of available tanks in some operations.

This point should, however, be treated with a degree of caution. There was a significant period between 4 and 18 September 1918 in which there was an absence of tank operations. Although this period offered benefits in the planning and execution of attacks involving armour, it did not witness a significant rise in available tank numbers for the next phase of operations. Indeed, Brigadier-General Anthony Courage commanding 5 Tank Brigade was especially vocal in criticizing a relatively small number of tanks from one battalion being spread too thinly for action on 18 September. Clearly, other factors were at work in connection with the tanks available for action from mid- to late September.

Tanks were usually deployed in larger set piece battles because of their relative mechanical fragility (especially in terms of durability of their tracks) and their heavy logistical requirements which basically demanded the use of railways for their transportation and supply. Moreover, limiting the distance tanks were required to travel under their own power and on their own tracks before being committed in action increased the likelihood of both them and their crews arriving at the point required in good enough physical condition to be of use in combat.

It is also important to recognize that in the operations in the last months of the war the set piece battles for which tanks were most suited by logistical considerations took place over a greatly reduced time period between initial planning and execution. As Andy Simpson points out, ‘Even though the set piece operations, such as the attacks on the Hindenburg Line, look ponderous, they were launched at a far greater speed than their equivalents in 1917.’ In one example, mentioned by Simpson, the planning and execution of an attack by units of IV Corps on 23 August took a total time of 19 hours from the start of the initial conference involving divisional and corps commanders to zero.

While this reduced cycle was an indication of how efficient the staff work of the BEF had become by 1918 and how strong was its logistical base, it gave rise to unexpected problems for the tank arm. Since the planning and execution of operations was now considerably devolved, with corps and divisions making arrangements with tank brigades, this should, in theory, have produced a closer mutual understanding between the various arms undertaking an attack. However, when combined with the desire demonstrated by all corps commanders to have some tanks (even in small numbers) allotted to them for any assault they were about to make, the end product instead was a return to the much-derided ‘penny-packeting’ of tanks in sections or companies among each of the attacking corps.

In 1916, given the paucity of tanks available and a desire to make use of what there was, there had been little option for army and corps commanders other than to commit tanks in ‘penny-packets’. Indeed, there had been a number of successful occasions of their use in this manner in the latter part of the Somme battle in that year. However, by 1918, a more robust Tank Corps defence was mounted against this approach, which disrupted command and control arrangements. The response from GHQ, in the person of Lieutenant-General the Hon. Sir Herbert Lawrence, was to warn of the consequences of penny-packeting but, once more, to be pragmatic in accepting the need if operational circumstances demanded. In a memorandum of 1 September, intended to address ‘the unfamiliarity of many Divisional and Brigade Commanders with the functions and limitations of Tanks’ which had produced disproportionate tank losses to the results attained, Lawrence wrote:

The units and formations of the Tank Corps have been so organized in order to facilitate their handling both tactically and administratively. This organization has been frequently departed from in order to meet local conditions. Although at times this may be unavoidable it should be borne in mind that such a departure from the normal organization must result in a loss of fighting efficiency.

This pragmatic approach was in keeping with Lawrence’s intention not to manage the detail of tank operations but more appropriately to ensure a defined general framework for the conduct of such operations.

Inevitably, given the continuing requests for tanks to be made available for operations, the splitting up of tank battalions and companies continued and, equally inevitably, produced its own problems. A complaint from one tank battalion commander after operations on 18 September that ‘The time for preparation was insufficient for proper co-operation with three corps’ did not prevent this distribution of tanks from one unit among several corps being repeated on 8 October 1918 when those from two battalions were split across four corps and eight divisions.

The demand for tanks in the fighting of late 1918 in combination with the logistical constraints on their availability and further compounded by tank casualties from action, necessitated another pragmatic response from both the tank arm and the higher echelons of the BEF. While both continued to expound the tactical principles associated with each type in use, the realities of the fighting prompted occasions when any tanks of whatever type were what were demanded and used – especially on the second and subsequent days of a major offensive. This was particularly the case with the Medium A ‘Whippets’. Notionally ‘intended to exploit success when the enemy’s line has been broken’, they were sometimes used in direct contact with the infantry.

Notwithstanding these issues, significant tactical successes were achieved in a variety of circumstances. In common with the units of all arms, the Tank Corps was industrious in post-battle analysis and used the results of this analysis to inform tactics and technology for future operations. From these after-action reports it is instructive to see how the tank arm regarded its own performance in the fighting from late August 1918 to the war’s end.

The Royal Air Force and the Tank Corps

One undoubted area of significant progress and success (and perceived as such by both the Tank Corps and Royal Air Force) was in the use of RAF aircraft to assist tanks in offensive operations. In January 1918, Brigadier-General Hugh Elles, commander of the Tank Corps in France, had asked Major-General Sir Hugh Trenchard for Royal Flying Corps assistance for tanks in offensive operations. Elles’ principal point was that when the tanks got ahead of the infantry and beyond, or at the extreme range of, the British artillery fire, they were likely to encounter German anti-tank guns at point blank range as happened at Flesquières during the Cambrai Battle. His request was for aircraft to give warning of such anti-tank guns and, if possible, to keep down their fire. The two general principles of Elles’ request encapsulated the areas in which tactics and technology were developed for aircraft to co-operate with tanks during the rest of the war and which, in the Final Advance, were being refined and conducted with considerable success.

The principal step forward in air co-operation after Amiens was the allocation of a second RAF squadron to the role of counter anti-tank work. No. 73 Squadron, which flew Sopwith Camels, was specifically assigned to the task of ground attacks on German field guns that might operate in an anti-tank role and first saw action in this capacity on 23 August. Between that date and 11 November, according to the squadron history, ‘continuous low-flying was carried out, and a total of 153,600 rounds were fired at, and 1,176 25-lb bombs were dropped on, ground targets. The record for one day being 25,000 rounds fired and 160 25-lb bombs dropped’. This ground-attack work complemented that of the existing tank co-operation squadron, No. 8, which flew Armstrong-Whitworth FK8 two-seater general purpose aircraft.

Once again, this decision to employ a second squadron arose from the actions of Elles who, on 16 August, had written to Major-General John Salmond, now commanding the RAF in France, regarding ‘the importance of having definite aircraft units detailed to engage anti-tank guns’. In suggesting that, ‘for the protection of the Tanks from the fire of hostile guns, a complete unit of fighting machines should be detailed to carry out this work and have no other mission’, Elles acknowledged that he might be perceived to ‘be making too great a demand on the RAF but I put it forward as I consider it to be of such importance if the Tanks are to be able to fulfil their tasks to the full’.

Typically, the role of the two squadrons in operations where they both participated was as follows. At least one machine from No. 8 Squadron was to be up during the attack to keep the co-operating tank brigade or battalion headquarters informed as to the advance of the tanks and any points where they were held up. Other aircraft from this squadron were to bomb and machine gun any German artillery which they witnessed firing at tanks, and keep under periodical machine gun fire all places where German guns were likely to be. Meanwhile, No. 73 Squadron machines actively roamed the attack front looking for likely locations for anti-tank guns to be sited. Major Trafford Leigh-Mallory, who had overall responsibility for the two RAF tank co-operation squadrons, had swiftly concluded that using such aircraft to send down Zone Calls for British artillery to open fire on the hostile batteries was out of the question since the anti-tank guns did not generally start to fire till the tanks were between 1,500 and 1,000 yards away. Consequently,

the only sound principle was obviously immediate action on the part of the aeroplane with bombs and machine guns, with a view to driving the German gunners from their guns until the Tanks had overrun the position.

This scheme to which the two squadrons worked derived first from a thorough post-action reconnaissance of the Amiens battlefield conducted by Leigh-Mallory and Brigadier-General Henry Karslake, the Tank Corps’ senior staff officer. The two examined the positions of German anti-tank guns during the various phases of the fighting. The principle then established was for the RAF squadrons and Tank Corps officers to discuss the location of anti-tank guns and draw up a map showing the places where the tanks would be most exposed, and the most likely places from which they could be fired at by anti-tank guns. It was soon found that the tanks did not require as much assistance at the commencement of a battle, when they were advancing under a barrage as they did later, when the artillery protection thinned and gradually diminished.

As the fighting went on, the airmen’s knowledge of the anti-tank guns increased and was greatly enhanced after 2 September with the capture of a document detailing the German methodology for allotting guns for anti-tank work and the kinds of positions they were to take up. According to Leigh-Mallory:

This gave us a great deal of valuable information, and greatly increased the use we could make of the existing Counter Battery Maps, which of course only lasted while we were fighting to break down the Trench System. The scheme worked in the following way: The pilot and observer copied the likely anti-tank positions on to their own maps, in the area over which they were going to fly, each machine only having about 2,000 yards of front to watch. They were to machine gun and bomb periodically all the likely places in their area, whether they were seen as active or not, and then when they actually did see a gun firing, to attack it with everything they had got. In this way, by looking at the right sort of places, a great number of anti-tank guns were spotted as soon as they opened fire. As time went on we could reckon on at least 50% of the places so marked actually being active.

RAF co-operation with the Tank Corps was only one strand of the very important ground-attack work the air arm undertook in 1918. But it was, without doubt, one of the great tactical successes of the last months of the war.