(This article first appeared in the June 2020 edition of Stand To! Number 118 pp.10-14.)

By 1916 both sides were seeking a solution to the stalemate which had developed on the Western Front as a result of the almost insuperable advantage enjoyed by garrisons in carefully constructed defences over enemy troops attacking across prepared fire zones. Both sides were resorting to increasingly heavy artillery bombardment to ‘soften up’ the opposing lines, although it was recognised that preliminary bombardments advertised the locality of the imminent attack. Both sides were also exploring ways in which opposing defences could be overcome without the expense – and the loss of the ‘surprise’ factor – involved in an artillery barrage. The solution to the problem favoured by the British (and the French) was the tank. That favoured by the Germans was cheaper and more immediately available, but, like a moral precept, simultaneously very simple and very difficult: nothing more, and nothing less than a complete rethink of infantry tactics. 

Mixed arms assault units

The keys to the German Army’s new tactics were one weapon that was much less favoured by the British and the French – the flamethrower – and two weapons that were available only from captured stocks of Russian and British equipment – the Russian model 1902 76.2mm gun and the Lewis Gun. The latter – an easily portable machine gun – was, according to British thinking, not to be included in the first wave of attacks but was to be primarily regarded as a defensive weapon, with a role in offensive operations confined to ‘the covering of the consolidation.... of the newly captured position, and the beating off of counter attacks.’ (1) 

A German soldier practises clearing a trench with a flamethrower during a training session near Sedan in May 1917. IWM Q23759

Bernhard Reddemann, a fire service official and army reservist, had been experimenting with flame–throwing devices in Germany before the war and obtained permission to establish a 48–man flamethrower unit partly consisting of professional firemen and under the leadership of a Feuerwehrfeldwebel (approximately equivalent to a fire brigade station commander). This unit first saw action near Verdun on 26 February 1915. After serious losses had been suffered advancing with the cumbersome flamethrower over open ground, thought was given to the development of tactics involving more support from infantry units.(2) Meanwhile a regular officer, Hauptmann Willy Rohr had been developing the concept of mixed–arms assault units from a different direction — ie by adding flamethrower teams to infantry units rather than by adding infantry to flamethrower teams — and by September 1915 was operating a unit with six machine guns, four grenade throwers and six flamethrowers.(3) 

Stormtroopers practising the attack from a trench supported by flamethrowers. Near Sedan, May 1917. IWM Q45341

On 23 October 1916, the German Army High Command (Oberste Heeresleitung) ordered instruction along the lines of Rohr’s methods throughout the army, with Rohr’s own unit Sturmbataillon Rohr as a training unit for Army Group Imperial Crown Prince, (4) By the beginning of 1917 there were 16 fully equipped Sturmbataillone (storm battalions) with machine guns – usually Lewis guns captured from the British – light and heavy mortars, explosive packs and satchels, grenade throwers and flamethrowers, the latter mostly operated by teams seconded from Reddemann’s unit, which had been designated Garde–Reserve–Pionier–Regiment. Later most Sturmbataillone also acquired a battery of four to six model 1902 76.2mm ‘infantry guns’, captured from the Russians and fitted with shortened barrels.(5) 


The precise role of the Sturmbataillone within the German Army has been the subject of some confusion. First of all, each Sturmbataillon had its own peculiarities of internal organisation and equipment. Secondly, they usually operated split up into separate sub–units, often under the command of a Leutnant. Reddemann’s Garde–Reserve–Pionier–Regiment meanwhile continued its independent existence, with its seven constituent companies distributed as required. As a report from the 45. Reserve– Division testified, attacks with flamethrowers were more effective than those with grenades: ‘Where support by flamethrowers was lacking, the conquest of machine gun nests by means of hand grenades generally took more time and cost more casualties.’(6) The Sturmbataillone should not, however, be seen as spearhead troops: essentially they were specialist troops assigned to attack key points, or called forward when an advance had been held up at a particular location. The one–ton 76.2mm guns and the quarter–ton 76mm mortars deployed by Sturmbataillone were obviously far too heavy to be at the forefront of attacks but were readied close enough to initial jumping–off points to be available to tackle defensive positions that had successfully resisted lighter weapons. Ordinary infantry battalions had also been trained in storm–troop tactics, partly by teams from Sturmbataillone, and by the spring of 1918 many Army commanders in the two German Army Groups on the Western Front judged the training role of the Sturmbataillone as no longer required, the general principles of the new tactics having been promulgated in an official General Staff pamphlet, Der Angriff im Stellungskrieg, by a staff officer named Hermann Geyer (later a corps commander in 1940).(7) Troops not in Sturmbataillone employing storm–troop tactics were known as Stosstrupps – shock troops ̶ a term apparently coined by Reddemann.(8) 

Hay Flame gun in action; operator wearing protective clothing. IWM Q64218

Whatever the precise role of the Sturmbataillone in Stosstrupp–type attacks, the point to emphasise is that the Sturmbataillone (if not the Stosstrupps of ordinary infantry regiments) were mixed–arms units, and also that a principal feature of their methods was the employment of a weapon, the Lewis gun, captured from the British, but used by the British primarily in a defensive role.(9) The Germans also made far more use of the flamethrower than the British did. In 1916 the Special Brigade of the British Expeditionary Force (BEF) had experimented with large flamethrowers pushed forward through underground galleries but had found this tactic far too cumbersome.(10) Some man–portable devices were being readied for 1919 when the Armistice came.(11) Perhaps as many as 15 Hay Flame Guns, each weighing 66lb and with a range of 22 yards, were used by the naval landing party, along with two larger shipboard devices, in the famous raid on Zeebrugge on 23 April 1918, though there may be more than symbolic significance in the fact that the officer in charge of the flamethrowers was killed attacking a German machine–gun post with his bare hands.(12) The interesting point, however, is not the comparative lack of interest in flamethrowers on the British side, but the army’s failure to recognise the utility of their own existing weapon, the Lewis gun, as a weapon to be used in combination with other weapons in an attack. 

Tactical potential

It may even be that the development of the tank – the war years’ most striking invention – contributed to the failure to develop the tactical potential of the Lewis gun to the full. The tank seemed to offer the best hope of breaking the deadlock on the Western Front and the need to train and exercise tank personnel in conjunction with infantry was certainly recognised; but one cannot but be struck by how inadequate the joint training was. In June 1918, for example, 11/Battalion of the 1st Tank Brigade shared a Field Day with the 4th Canadian Division, and 12/Battalion trained for four days with the 74th Division.(13) Even this may have been exceptional: the staff of 5th Division blamed ‘not entirely satisfactory’ co–operation between tanks and infantry during operations in late August 1918 partly on the fact ‘that the 5th Division had never had an opportunity of training with Tanks.’(14) A staff memo on ‘Co– operation of Tanks with Other Arms’ noted that:

 ‘Except for the August 31st [1918] battle, when Infantry Officers were met the day previous to the ‘show’, all the liaison has been in the hour or two hours immediately preceding the attack, after tanks have reached their starting points and the tank commander has found out the infantry officers with whom he is working.’(15) 

Major General C J Deverell, commanding 3rd Division − later Field Marshal Sir Cyril Deverell, CIGS 1936–7 − thought ‘the want of training and actual experience with Tanks was very much felt both by the Infantry and the Tanks.’(16) Eighteen months previously a joint Anglo–French conference at the War Office in London had concluded that:

‘The infantry Division is the chief fighting unit . . . the action of the tanks must be co–ordinated with that of the divisions. The tank units are generally at the disposal of the Army Commander; but the tank unit operating with any division of this Army is, during the attack of this division, under the orders of the divisional commanders. British experience indicates that it is better to allot the tanks to a Corps front, controlled by the Corps, than to have a detachment of tanks definitely allocated to a division.’(17)

But while much was said about assigning tanks to corps between battles, and to divisions during battles, all that was said about co–operation between tanks and infantrymen under fire was that in an attack on a 1,600m divisional front an infantry company should be attached to each of two tank units ‘to assist the tanks in crossing obstacles, trenches, etc.’(18) There seemed to be no notion that the tanks would benefit from infantry in close support once they were in contact with enemy troops armed with anti– tank weapons and, consequently, there were frequent instances later on of tanks running too far ahead of advancing infantry and being captured.(19) 

According to a memorandum by the General Staff of the Tank Corps in January 1918, tanks would attack in sections of three tanks, one of which would be the ‘Advanced Tank’ and the other two ‘Main Body Tanks’, and it stipulated that: ‘Normally infantry will only follow the Main Body Tanks’. It also advised that ‘Once the Infantry have been placed in the enemy’s trenches by the Tanks, they must depend on themselves to do all the minor cleaning up work.’ In other words, the tanks would take positions and the infantry merely follow and occupy.(20) 

Sketchy notions

The British Army’s notions of what constituted adequate liaison between different arms is illustrated by attempts made in 1918 to co– ordinate the action of tanks and aircraft. Not too much should be made of this: even with much superior radios such co–ordination was still proving difficult more than 50 years later. Nevertheless, the week or so in which, ‘several demonstrations with aeroplanes [were] carried out. Tank Officers going in aeroplanes and Flying Officers in Tanks’, then followed up, just two weeks later, by ‘interchange of two 3rd. Battalion Officers and two Officers of ‘A’ Flight, No.8 Squadron, RAF, for two days’ visit’, as recorded in the war diary of 3/Battalion, 3rd Brigade Tank Corps for July 1918, suggests only the sketchiest notions of what might constitute combined training.(21)

Much has been made in recent years of the British Army’s ‘learning curve’ during the period August 1914 to November 1918, that is, the supposed process whereby a small peace–time army was expanded into a huge military machine schooled in the techniques required for victory.(22) 

It would be an error, however, to suppose that, even compared to the Americans, the British Army of the late summer of 1918 had been honed, hardened and brought to an adamantine edge of keenness and tactical competence by years of gruelling combat, as one subaltern’s explanation of the circumstances of his being captured in September 1918 makes clear:

‘We attacked & a fair proportion of us succeeded in gaining the objective. An immediate counter attack, however, found the majority of our rifles useless as they were covered in mud. The men were new recruits almost to a man & those not killed or wounded by enemy fire failed to stick it & attempted to get back to our lines . . . .’(23) 

In fact it is difficult to believe that ‘the whole BEF had by 1918 attained a high level of professionalism,’ as one authority claims, when one reads Second Lieutenant Cyril Victor Longland’s account of his capture on 8 August 1918, the day General Ludendorff described as ‘the Black Day of the German Army.’ As second in command of A Company, 2/Lancashire Fusiliers, but presumably following the orders of his company commander Longland took the acting company sergeant major and three sergeants forward to reconnoitre the corner of a copse and, while trying to find it, was captured along with the sergeant major by a German patrol: Longland may have been inexperienced enough to need the advice of all four sergeants in his company but sending them out all together on a reconnaissance and leaving only the corporals with the main body of troops was definitely not ‘professionalism’.(24) There were numerous other instances during the final weeks of the war of men pushing forward till they found themselves cut off from the rest of their battalion and forced to surrender by German troops who surrounded them in overwhelming numbers.(25)

Infantry practising an attack behind a smoke screen and a tank. Photograph taken at Sautricourt, 12 July 1918. IWM Q9819

Quality of learning

It had not been like that with the German Army during the spring offensives of that year. The fact was that German troops, from being demonstrably less well schooled than the British in 1914, thereafter edged ahead and maintained a distinct tactical superiority till at least August 1918; this means that the most interesting ‘learning curve’ exhibited in these years was not that of the British Army, but the one demonstrated by the German Army. Thinking in terms of ‘learning curves’ may however distract from the key point that what mattered was not how fast the different armies learnt but what they learnt. 

In 1890 the American naval theorist A T Mahan wrote:

‘[One] will observe that changes in Tactics have not only taken place after changes in weapons, which necessarily is the case, but that the interval between such changes has been unduly long. This doubtless arises from the fact that an improvement of weapons is due to the energy of one or two men while changes in tactics have to overcome the inertia of a conservative class.’(26) 

This analysis seems hardly applicable to the contrast between the German and British armies in tactical innovation during the Great War. The German Army circa 1914–1916 cannot be said to be less conservative, less hide–bound, than the British Army. The process whereby middle–ranking officers like Bernhard Reddemann and Willy Rohr were able to have their ideas taken up and promoted higher up the chain of command had its counterpart in the British Army.(27) In the German Army, however, it was unusual for junior officers to receive more than a single step in rank during the four years of fighting, whereas at least 60 battalions in the British Army were commanded by men who had been civilians with no military experience in July 1914, and another dozen by officers who had been commissioned from the ranks after long service in the peacetime army as NCOs; at least four front line brigades were commanded in 1918 by men who had had no army experience before the war.(28) Two of Haig’s most successful corps commanders were colonial militiamen; the principal General Staff Officer of the elite Guards Division was a pre–war civilian whose occupation was that of magazine editor.(29) 

One can multiply such instances, all of which suggest a much greater openness and adaptability in the British Army when compared to the German.

Edward William Macleay Grigg, 1st Baron Altrincham by Walter Stoneman The only pre–war civilian to become the GSO1 of a division during the Great War. Bromide print, 1921 © BY NC ND National Portrait Gallery, London


There was, however, one feature of the German Army that, though related to its more hierarchical, more caste–bound characteristics, gave it an advantage over the British. Before the war Germany had conscription, Britain did not. The annual influx into the German Army of thousands of young conscripts had fostered the notion that it was perfectly feasible to train up inexperienced recruits to a fixed standard in a determinate period of time. Senior British officers on the other hand never quite believed in the ability of a citizen army to come up to even a barely adequate standard:

‘We must remember that owing to a large expansion of our army and the heavy casualties in experienced officers, the officers and troops generally do not now possess that military knowledge arising from a long and high state of training which enables them to act instinctively and promptly on sound lines in unexpected situations.’(30) 

Such views had the effect of becoming self–fulfilling prophecies. The Sturmbataillone and regimental Stosstrupps of the German Army may be seen as an application of the principle of initiative in junior leaders; the developing practice of the British Army on the Western Front was not merely to stifle initiative but to sidestep the whole question of the role of initiative.(31) The ‘Report of the Conclusions Reached at a Conference on the Tactical Employment of Tanks, held on 4 March 1917, at the War Office’ is indicative of this. As already mentioned attention was given to the question of which level of headquarters should control the available tanks, and it was decided that during the attack ‘the tank unit operating with any division’ should be temporarily under the orders of the division, but that previous to an attack tanks should be controlled one level higher up, by a corps headquarters. It was determined that: 

‘The tanks will not, under normal circumstances, come into action until the infantry advance is checked by the enemy strong points . . . The tanks then do the work which the artillery has been unable to accomplish. They move forward in lines of columns and then deploy on coming into action . . . .’(32) 

Since the divisional headquarters temporarily in control of the tanks would be several miles behind the front line, and dependent for its information on intermediate brigade headquarters that might well have lost contact with attacking battalions during a contested advance, there would obviously be a question of some sort of time lapse – several hours? – between infantry being checked by a strong point and the tanks coming up. But there was a good ‘professional’ reason for the need to specify that the tanks move forward in columns and only deploy when coming into action. The tank unit commander might seem to be in a better position to decide formation and timing on the basis of the ground conditions he encountered: but in most cases he would be an officer commissioned since the start of the war, not a professional soldier, and therefore deemed to lack the educated judgement of the real professionals at command and staff level whose job it was to decide the details of complex operations: he needed trained officers in the rear to do his thinking for him. In effect the question of what might actually happen when tanks were sent forward was subordinated to organisational arrangements at a higher level: we have already noted that no thought had been given to infantry giving direct combat support to tanks. Unlike tanks, German storm troopers could be deployed in front line trenches at the beginning of an attack, but the key point is that the decision to commit them to action was generally made at battalion or regiment–level (German regiment–level corresponding with brigade–level in the British Army), or even at company level, and did not have to be referred to a headquarters located back out of artillery range. But even 25 years later, the lack of scope given to junior commanders to exercise their own initiative was reported by German officers as characteristic of British methods.(33) 

Leaving behind it the simplicities and improvisations of colonial warfare the British Army in the Great War had grown into a vast over–compartmentalised hierarchy, with a bureaucratic mind–set that functioned well enough as regards the routines of supply and replenishment but was perhaps less effective when it came to defeating the Germans on the Western Front. Circumscribing the initiative of junior officers while failing to think through the problem of co–ordinating the different weapon arms of the army was the result of a failure to see that though integration of arms needed to be prepared in advance, with memoranda by staff officers and inter–departmental consultations – in effect, initiatives at bureaucratic level – it was subordinate commanders acting in real time in the firing line who actually had to make it happen. 


(1) ‘Lewis Gunner’, Tactical Handling of Lewis Guns: With Notes on Instruction, etc., (London: 1918), pp.15–16, 33.

(2) [Bernhard] Reddemann, Geschichte der deutschen Flammenwerfer–Truppe, (Berlin: c. 1933), pp.11, 17–18.

(3) Helmuth Gruss, Die deutschen Sturmbataillone im Weltkrieg: Aufbau und Verwendung, (Berlin: 1939), p.21. English language texts on the subject of German storm troops, generally recycling a somewhat limited range of original German material, include Ian Drury and Gerry Embleton, German Stormtrooper: 1914–1918, (London: 1995), Bruce I Gudmundsson, Stormtroop Tactics: Innovation in the German Army, 1914– 1918, (Westport Connecticut: 1995) and Stephen Bull, German Assault Troops of the First World War: Stosstrupptaktik: the First Storm Troopers, (Stroud: 2007).

(4) Gruss, Die deutschen Sturmbataillone p.60 and p.156 Anlage Nr. 7. Jonathan Boff, Winning and Losing on the Western Front: the British Third Army and the Defeat of Germany in 1918, (Cambridge:2012), pp.153–9 emphasise the diversity of the tactical methods employed by different divisions of the British Third Army in the final three months of the war: but this was nearly two years after the Germans had implemented a standardised tactical reform along Stosstrupp lines throughout their armies on the Western Front. In his most recent book – Haig’s Enemy: Crown Prince Rupprecht and Germany’s War on the Western Front, (Oxford: 2018), p.279 – Boff acknowledges the overall tactical shortcomings of the British Army in 1918.

(5) Ibid, pp.73, 103, and [Christof] Theune, Flammenwerfer und Sturmtruppen, (Berlin: nd [c1920]), pp.119–121. Cf. Drury and Embleton, Stormtrooper, p.26. See also Georg Grosskopf, Sturmbataillon Nr. 1(einschl. bayr. Infanterie– Geschütz– Batterie Nr. 2) im Weltkrieg 1916/18, (Landsberg: 1938).

(6) Reddemann, Geschichte der deutschen Flammenwerfer–Truppe, pp.49–50.

(7) Robert T Foley, ‘A Case Study in Horizontal Military Innovation: the German Army 1916–1918’, Journal of Strategic Studies, vol. 35 (2012), pp.799–827, ignores the institutional arrangements made to disseminate storm–troop tactics throughout the German Army and suggests that it is an example of ‘horizontal military innovation, rather than a top–down, ie ‘vertical’ process.

(8) Martin Samuels, Doctrine and Dogma: German and British Infantry Tactics in the First World War (London: 1992), p.50. The German word Trupp, plural Trupps, refers to a small body of soldiers, the word Truppe, plural Truppen, with the same derivation, refers to soldiers in a larger formation or a branch of the army. The memo ‘Fortsetzung zu den Erfahrungen aus dem Angriffskämpfen bei Cambrai für die “Angriffsschlacht”,’ dated 24 December 1917, in Kronprinz Rupprecht von Bayern, Mein Kriegstagebuch, ed. Eugen von Frauenholz, 8 vols., (Munich: 1929), vol. 3, pp.207–222 refers to Stosstrupps not to Sturmbataillone.

(9) See Ernst Jünger’s descriptions of action in In Stahlgewittern : aus dem Tagebuch eines Stosstruppführers and Das Wäldchen 125 – The Storm of Steel, (London: 1929), p.252 foll., Copse 125: a Chronicle from the Trench Warfare of 1918, (London: 1930), p.250. foll., which makes it clear that Stosstrupp leaders in the regiments were reliant on the same rifle and grenade combination as attacking British infantry. For an optimistic account of the tactics of Lewis gun, rifle, grenade and rifle grenade sections in the British Army following the issue of SS143, ‘Instructions for Training of Platoons for Offensive Action’ see Chris McCarthy, ‘Queen of the Battlefield: the Development of Command, Organisation and Tactics in the British Infantry Battalion during the Great War’ in Gary Sheffield and Dan Todman eds. Command and Control on the Western Front: the British Army’s Experience 1914–1918 (Stroud:2004), pp.173–93 at pp.181–2.

(10) C H Foulkes, ‘Gas!’The Story of the Special Brigade, (Edinburgh: 1934), p.111.

(11) TNA, MUN 5/198/1660/11 and Donald Banks, Flame Over Britain: a Personal Narrative of Petroleum Warfare, London: [c1946], p.62.

(12) http://www.1914–1918.invisionzone. com/118787–naval–flame–throwers cf. Paul Kendall, The Zeebrugge Raid: ‘the Finest Feat of Arms’, (Brimscombe Port: 2009), p.224–5

(13) TNA, WO 95/99, War Diary of 1st Tank Brigade. WO 158/803, ‘Tanks Corps Training: Infantry and Tank Co–operation’ which dealt with the issue of co–operation in some detail but neglected the issue of armour’s dependence on infantry to suppress enemy anti–tank weapons.

(14) TNA, WO 95/1516, War Diary, General Staff, 5th Division, ‘Resume of Reports of Operation of 95th, 15th and 13th Infantry Brigades’, pp.17–19 Appendix, ‘Some Lessons from the Operations; August 21st to September 4th’ at pp.17–18.

(15) TNA, WO 158/855, ‘Co–operation of Tanks with Other Arms,’ [1918] p.1.

(16) TNA, WO 95/1381, War Diary, General Staff, 3rd Division, ‘Report on Operations by 3rd Division: 1st Phase − August 21st to 24th 1918’ Appendix, ‘Lessons’ p.2, cf. Boff, Winning and Losing, pp.144–5.

(17) TNA, WO 158/841, ‘Report of the Conclusions Reached at a Conference on the Tactical Employment of Tanks, Held on 4th March 1917, At the War Office’, paragraph 4 [p.3], paragraph 8 sect. 2 [p. 6].

(18) Ibid, para.4 [p. 4].

(19) Cf. A D Harvey, ‘First Hand Accounts of Tank Action in WW1,’ Royal Tank Regiment Journal, vol. 95, no. 800, (Winter, 2013), pp.54–55.

(20) TNA, WO 158/803, ‘Tank Corps Training: Infantry and Tank Co–operation... Issued to TANK BRIGADES by GS TANK CORPS’ 27th January 1918, p.7 (para. 8) and p.9 (para. 14) – paras. 10 and 16 in later, undated version of same text in same file.

(21) TNA, WO 95/106, 5–13 July and 29–30 July, 1918. With regard to the exchange of two officers from each formation: according to 3rd Battalion, 3rd Brigade’s war diary for 30 June the battalion then had 96 officers on strength, and A Flight, No. 8 Squadron probably had a fifth of that number. For a probably optimistic account of the results see TNA, AIR 1/1074/204/5/1665, ‘Notes on the Co–operation of Aeroplanes and Tanks in Recent Operations’.

(22) For a discussion and elaboration of the ‘learning curve’ argument see Boff, Winning and Losing on the Western Front pp.11–12, and see also Fox, Learning to Fight, pp.7–9. There has never been any talk of a ‘learning curve’ with regard to the Battles of Alma and Inkerman, fought and won six and 62 days respectively after the British had landed in the Crimea in their first continental campaign for almost 40 years.

(23) TNA, WO 374/57499, explanation of the circumstances of his capture on 10 September 1918 by Second Lieutenant F O Rideout. His battalion, 1/East Yorkshire Regiment, suffered 186 casualties in this action, mostly PoW. This was less than nine weeks before the Armistice.

(24) Brian Bond, Britain’s Two World Wars against Germany: Myth, Memory and the Distortions of Hindsight (Cambridge:2014), p.214: cf. Boff, Haig’s Enemy, p.279. Note that Professor Bond does not propose the BEF’s ‘high level of professionalism’ in 1918 as a myth or a distortion of hindsight. For Cyril Victor Longland see TNA, WO 374/42838 cf. Erich Ludendorff, Kriegserinnerungen (Berlin, 1919), p. 547. Exponents of the ‘high level of professionalism’ interpretation of the British Army’s performance in the late summer and autumn of 1918 often cite the influence of instruction manuals issued by GHQ such as SS135 and SS148. Fox, Learning to Fight, p.64 mentions the view of some officers that these were mere ‘Bumf’, or tended to lead to confusion

(25) See for Example TNA, WO 339/1704, WO 339/101438, WO 339/30889, WO 339/32126, WO 374 /2878, WO 374/31667, WO 374/51636.

(26) A T Mahan, The Influence of Sea Power on History 1660–1783, (London: 1890), pp.9–10.

(27) See for example C D Baker–Carr, From Chauffeur to Brigadier, (London: 1930), pp.102–7, 127–8, 134–5, Edmonds, Military Operations: 1915 vol. 1, p. 31 and TNA, WO 32/11239. See also TNA, WO 158/831 ‘The Necessity of Machine Gun Destroyers,’ June 1915. Robert T Foley,’ “Dumb Donkeys or Cunning Foxes?” Learning in the British and German Armies during the Great War’, International Affairs, vol. 90 (2014), pp.279–298 argues that in contrast to the British Army the more ‘professional’ German Army made little use of outside expertise ̶ which overlooks Fritz Haber’s role in gas warfare, Bernhard Reddemann’s role in the adoption of the flamethrower, the young Dutchman Anton Fokker’s role in fighter design, and the German Army’s use of captured Lewis guns and tanks, etc. Jonathan Boff, ‘Culture Clash: British and German Military Innovation at War, 1914 – 18’, British Army Review no. 174 (Winter 2019) is the latest to claim that the German Army was more top down than the British: but he overstates his case and adduces little evidence.

(28) A D Harvey, ‘A Good War: Wartime Officers Who Rose to Command Level in the First World War,’ RUSI Journal vol. 153, no. 2 (April 2008) pp.76–80 and P E Hodgkinson, British Infantry Commanders in the First World War, (Farnham, Ashgate: 2015), p.159: the four brigadier generals who were civilians in July 1914 were Arthur Asquith, George Gater, George Rollo and Spencer Vaughan Weston.

(29) The two colonial lieutenant generals were, of course, Sir John Monash, a civil engineer, and Sir Arthur Currie, whose background was in insurance and real estate. The GSO1 of the Guards Division was Edward Grigg, formerly of The Times, editor of The Round Table, described by Winston Churchill as ‘one of the best officers the new army produced.’ (TNA, WO 339/31633 Churchill to Military Secretary 27 June 1919).

(30) Edmonds, Military Operations, 1916 Appendices volume, pp.131–147, appendix 18, ‘Fourth Army Tactical Notes’ p.131.

(31) The classic discussion of this issue is in Martin Samuels’, Doctrine and Dogma, pp138–148.

(32) TNA, WO 158/841 ‘Report of the Conclusions. . . 4th March 1917. . . .’ para 4 [p. 4], cf. Samuels, Doctrine and Dogma, p.139: ‘This practise of controlling “three down”, that is, a headquarters determining the actions of units three levels down the chain of command, appears to have been standard [in the British Army]’

(33) See the remarks of Fritz Bayerlein, Rommel’s chief of staff, in B H Liddell– Hart ed. The Rommel Papers (London: 1953) p.184 and cf. C P Stacey, Official History of the Canadian Army in the Second World War, 3 vols., (Ottawa: 1955–66), vol. 1, p.391.