On 4 April 1918, Sir John Monash, then a major-general commanding the 3rd Australian Division, wrote:
A new British Division came into the line on my right flank, South of the Somme last night, and was this morning biffed out, making a bad break in the line, and exposing my right flank … Some of these Tommy Divisions are the absolute limit, and not worth the money it costs to put them into uniform. However, – I mustn’t let myself go. You can doubtless read between the lines of all I have written in the recent past – bad troops, bad staffs, bad commanders.1
These words were written at the time of the first German attack at Villers-Bretonneux. On 26 April, shortly after the second German attack on the town, he commented:
It was the same old story. My 9th Brigade had held securely, and kept the Bosch out of the town of Villers-Bretonneux for three weeks. They were then withdrawn for a rest on April 23rd, and the 8th British Division (regulars) took over the sector from them.
Naturally, on April 24th, the Bosch attacked … and biffed the Tommies out of the town. Late at night we had to organise a counter-attack. This was undertaken by 13th and 15th Brigades in the early hours of Anzac Day. They advanced 3,000 yards, in the dark, without Artillery support, completely restored the position, and captured over 1,000 prisoners … It was a fine performance.2
Both of these extracts, written during, or in the immediate aftermath of, the two battles of Villers-Bretonneux, are certainly less than complimentary to British troops and formations and they give no credit whatsoever to British units for their part in the defensive operations or counter-attacks which finally ensured the security of the town and, with it, the safety of Amiens. Even in his book The Australian Victories in France, published after the war, Monash is equally ungenerous and makes scarcely any references to the British units involved in the fighting for, and around, Villers-Bretonneux. The impression given by Monash is that the successful defence of the town was almost entirely due to Australian troops.
The evidence available in surviving letters, diaries and unpublished accounts indicates that a large proportion of the officers and men of the AIF shared Monash’s view of the poor quality and performance of many British troops in the face of the German spring offensives in 1918. Major-General Sir J. Talbot Hobbs – himself British-born – commanded the 5th Australian Division, whose 15th Brigade played such a key part in the decisive counter-attack at Villers-Bretonneux on the night of 24–25 April 1918. Towards the end of that month he recorded: ‘the conduct of some of the [British] troops through the ignorance neglect or I am almost tempted to say – but I won’t, I’ll say nervousness of their officers has had a very depressing effect on me and disgusted many of my officers and men’.3
Such sentiments were not confined to senior officers. Lieutenant Sydney Traill, of the 1st Battalion AIF, noted in his diary on 9 April: ‘Every day come fresh rumours about the Tommies. It seems that whole divisions retreated for miles without putting up any show of fight. The cold-footed hounds, it is enough to make one weep’. Traill wrote in a similar vein on 25 April, after the German capture of Mount Kemmel, observing: ‘The name of the Tommy stinks in a good many quarters now …’ .4 Lieutenant J.W. Axtens, of the 8th Australian Machine Gun Company – writing home to his parents in New South Wales on 2 May 1918 – pointed out that there was ‘a lot of strong feeling among our chaps against the Tommies … There are some very scathing jokes against them; this is one. A Tommy brigadier is reported to have overtaken a hare on the road towards Amiens and said savagely “Get out of the road you brute and give a man a chance who can run” …’.5 The prevalence of this type of comment was sufficient to persuade Sir William Birdwood, then still commanding the Australian Corps, to urge his senior officers on 30 April to do everything in their power to put a stop to ‘disparaging comparisons’ between British and Dominion troops. ‘The Dominion soldier’, he emphasised, ‘has so established his merits that depreciation of his kith and kin is not necessary for the acknowledgement of the great work Dominion troops are doing. We are of the same blood, and the creation of friction by criticism is only playing the German’s game’.6
A few pertinent points of my own need to be made at this juncture, I feel, for it is very far from being the intention of this chapter to underplay the role of the Australians in the defence of Amiens and the operations at Villers-Bretonneux and elsewhere in March and April 1918. Indeed, Australian officers and other ranks had every reason to be proud of their performance in April 1918 and there were at least some grounds for their criticisms of that of the British units alongside them. It also has to be said that, on the whole, the Australian official historian, C.E.W. Bean, is very fair in his description and assessments of the First and Second Battles of Villers-Bretooneux, apportioning credit to British units, and individual officers and men, where it is due and levelling criticisms against them where they are justified.7 In turn, the Australian deeds at Villers-Bretonneux – particularly those of the 35th and 36th Battalions of the AIF on 4 April, and of the 13th and 15th Australian Brigades in the counter-attack of 24–25 April – are given reasonable prominence in the relevant accounts in the British official history.8 However, Sir James Edmonds rarely, if ever, attempts to cover the exploits of individual junior officers, NCOs and men in the same way as Bean does, and much of the spirit and character of the defence and counter-attacks is thereby lost or obscured in Edmonds’ drier and less vivid narrative. Moreover, whereas Bean devotes some 212 pages to the fighting in and around Villers-Bretonneux, Edmonds deals with it in only 34 pages. But neither Bean nor Edmonds would be classed as light reading, and, of the more recent popular accounts of 1918, only one or two – such as Gregory Blaxland’s Amiens 1918 and Peter Pedersen’s Villers-Bretonneux – pay any real attention to the efforts of British units in these operations.9 In these circumstances, it is hardly surprising that the story of Villers-Bretonneux does not loom large in the British national folk-memory. Outside of specialist miltary historians and members of the Western Front Association, one ventures to suggest that few British people would even have a clear idea as to where Villers-Bretonneux is located, let alone know what happened there.
My own observations and experiences would also suggest that this assessment would be much less true in the case of the Australian public. Today, visitors to Villers-Bretonneux, which lies some 10 miles east of Amiens, can see several potent symbols of Australia’s connections with the town, not least the Australian National Memorial – unveiled by King George VI in 1938 – which stands on the crest of Hill 104 just to the north of the town and very close to the ground crossed by the 15th Australian Brigade during its night counter-attack on 24–25 April 1918. Villers-Bretonneux was adopted by Melbourne after the First World War and the local school, situated in a street called the Rue Victoria, was rebuilt in the 1920s with money collected by children from the schools of that state. Since the 1970s, the building has housed a museum with displays on the AIF and its operations, especially those in Picardy and the Somme region in 1918. The town is twinned with Robinvale, Victoria, and inside the Mairie is a room containing further material marking the links with Australia. In contrast, apart from graves in the area’s war cemeteries and a memorial on the Cachy road to the first-ever tank versus tank action, it is very difficult to find any comparable physical reminders of the presence of British troops in and around the town in 1918.
How significant, then, was the contribution of British units to the defence of Villers-Bretonneux, and to what extent were Australian criticisms of the quality and performance of the British troops there actually justified?
It is still difficult to understand why the Germans did not make the key rail centre of Amiens a primary objective of their March offensive from the very outset. Approximately half of the BEF’s supplies came in through its three main southern ports (Rouen, Le Havre and Dieppe) and had to pass through or skirt the rail hub at some stage, as did 80 per cent of the north–south traffic, including trains from the Lens coalfields for French munitions factories.10 The loss of Amiens would therefore have seriously, perhaps fatally, impaired the Allied war effort and the BEF’s logistics and operational capacity. In the plans for Operation Michael, the German March offensive, Amiens does not appear to have featured as a main objective, though General von Hutier predicted in his operation orders for the German Eighteenth Army on 14 March that the French would use its vital rail centre to bring up reserves. On 21 March, the opening day of the German offensive, von Below’s Seventeenth Army and von der Marwitz’s Second Army, of Crown Prince Rupprecht’s Army Group, were to attack south of Arras, pinching off the Flesquières salient and advancing towards Bapaume and Péronne and across the old Somme battlefield before swinging north-west in a great left hook, enveloping Arras in the process. The object of the wheel to the north was to roll up the BEF and press it back against the sea. Von Hutier’s Eighteenth Army, of Crown Prince Wilhelm’s Army Group, was to advance beyond the Somme and the Crozat Canal to protect the left flank of the offensive, defeating any French reserves which might come up from the south to help the right flank of the British Fifth Army. Von Hutier might thus also sever the connection between the French and British armies.11
By 23 March, parts of the British Fifth Army had been driven back over 12 miles and the German Eighteenth Army was pushing on westwards to seize crossings over the Somme and Crozat Canal. At this point, Erich Ludendorff – who as First Quartermaster-General at German General Headquarters (OHL) was the de facto director of Germany’s war effort – allowed the tactical opportunity presented by von Hutier’s impressive progress to deflect him from the declared strategy, a tendency which bedevilled the German high command throughout the war. Ludendorff now issued new orders, directing the Seventeenth Army towards St Pol and Abbeville and the Second Army westwards to Amiens. The Eighteenth Army, originally given the flank protection role, was now to drive north-west, pushing back the French as well in a much more clearly defined attempt to separate the French and British armies. Instead of being concentrated for the gigantic left hook, the German armies would henceforth be advancing in divergent directions – to the north-west, the west and south-south-west – like a hand with the fingers spread wide.12
On 25 March, with the centre and left wing of the German Second Army, on either side of the Somme, beginning to lag behind, Ludendorff agreed with a suggestion from Crown Prince Wilhelm that, while the Second Army remained directed on Amiens, the weight of the Eighteenth Army should be transferred to its right wing so that, together with the Second Army, it might gain the old French line from the Avre valley to Caix (6 miles south-east of Villers-Bretonneux). The following day, with the French line being dangerously thrust back and the right of the German Eighteenth Army having advanced over 9 miles, Ludendorff continued to probe for the weakest spots. That evening, he ordered Second Army to make its principal effort south of the Somme and to press forward to the Avre with its left on Moreuil, a few miles south-east of Amiens, while its centre captured Amiens itself. The Eighteenth Army was also to maintain its drive from the line Noyon-Chaulnes south-west to the Avre, where it would secure crossings with a view to a further advance.13 As a 9-mile gap opened between the French First and Third Armies, Montdidier was abandoned by the French on 27 March, forcing the French First Army, under General Debeney, to move its assembly area back to St Just, on the only remaining direct railway line from Paris to Amiens. The focus of attention temporarily switched back to the north with the launching of Operation Mars, an attack against Arras, on 28 March, but the failure of this operation with heavy losses effectively sealed the fate of the Michael offensive as a whole. The only direction which offered any prospect of success was where the front was still fluid – namely opposite the left wing of the German Second Army and opposite the Eighteenth Army. It was no longer a question of what was desirable but what was possible.14
The grandiose objectives of Operation Michael had plainly proved to be beyond the strength and capacity of the forces employed and the colossal sweep of the original plan had been reduced in a few days to the scope of a local operation to seize a railway centre. Yet, as Bean notes, Ludendorff, despite the advice of his subordinate commanders, was now simply too close to the prize of Amiens to resist an attempt to grasp it, even if it meant only that the railway centre was brought under artillery fire. To this end, the inner wings of the Second and Eighteenth Armies were to push on towards Amiens by the shortest routes. This involved an advance on both sides of the Avre in a north-westerly direction beyond Moreuil to throw the enemy back over the Noye, some 5 miles west of the Avre.15 Even these more modest objectives were beyond the immediate powers of the exhausted Second and Eighteenth Armies and, although both armies carried out a series of small local attacks after 30 March, Ludendorff postponed further serious operations until 4 April, to allow time for rest and reorganisation. Above all he needed to bring up fresh supplies of ammunition along his over-stretched lines of communication over the old Somme battlefield, so that the next big attack would have adequate artillery preparation.
It should be borne in mind that the increasing resistance which the divisions of the British Third Army, stiffened by the 4th Australian Division and the New Zealand Division, were able to mount between Arras and the Somme – particularly after 26 March – was a not insignificant factor in causing the Germans to shift the weight and direction of the Michael offensive further to the left, or south-west. For example, Bean fully acknowledges the part played by the British 19th, 42nd and 62nd Divisions, with the New Zealanders, in blocking the German advance near Hébuterne in the Third Army area on 26 March. The British stragglers whom the Australians encountered in scenes of near-panic and confusion near Hébuterne that day – and who were roundly condemned by Australians in their letters home – appear mostly to have belonged to tramsport echelons, labour battalions and a divisional ammunition column. ‘The panic nowhere extended to the front-line troops’, Bean asserts.16 Even on the extreme British right flank, astride the Oise, where the British III Corps had been temporarily placed under French command, the hard-pressed formations of the British Fifth Army remained capable of springing some unpleasant surprises on the Germans, a case in point being the counter-attack delivered by the 18th Division at Baboeuf on 25 March, when its 54th Brigade retook the village and captured ten machine-guns.17
Between the Oise and the Avre, however, the front of the French First and Third Armies on the British Fifth Army’s right, was, for the moment, comparatively soft, encouraging the Germans to focus their attention on the Noyon–Montdidier–Moreuil sector. By 27 March, twelve French infantry divisions and three cavalry divisions had arrived, or were arriving, on the front held by Fayolle’s Reserve Army Group, and another thirteen infantry divisions and a dismounted cavalry division were either on their way or had been warned to prepare to move to the threatened sector. Nevertheless, these units invariably arrived with limited supplies of small arms ammunition, little or no artillery and without their cookers and other transport, and they were often committed piecemeal to the battle. When the British were driven back, pivoting on Arras, the French – for a crucial few days – formed a line facing not east but north, between the original right of the British Fifth Army at La Fère and its subsequent positions as far back as Moreuil. Thus, as Edmonds puts it, ‘instead of the British holding the base of a triangle, La Fère-Arras, they held one side, Moreuil-Arras, while the French occupied the other side, La Fère-Moreuil, with the Germans in the angle’.18
The brief, but alarming crisis of 24–25 March, when it seemed that the French and British armies might retire in different directions, led, on 26 March, to an inter-Allied conference at Doullens, at which General Foch was made responsible for the co-ordination of Allied operations on the Western Front.19 At a time when the German offensive was beginning to lose momentum through fatigue, heavy casualties and supply problems, Foch unquestionably helped to boost Allied morale and inspire confidence, and he acted promptly to ensure the maintenance of the link between the French and British armies, giving priority to the protection of Amiens. Even so, there was little he could do, in practical terms, to influence the immediate tactical situation. The twenty-nine British divisions which had been in the line on 21 March were now, six days later, greatly weakened by battle losses, sickness and straggling and were very tired, having had precious little rest or sleep. On 27 March, nine divisions, including the 4th Australian and New Zealand Divisions, had arrived from GHQ Reserve or other armies to reinforce them, and the 3rd and 5th Australian Divisions were also on their way south from the Second Army area and about to reach the Somme–Ancre region. But there was not much else left in the British locker.20 In fact, Foch was anxious to build up a strong reserve in the Amiens area for a future counterstroke and, with the French then under pressure in the Montdidier area, he could not yet fulfil his promise to relieve the British Fifth Army up to the Somme. All this meant, in essence, that the British Fifth Army would not only have to struggle on, existing at least partly on its own dwindling resources and reorganising itself where it stood, but would also, for some days to come, actually have to take over more of the front, from the Luce southwards to the line Mezières–Moreuil. Such developments clearly had a powerful bearing on the condition and performance of the British units at Villers-Bretonneux the following month.
To adhere to Foch’s directive that there was to be no further withdrawal was easier said than done. All the same, on 27 March, when the French gave way at Montdidier, the troops of the British Fifth Army managed to cling to a line not far short of that which Foch, the previous day, had ordered them to hold. On 27 March the 8th Division, under Major-General W.C.G. Heneker, stood its ground in the morning at Rosières, about 9 miles south-east of Villers-Bretonneux. In the early afternoon, two of its battalions, which had been sent hurriedly northwards to restore the situation around Proyart – just north of the Roman road from St Quentin to Amiens – made a gallant counter-attack that drove the advancing Germans back 1,000 yards and temporarily halted the enemy’s progress along the main Amiens road. And, as if these achievements were not enough for one day, three more of the 8th Division’s infantry battalions, together with engineers and the personnel of three brigade headquarters, helped to re-establish the 50th (Northumbrian) Division’s line and briefly recapture Vauvillers in another counter-attack.21 Unfortunately, German successes to the north and south rendered the position of the British XIX Corps in this key sector virtually untenable.
A misinterpretation of orders by VII Corps in the sector south of Albert had resulted in much of the right wing of the British Third Army falling back towards the Somme at Sailly-le-Sec, about 6 miles further west than the left flank of the Fifth Army. Exploiting this situation, the Germans crossed the Somme at Cérisy and Chipilly and thrust south-westwards, occupying Lamotte and the adjoining village of Warfusée Abancourt on the main Roman road in the evening, placing them only some 3 miles east of Villers-Bretonneux and barely a dozen miles from Amiens itself. A counter-attack on Lamotte and Warfusée on 28 March by the 61st (2nd South Midland) Division, supported by the 1st Cavalry Division, made some initial progress but was stopped 200 yards short of the Germans. The 61st Division subsequently withdrew south-westwards beyond Marcelcave. On the left of the XIX Corps line, infantry of the 16th Division, with the assistance of the 1st Cavalry Division – and men of the improvised formation of British, Canadian and American engineers and railway troops known collectively as Carey’s Force – repulsed a strong German attack on Hamel, to the north-east of Villers-Bretonneux. By the end of 28 March, having absorbed the rump of XVIII Corps, the British XIX Corps – under Lieutenant-General Sir Herbert Watts – was, to all intents and purposes, the sole surviving remnant of the Fifth Army still operational. Its front-line troops now only comprised the 39th, 61st and 66th Divisions, with the 20th Division in close support of the French on the right, and the 16th Division and 1st Cavalry Division to the left of Carey’s Force on the other flank.22
One could argue, with some justice, that these actions by the divisions of the Fifth Army in the last days of March ought to be borne in mind in any discussion of the defence of Amiens, and – with the subsequent operations between the Avre and the Luce from 30 March to 2 April – they could be said to constitute the first of the four phases of that defence. Given the exhaustion of the infantry, their inadequate level of training in open warfare, and the heavy casualties they had suffered, the troops of Fifth Army had displayed remarkable powers of resistance in this period. As Edmonds comments, not only ‘did they hold the enemy’s advance with little loss of ground, but they also often counter-attacked with success’. Bean too praises the tenacity of the Fifth Army at the end of March: ‘its unrelieved remnant was stubbornly holding an almost equally exhausted enemy … in maintaining the one condition vital to its side – a generally unbroken front – it was entirely successful; and … with reinforcements denied, its surviving wing still managed to keep the British right in touch with the arriving French’.23
For all that, Fifth Army was desperately weak by 28 March, when General Sir Hubert Gough was removed from command and replaced by General Sir Henry Rawlinson. Five days later, on 2 April, the Fifth Army was officially renamed the Fourth Army. On taking over from Gough on 28 March, Rawlinson telephoned Watts at his headquarters for a briefing on the situation facing XIX Corps. Watts replied breezily that ‘they may well get us by lunch-time and you by tea-time’.24 The prospects certainly did not appear to be rosy. Returns left by Gough put the total infantry strength of Fifth Army on 27 March at an estimated 21,650, of whom around 8,000 were still with the absent III Corps, and one should remember that this calculation had been made before the actions of 28 March diminished the strength still further. Of the individual divisions, the 8th, 30th and 50th were reckoned to have a fighting strength of perhaps 2,000 men each, while the 66th had only some 500. The divisions from Lieutenant-General Sir Ivor Maxse’s XVIII Corps, which was then being absorbed by the XIX Corps, were all graded as ‘unreliable’, ‘quite unreliable’ or ‘rather unreliable’ and the 61st Division, with a strength of perhaps 1,500, was classed as ‘all but tired out’. By 30 March some of these figures had been adjusted upwards, with the 50th Division now estimated to have a fighting strength of 3,000, the 8th Division 2,600, the 24th Division 1,800, the 66th Division 880 and the 39th Division 800 men. The position with regard to artillery was not quite so serious. Roughly half the 500 guns lost since 21 March had already been replaced and Gough bequeathed 783 to Rawlinson, including 225 medium and heavy pieces. In the words of Gregory Blaxland, these ‘could provide a much greater density of fire on the reduced frontage than was available on March 21st’.25 Even so, the overall situation was sufficiently gloomy for Rawlinson to write urgently to Foch on the evening of 28 March. Unless fresh troops were sent to him in the next two days, Rawlinson said that he doubted ‘whether the remnants of the British XIX Corps which now hold the line to the east of Villers-Bretonneux can maintain their positions’. Expressing his anxiety for the security of Amiens and seeking to impress upon Foch the seriousness of the threat to it if the Germans renewed their attacks from the east before the desired reinforcements arrived, he also predicted that XIX Corps was no longer capable of executing a counter-offensive.26
In the event, Rawlinson’s worst fears proved unfounded. Over the next three or four days more units did arrive to bolster the defences around Villers-Bretonneux, some of these formations having been requested by Gough before his departure. These included the 2nd and 3rd Cavalry Divisions, which had previously been operating with the French, much of the time in a mounted infantry role. Less emasculated than the standard infantry formations by the huge battles of attrition of earlier years, they contained a greater proportion of surviving Regulars who were expert with small arms and their Hotchkiss machine guns. The three divisions of Lieutenant-General Sir Richard Butler’s III Corps had also been relieved by the French, with whom they had been operating on the Oise, and had been moved around to the Amiens sector. All three had been in the front line on 21 March – when the 14th (Light) Division had been faced by no less than four German divisions with another two in the enemy’s second line and one in the third (a total of seven German divisons in all). Butler’s divisions had each incurred heavy casualties in the week following the start of the German offensive, the 58th (London) Division having lost 832 killed, wounded and missing, the 18th Division 2,445 and the 14th Division 3,197.27
These divisions had had a few days’ respite from battle – if not much actual rest – and had been partly filled up with new drafts, but they were still under-strength and particularly weak in officers and experienced NCOs. There had been no real opportunity to absorb or train the new arrivals properly, and the historian of the 18th Division recorded that its fresh recruits, ‘most of them raw youths’, had not yet had time to become imbued with ‘the corporate spirit of the division’.28 However, the 18th Division possessed some able senior officers, including its commander Richard Lee, a sapper of ‘quick grip and decision’ according to the divisional historian.29 They also included H.W. Higginson, the commander of the 53rd Brigade, a gifted tactician who was soon to take over the 12th (Eastern) Division; and the doughty Brigadier-General E.A. Wood, an extraordinary officer and one of those individuals for whom the First World War was their crowning moment. Wood, who stalked the battlefields in 1918 with a cigar clenched between his teeth and using a lance as an alpenstock, captured over twenty Germans later in the year by pelting them with lumps of chalk and old boots. Bean describes him as an officer ‘of stout build and of most stalwart disposition’.30 During the Great War, Wood won four DSOs, was Mentioned in Despatches nine times, was wounded five times, gassed twice and buried once. Small wonder, then, that Robert Cude, a runner with Wood’s 55th Brigade in 1918, wrote that he ‘would not mind going through Hell itself’ so long as Wood was in command.31
The 14th Division, however – which had been hard-hit on 21 March – had some problems in this regard. Its commander at the start of the German offensive, Major-General Sir Victor Couper, had been relieved on 22 March by Butler, who judged him to be ‘not in a fit state to handle the situation for the time being’.32 Couper was succeeded by Major-General W.H. ‘Bob’ Greenly from the 2nd Cavalry Division, but he too broke down under the pressure and was thought ‘not fit to continue operations’ by 28 March. Haig’s diary says, more bluntly, that Greenly ‘went off his head with the strain’.33 By early on 25 March Butler believed that the 14th Division should be pulled out of the fight. The division’s performance in March was damned with faint praise by Butler in his subsequent report, and Couper’s eventual successor, P.C.B. Skinner – previously the commander of its 41st Brigade – did not take over until 31 March. The 58th Division, the last to be released by the French, was on its way to XIX Corps, but only two of its battalions, the 6th and 7th Londons, would be close at hand on 4 April. Last, but by no means least, there was the 9th Australian Brigade, which had been detached from Monash’s 3rd Australian Division and hurried to Villers-Bretonneux on the night of 29 March. Its commander, Charles Rosenthal, brought to the brigade, in Bean’s words, ‘a robustness and audacity intensely welcome to its members’.34 The Australians had not been in a major action since Third Ypres and they were therefore fresher than the troops of III Corps. In addition, while the manpower crisis of early 1918 had forced the British to reduce the number of battalions in a brigade from four to three, the Australians still retained the old four-battalion organisation, and, although their divisions were also under-strength and the reservoir of trained reinforcements was running low, the rate at which their drafts arrived was nevertheless above that for British divisions.
During the period 30 March–2 April a series of actions took place to the south and south-east of Villers-Bretonneux, especially between the Avre and the Luce, as the Germans strove to seize features of local tactical advantage and to straighten the line preparatory to the next big assault. On 30 March, the Canadian Cavalry Brigade, of the 3rd Cavalry Division, fighting both mounted and on foot, distinguished itself by driving the Germans out of Moreuil Wood and recapturing it for a time.35 The same day, on the other side of the Luce, the 33rd Battalion of the AIF, under Lieutenant-Colonel Leslie Morshead – preceded by the 12th Lancers from the 2nd Cavalry Division – counter-attacked from the east of Cachy in an attempt to retake Aubercourt. The Lancers, galloping forward in lines of squadrons, succeeded in forcing the Germans out of a feature henceforth known as Lancer Wood, but the 33rd Battalion AIF, supported by companies of the 34th Battalion, came under heavy machine-gun fire 200 yards beyond the wood and failed to reach the objectives, the Australians losing 200 killed and wounded in the process.36 On the credit side, the very presence of the Australians and cavalry gave a lift to the spirits of the men of XIX Corps at a critical moment, and Morshead was quick to record his admiration for the 12th Lancers, writing that it was a privilege for him to work with such a fine regiment. ‘All ranks’, he noted, ‘were eager to give every possible help to us … One was able, too, to judge of the splendid work they are doing for the army at the present time, and they cannot be too highly praised’.37 To the north, between the Roman road and the Somme, engineers and pioneers of the 16th Division, and the 5th Dragoon Guards from the 1st Cavalry Division, combined to eject the Germans from trenches they had entered in front of Hamel, beating off a second attack later that afternoon. Next day, the Germans made some headway between the Avre and the Luce, gaining possession of Moreuil Wood and Rifle Wood, though a counter-attack of the 8th Division’s 25th Brigade – only some 200-strong – retook the copse at the north-west corner of Moreuil Wood at 4 p.m. On 1 April, around 1,000 men of the 4th, 5th and Canadian Cavalry Brigades, after an intense artillery bombardment and covered by a machine-gun barrage – advanced on foot in 3 waves to recapture Rifle Wood, along with over 100 prisoners and 13 machine-guns.38 Finally, on the evening of 2 April, companies of the 11th Royal Fusiliers and 7th Bedfords of the 54th Brigade, 18th Division – backed up by the 7th Royal West Kents from the 53rd Brigade – made an attempt to take a German trench on the high ground north-west of Aubercourt. They were spotted by German observers as they were forming up and were subjected to severe artillery and machine-gun fire, the leading companies losing over 100 officers and men before the attack was called off.39
It is worth re-emphasising, then, that the operations between Moreuil and Hamel at the end of March and the beginning of April, and the performance of the British divisions – particularly the cavalry formations – in these actions, were of no little significance to the defence of Amiens and Villers-Bretonneux. The determined stands and counter-attacks made during this period all played a part in keeping the Germans at arm’s length and in preventing an enemy breakthrough along the Roman road or from the south-east. They delayed the Germans sufficiently to allow time for further French reinforcements to arrive and take over, by 3 April, the sector from Moreuil to Hangard (inclusive); and they provided a bit of extra breathing space for Rawlinson and Watts to build up the defences at Villers-Bretonneux.
By the eve of First Villers-Bretonneux, the front line east of the town was held by the 18th Division on the right and the 14th Division on the left. Between them, from the Chaulnes–Amiens railway to the Roman road, was the 35th Battalion of the 9th Australian Brigade, now attached to the 18th Division. In support and reserve were the 6th Cavalry Brigade and the 33rd Battalion AIF. Behind them, in the so-called Gentelles Line, were the 24th Division, the 11th King’s (Liverpool Regiment) – who were the Pioneers of the 14th Division – the two remaining battalions of the 9th Australian Brigade and the 7th and Canadian Cavalry Brigades. Further back still, near Amiens itself, was the 2nd Cavalry Division. Thus, if not in overwhelming strength, XIX Corps – as Edmonds comments – was at least ‘disposed in considerable depth’.40 The front was also covered by the field artillery of five divisions as well as five heavy artillery brigades. But the unfortunate 14th (Light) Division represented a potential weak spot in the defence. It had first re-entered the line to the south, near Hangard, on the night of 1–2 April. The next night it was relieved by the French and was then transferred northwards to the left flank where, on the night of 3–4 April, it relieved the 1st Cavalry Division between the Roman road and the Somme.41 When considering its performance in the coming battle, it is therefore wise to remember that the 14th Division had been involved in reliefs on three successive nights and had not even seen, in daylight, the ground it was now expected to defend.
The first major German assault on Villers-Bretonneux began at 6.30 a.m. on 4 April, after a heavy 75-minute artillery bombardment on the British–Australian front line, artillery positions and rear areas. If successful, the Germans intended to follow it up next day with a drive against Amiens itself. Twelve German divisions attacked on a 9-mile front against the French and five more on a 6-mile front against the British and Australians. On the front held by the Australian 35th Battalion and the 55th Brigade of the 18th Division – where the 7th Buffs and the left of the 7th Royal West Kents were in the line – the Germans made little or no progress for most of the morning. At least three German attacks, in which the enemy infantry came forward in dense masses, were stopped by intense artillery, machine-gun and rifle fire. Eventually an attack against the line of posts occupied by the 7th Buffs, on the immediate right of the Australians, caused the Buffs to waver and leave their trenches, but they returned to the line when provided with covering fire by the Lewis guns of the 35th Battalion. According to Bean, the Buffs retired again shortly afterwards, although the 7th Royal West Kents stood firm.42
It was on the left flank, to the north, that the main crisis occurred. The 41st Brigade of the weak 14th Division, immediately north of the Roman road, fell back hurriedly in some disorder and, after rallying for a while on a support position 500 yards behind the front line, then retired another 3,000 yards to a ridge west of Vaire Wood, reaching this point about 10 a.m. The neighbouring 42nd Brigade, under Brigadier-General G.N.B. Forster, stood its ground but – with its flank exposed by the retirement of the 41st Brigade – it too withdrew in some haste, abandoning Hamel. Forster himself remained in Hamel to the last, was captured about midday and was killed by a stray bullet shortly afterwards.43
As the 14th Division gave way, the Australian 35th Battalion found the Germans behind their left flank and carried out a fighting withdrawal to a support position roughly a mile east of Villers-Bretonneux, where it linked up again with the men of the 7th Buffs who had retired earlier. Meanwhile, Lieutenant-Colonel Henry Goddard of the 35th Battalion – who had been placed by Rosenthal in local command of the Australian reserves in Villers-Bretonneux – ordered up three companies of the 33rd Battalion to protect his own left north of the railway and the Roman road. XIX Corps sent forward the two remaining battalions of the 9th Australian Brigade (the 34th and 36th) which, at Rosenthal’s urging, were kept concentrated for a possible counter-attack and placed in positions northwest and south of Villers-Bretonneux. General Lee, of the British 18th Division, commanding all the troops south of the Roman road, had summoned up the 6th Londons, from the 58th Division, giving them to Brigadier-General Wood of the 55th Brigade as a reserve. Wood, who was always at his best in the heat of battle, moved part of his own reserve battalion, the 7th Queen’s, and a company of the 6th Londons to his left flank to support the Buffs and hold the line of the railway south of the town, warning the other two companies of the 7th Queen’s to be ready to counter-attack in the event of further withdrawal.44 In the critical sector north of the Roman road, the day was saved by the British artillery, sometimes firing over open sights and with the Germans only 1,500 yards away; by the efforts of the 58th Battalion of the 5th Australian Division, holding the Somme crossings at Vaux; and by the three regiments of the 6th Cavalry Brigade, which came up to plug gaps in the line and whose Hotchkiss guns helped to beat off the attacking Germans. By the early afternoon, therefore, a new line (of sorts) had been formed and the situation temporarily stabilised.
This, however, was not the end of the day’s fighting. At around 4 o’clock in the afternoon, following an hour’s hurricane bombardment, two German divisions made a new attack from the south-east towards Cachy and Villers-Bretonneux, striking the 18th Division’s front hard. On the right, where the French were simultaneously driven out of Hangard cemetery, the 6th Northamptonshires, of the 54th Brigade, were forced back to the Cachy–Hangard road and the 8th East Surreys, of the 55th Brigade – now down to under 100 officers and men – likewise yielded ground. The 55th Brigade, hampered by the rain and mud which clogged many rifles and made them unserviceable, was pushed out of its positions and the Germans entered the northern part of Lancer Wood, threatening to outflank the 7th Royal West Kents, who were holding a line in front of its south-eastern corner. The West Kents therefore withdrew to the eastern edge of Hangard Wood to escape being surrounded. A swift counter-attack by the 8th Royal Berkshires of Higginson’s 53rd Brigade brought the Germans to a temporary halt but the battalion commander, Lieutenant-Colonel R.E. Dewing, was killed and the Berkshires were reduced to less than sixty men as a result of this action. Thereafter, the 53rd and 54th Brigades were able to hang on to their new positions and resist further attacks. The 55th Brigade, however, was still in some confusion and as the Buffs and East Surreys continued to retire, they took the Australian 35th Battalion, and the right of the 33rd Battalion, with them. The Germans had now nearly reached the Demuin road, leading directly to Villers-Bretonneux from the south, and, but for a few Canadian motor machine-gun batteries and six armoured cars, the way into the town appeared to be open.45
At precisely this moment – the second truly critical point in the battle – Lieutenant-Colonel Goddard ordered the Australian 36th Battalion, under Lieutenant-Colonel John Milne (once a private in the British army), to counterattack south of the railway. Launched at around 5.45 p.m., it was supported by the reserve company of the 35th Battalion on the left and 180 men of the 7th Queen’s – rallied and led forward by Brigadier-General Wood himself – on the right, with the 6th Londons acting as a ‘second wave’. Advancing with great dash in the face of intense small arms and machine-gun fire, the Australians and 7th Queen’s, despite taking heavy casualties, caught the Germans off-balance and swept them back nearly a mile on the left and half a mile on the right. Within an hour or so, a line east of Villers-Bretonneux was once more securely held.46 Even so, units of the 18th Division were called into action again on the evening of 5 April. At 7.20 p.m. that day, the 54th Brigade delivered a counter-attack in conjunction with the French, who had been thrown out of Hangard cemetery and wood in the late afternoon. The 6th Northants had to cross 1,500 yards of ground and were silhouetted against the light of a burning haystack, yet they got to within 50 yards of their objective – the sunken road running from Hangard to Hangard Wood – where they dug in. The French, on the right, pushed the Germans out of Hangard cemetery and regained the village.47
What then were the results of the first German attempt to seize Villers-Bretonneux and how well had the British units performed? The town had indeed been held, although Rawlinson’s Fourth Army had been driven back along its whole front and up to 2 miles at some points. The French XXXVI Corps on the right had similarly been pushed back 2 miles beyond the Avre, to the west of Moreuil, and the Germans were in part of the Bois de Sénécat, from the edge of which the outskirts of Amiens could be seen. On the other hand, more French reinforcements were arriving in the sector and the German penetration at this part of the line was thin, if deep. Bean reasonably gives most credit to the 3rd Cavalry Division and the 9th Australian Brigade for averting the danger to Villers-Bretonneux along the most direct line of approach north and south of the Roman road. The 14th Division, even when all the mitigating circumstances have been taken into consideration, cannot be said to have performed well. Parties of the Australian 58th Battalion near Vaire, south of the Somme, had tried in vain to stop officers and men of the 14th Division from withdrawing and many had already dumped their rifles and equipment. ‘Pompey’ Elliott, of the 15th Australian Brigade, issued an order to the 58th Battalion ‘to stop all stragglers and compel them to fight’. The Australians on this flank noted the difference in the spirit of the cavalry and Captain H.D.G. Ferres of the 58th Battalion later remarked that no men ‘could have done more than these cavalry men did’. Even in the centre, where the 35th Battalion and elements of the British 55th Brigade had retired, one unnamed Australian officer – who is quoted at length by Bean – was struck by the contrast between the British and Australians. Every single Australian still carried a rifle, recorded the officer in his diary, and they were easier to rally. The British, by way of comparison, ‘though only walking as if from a football match’, were, in his opinion, ‘quite spiritless’.48
This may all be justified comment, but it cannot be fairly applied to the British gunners, some of whom had kept their guns in action until the Germans were less than half a mile away. Neither would the 18th Division fully deserve such criticism. The division had suffered only slightly less than the 14th in the March Retreat, yet, with the possible exception of the 7th Buffs – who were unsteady and difficult to handle throughout much of the day – most of its battalions had fought hard. The 7th Queen’s had played a useful part in the decisive counter-attack in the late afternoon and early evening, while the 8th Royal Berkshires and 6th Northants also had enough spirit and determination left to mount telling counter-attacks of their own during the later stages of the fighting. On the command side, Watts and Lee, at corps and divisional level, had, in the main, contributed effectively by feeding in reinforcements and moving reserve units at appropriate moments. But, in the final analysis, it was the brigade, the battalion and, often, the company commanders who conducted the battle. Officers such as Rosenthal, Goddard and Milne among the Australians and, for the British, Wood, Higginson, Dewing and Lieutenant-Colonel C.B. Benson of the 6th Londons, all provided control or front-line leadership when and where it mattered most and certainly helped to compensate for the lack of experienced junior officers and NCOs, especially in the British units. It is interesting to note, in passing, that Goddard of the 35th Battalion AIF was born in England while Benson of the 6th Londons was born in Queensland.
The most important result of all from the defence of Villers-Bretonneux on 4 April was that it played a major part in persuading Ludendorff to call off Operation Michael and to transfer the main German effort to the Georgette offensive on the Lys, preparations for which were already in train. However, this did not mean that the front north and south of the Roman road at Villers-Bretonneux was altogether quiet. Rawlinson was now better off with regard to reserves but the tactical position still left much to be desired, since there were two big dents in the Allied line and the German artillery was now a mile closer to Amiens. South of the Luce, the alignment of the French formations ran back well behind that of the British Fourth Army, this being a constant source of anxiety to Haig and Rawlinson over the next two or three weeks, as it provided the Germans with a tempting route of attack from the south-east towards Cachy to outflank both the Fourth Army and Villers-Bretonneux. Immediately after First Villers-Bretonneux, tactical control of the Fourth Army’s front at Villers-Bretonneux passed from XIX Corps to Butler’s III Corps, though the front line itself was taken over entirely by Australian units – with the Australian Corps as a whole being transferred from the Third Army to the Fourth Army on 7 April. The 15th Australian Brigade, from the 5th Australian Division, occupied the new line between the Somme and Villers-Bretonneux; the 5th Australian Brigade, from the 2nd Australian Division, came in between Villers-Bretonneux and the French; and the 8th Australian Brigade – also from the 5th Australian Division – relieved the tired 24th Division as the reserve behind the southern sector of the Fourth Army’s line.49
Rawlinson was somewhat reassured by the presence of the Australians and confided to his diary on 5 April that he thought that the Fourth Army would now ‘be able to keep the Bosche out of Amiens’. But, whatever his inner thoughts, he continued to impress upon Haig and Foch his worries about the right flank. As he wrote to Haig on 15 April: ‘I have discussed this question with the III Corps commander and his Divisional Commanders and they are all agreed that unless something is done by the French to restore the situation on the right it may become serious and the safety of Amiens compromised’. Haig duly wrote to Foch in much the same vein two days later.50
To be fair to Foch, he was clearly aware of the need for some action on the Avre–Luce sector, although what could be achieved in practice was quite another matter. Even before First Villers-Bretonneux, Foch, on 3 April, had issued a general directive for continued Franco-British operations ‘with a view to freeing Amiens’ by driving the Germans ‘farther away’ from the important rail centre. The forces available would only permit this to be attempted in two stages. Another directive from Foch, issued on 6 April, laid down that, in preliminary operations, the French would seek to drive the Germans from part of the west side of the Avre, while the Fourth Army cleared up the rather messy tactical situation around the ravine and woods to the north and north-east of Hangard, near its junction with the French First Army. Then, once these initial steps had been taken, a joint attack would be launched to reach a line stretching from Moreuil, through Demuin and Aubercourt, to Warfusée – i.e. approximately the position held on 28 March.51 The combined attack was scheduled for 9 April but the French were not ready and it was postponed until 13 April. ‘The French do not seem out to do v[ery] much fighting’, Rawlinson noted on 8 April. He was also unhappy about Foch’s decision to place reserve divisions immediately west of Amiens in the Fourth Army area: ‘It is an infernal nuisance as it blocks all the roads and I feel sure they will be too far off there to be of much real value’. Nevertheless, Rawlinson grudgingly acknowledged that Foch was now ‘Generalissimo’ and had ‘issued the order in writing so it has to be obeyed’.52
As part of the preliminary operations called for by Foch, the 19th and 20th Battalions of the 5th Australian Brigade provided a company each on 7 April to attack and capture Hangard Wood. The attack was made with considerable dash and bravery – Lieutenant P.V. Storkey of the 19th Battalion winning the VC for his gallantry during the action – but it did not succeed and cost the battalions concerned 151 officers and men. Butler attributed the failure to enfilade machine-gun fire from south of the Luce but Bean disputes this and certainly the relative weakness of the attacking force, the thin and ragged supporting artillery barrage, and the poor selection of the objective for the 19th Battalion – which, in the event, proved untenable – were the main reasons for the lack of success, and all were at least partly Butler’s responsibility.53
The Germans too were not content to remain passive during this period. On the evening of 9 April they attacked Hangard, got into the village, and temporarily captured the cemetery to the east of it, until a French counter-attack, the same night, drove them out. At dawn on 12 April, a weighty bombardment heralded a fresh German attack, delivered in greater force than its predecessor. The 5th Australian Brigade’s sector of the front, next to the French, was at this time held by the 34th and 36th Battalions (lent by the 9th Australian Brigade), with two of the 18th Division’s battalions – the 7th Royal West Kents and 10th Essex – standing by in close support for a possible counter-attack. The German assault quickly penetrated into Hangard and the French managed to retain only the chateau. At 10.15 a.m. the Royal West Kents counter-attacked alongside the French. The latter failed to retake Hangard and some of the West Kents apparently wavered as they came under fire en route to the Australian front line, but the remainder went on and, with the French, reached and held on to the western edge of the copse above the village. At 6.30 in the evening, the Germans at last captured the chateau, only to be subjected, at 7.20 p.m., to yet another Franco-British counterattack, in which the 10th Essex played a distinguished part. Advancing steadily in artillery formation through a German barrage, the 10th Essex lost about half their strength, but ‘went straight through the German position, as did the French on their right’. Although the cemetery and the copse 200 yards to its north-east stayed in German hands, the much-disputed village of Hangard was recaptured with over 120 prisoners. ‘It was desperate, it was costly, it was successful’, observed the historians of the 10th Essex.54 The bitter fighting around Hangard on 12 April caused the cancellation of the joint attack which III Corps and the French were to have launched the following day, but the struggle in this sector flared up again briefly on 15 April, when the 18th Battalion AIF, from the 5th Australian Brigade, attacked with the French in a minor operation to recapture the cemetery and its neighbouring copse. The Australians failed to secure the copse, losing 84 out of 180 officers and men who participated in the attack, though the French seized and held on to the cemetery.55
The fighting for, and around, Hangard – which may be said to have constituted the third of the four phases of the defence of Amiens and Villers-Bretonneux – unquestionably contributed to Rawlinson’s unease about the security and alignment of his right flank and, as we have seen, impelled him to keep up the pressure on Foch, through Haig, to take appropriate action. At last, on 18 April, the French First Army, under General Marie-Eugène Debeney, launched an attack without British assistance, seeking to force the Germans back across the Avre. Edmonds does not waste the opportunity to comment tartly that this ‘was the first offensive action taken by the French to relieve the British since the fighting began on the 21st March’. Debeney’s operations were only partly successful, gaining just 500 yards on average, and no progress was made east of the Avre, yet, by clearing the enemy out of the Bois de Sénécat and the ridge north of it, the French deprived the Germans of their observation over the rear areas of the British III Corps. In this respect, therefore, the French attack was not totally without value to Rawlinson.56
While all this was happening on the southern flank, ‘Pompey’ Elliott’s 15th Australian Brigade – though not, in the end, required to take part in the proposed Franco-British attack – had, under Elliott’s dynamic and energetic leadership, made good use of its time in the sector between Villers-Bretonneux and the Somme. As Bean remarks, the brigade was then ‘at the zenith of its form; a magnificent instrument, fit, like Cromwell’s Ironsides, for the hardest military tasks’, even if Elliott’s frequently voiced contempt for the quality of British troops and his ‘hot-headed tendency to use his brigade as if it were independent of the rest of the BEF’ often caused trouble.57 From 7 April, by means of aggressive patrolling and methodical reconnaissance, the Australians were able to advance at night in successive bounds of 500 yards or so, bringing the front line closer to Hamel. When the 15th Australian Brigade went into reserve and was replaced at the front by its sister brigade, the 14th, Elliott – with the approval of Hobbs – kept one battalion in the Bois l’Abbé, in part of the Aubigny line a mile behind the town, ordering it to be in constant readiness to counter-attack should Villers-Bretonneux fall. Elliott ensured that his staff and battalion commanders knew the ground, of which he ordered contour models to be made, and he also drew up contingency plans for the counter-attack operation. After inspecting the Australian Corps on 14 April, Rawlinson remarked on the contrast between them and Butler’s troops. ‘They are a splendid body of men’, he noted, ‘and Hobbs and Monash are both very good commanders. They are ready for any emergency, which is comforting; for Butler’s III Corps, which has recently been reinforced, has little or no chance of training its drafts and is short of officers’.58
With Haig insistent that both sides of the Somme, to the north, should be held by one formation, the Australian Corps, the Villers-Bretonneux front was reorganised in mid-April. The Australian front now lay astride the Somme and, on its right, extended to the summit of Hill 104, the inter-corps boundary being about 1,300 yards north of the Roman road. The defence of the Villers-Bretonneux plateau and the town itself became the responsibility of Butler’s III Corps and, in particular, the 8th Division, which had been transferred to III Corps to replace the shattered 14th (Light) Division and which took over this key sector between 19 and 21 April. As the evidence increased that another German thrust against Villers-Bretonneux was imminent, Rawlinson had some cause for concern about the current condition of III Corps. The 8th Division, for example, had lost some 250 officers and 4,693 other ranks – roughly half its infantry – during its splendid fighting withdrawal in March. The two battalions which had suffered the most were the 2nd Middlesex and the 2nd West Yorkshires, the latter having recently received drafts of no fewer than 11 officers and 700 other ranks and the 2nd Middlesex almost as many.59 In addition, the division’s own artillery was currently refitting and training elsewhere, which meant that, in the immediate future, the formation would have to work with the less familiar gunners of the 20th (Light) Division. There was, however, no shortage of quality among its senior officers. Brigadier-General Clifford Coffin, of the 25th Brigade, had won the VC at Third Ypres while Brigadier-General George Grogan, of the 23rd Brigade, was destined to win the same decoration on the Aisne in May 1918. W.C.G. Heneker, the long-serving divisional commander, was described by one of his former battalion commanders, James Jack, as ‘a fine but exacting chief’. John Bourne says that his command style was that of an auditor, with nothing escaping his ‘penetrating and unrelenting attention’. Hubert Essame recalled, many years later, that Heneker expected to be saluted by everyone within eye range: ‘His eagle eye could detect an unshaven chin, the need for a hair cut, a grease stain or an unpolished button at a considerable distance. His comments were unequivocally clear, vividly expressed and long remembered’. Heneker’s GSO1 from November 1916 to December 1917, Lieutenant-Colonel E.H.L. ‘Moses’ Beddington, admitted that his commander was ‘a bit of a bully’ but thought him to be a good tactician and ‘a good man to serve so long as one stood up to him’.60 The performance of the 8th Division in the final week of March 1918 should leave no room for doubt about Heneker’s genuine capabilities as a divisional commander in battle.
The 18th Division faced similar problems in having to incorporate and train new drafts as quickly as it could. The 7th Royal West Kents at least received strong drafts of third-line Yeomanry and Kent Cyclists, who had undergone a fair amount of training in England and who brought the battalion’s ration strength up to 700; the 8th Royal Berkshires contained 350 youngsters under the age of 19; and 60 per cent of the 7th Queen’s were also boys under 19 who, until the week before, had never fired a round. On 23 April eight new officers joined the 10th Essex. Within three days all but one would be dead or wounded. Rawlinson was right to be worried by the need to rely upon those he subsequently described as ‘children’.61 It is true that the overall strategic situation, with the Germans preparing further operations in Flanders, left Haig and Rawlinson with a limited range of options. However, it is reasonable to ask whether they might have been wiser to leave the defence of Villers-Bretonneux in the hands of well-integrated and combat-hardened Australian formations rather than entrust it to the divisions of III Corps, which were weakened and embodied recent drafts of uncertain quality. Moreover, Richard Butler, who had previously spent three years on Haig’s staff, latterly as Deputy Chief of Staff at GHQ, lacked experience as an operational corps commander. Not surprisingly, ‘Pompey’ Elliott, invariably a vociferous critic of British troops and generals, questioned Rawlinson’s judgement in this regard and expressed his concerns forcefully to Hobbs.62
The reorganisation of the front was carried out in stages between 13 and 21 April. The 58th (London) Division – also barely rested and refitted after the March crisis – took over the southern sector from Monument Wood down through Hangard Wood to the junction with the French, holding the front line with all three battalions of the 173rd Brigade, and with the 2/10th Londons from the 174th Brigade attached for counter-attack purposes. In the course of the reorganisation, on 17 April and the following days, the Germans drenched Villers-Bretonneux, the Bois d’Aquenne to the south-west of the town, and the village of Cachy to the south, with mustard gas, phosgene and irritant gases, causing some 1,074 casualties, chiefly in the 58th Division and in Australian units. By the eve of the next German attack, the 8th Division completed its deployment, if only just. The 23rd Brigade, under Grogan, was on the right next to the 58th Division, and had the 2nd West Yorkshires and 2nd Middlesex in the front line, to cover Villers-Bretonneux, with the 2nd Devons, as counter-attack battalion, behind them and immediately to the south of the town. Coffin’s 25th Brigade, had the 2nd Rifle Brigade in the front line to the left of the Middlesex and on the other side of the Roman road; the 2nd East Lancashires were placed in trenches around the town for close defence; and the 2nd Royal Berkshires, the counter-attack battalion, were located to the north of Villers-Bretonneux. The 24th Brigade was in divisional reserve, but two companies of the 1st Worcestershires were in the Cachy Switch, protecting the village of Cachy with a scratch formation known as Shepherd’s Force – which included two companies of the 6th Northants and the 83rd Brigade, Royal Field Artillery, from the 18th Division.
A vaulable addition to the strength of Butler’s forces was made with the arrival of a dozen or so Mark IV tanks, some of which were placed in front of the Bois l’Abbé to prevent the Germans from gaining a foothold in the Bois d’Aquenne, while the remainder were held in reserve to check, by immediate counter-attack, any German advance near Gentelles and Cachy. Seven light Whippet tanks were also available in a wood a mile south of Blangy. The German gas bombardment, however, seriously affected the crews of the three tanks near the Bois d’Aquenne, causing the forward tanks to be moved south of the Bois l’Abbé to reduce the danger of this happening again. A total of 7 field artillery brigades and 89 guns of the III Corps heavy artillery supported the 8th and 58th Divisions.63
Thanks to a lot of hard work since 5 April, there were now four rear defence lines in front of Amiens between the Avre and the Somme. The Gentelles-Bois l’Abbé–Aubigny line, or reserve line, behind Villers-Bretonneux had one switch running to the front line from Gentelles past Cachy to the Bois d’Aquenne and the south-east face of Villers-Bretonneux, and another switch, as right flank protection in their area, dug by the Australians. Approximately 1,500 yards behind the reserve line was the Bois line, in front of the Bois de Blangy and Bois de Gentelles. Then, 2,500 yards behind this was the Blangy-Tronville line and, finally, a further mile to the rear, the Glisy line. Edmonds points out that the III Corps defences had some depth even though the trenches were shallow and lacked dugouts and adequate wire protection. He emphasises too that, for want of troops, there was no large body of men available for immediate counter-attack.64 Elliott and Hobbs, indeed, were disturbed to find that, on 20 April, when the 8th Division took over the southern part of Elliott’s area, only one platoon of the incoming 22nd Durham Light Infantry (DLI), a Pioneer battalion, was assigned to relieve the battalion which he had kept for counter-attack. When protests were made to III Corps, the response was that Butler had sanctioned different arrangements for the weaker brigades of the 8th Division. More anxious than ever, Elliott maintained one of his battalions, the 59th, at the southern end of the Aubigny line in the Bois l’Abbé, ready to counter-attack should the Germans break through.65
The second big assault by the Germans on the Fourth Army’s front at Villers-Bretonneux began in thick fog at 6 a.m. on 24 April, preceded by a bombardment – which included gas shells – lasting 2 hours and 15 minutes and covering the Allied front positions and rear areas between the Bois de Sénécat and the Somme. The attack had two principal objects: one was to straighten out and blunt the German salient facing Amiens and south of the Luce by capturing Villers-Bretonneux and the plateau on which it stood; the other was to divert Allied attention from German operations in Flanders, where a final attempt was to be made to seize Kemmel. It was hoped that, at least, the attack at Villers-Bretonneux would bring the German artillery within even closer range of Amiens, and if Villers-Bretonneux, the main objective, was taken, the success was to be exploited. Four divisions – the 228th, 4th Guard, 77th Reserve and 208th, and the left flank of a fifth, the 243rd Division – were to attack the front held by the British 8th and 58th Divisions while the Guard Ersatz Division struck Hangard and demonstrated on the front of the French 131st Division to the south. The German 19th, 9th Bavarian Reserve and Jäger Divisions were in reserve. The 228th and 4th Guard Divisions were to assault Villers-Bretonneux, the 77th was to take Hangard Wood and Cachy and the 208th would secure Hangard village and its neighbouring copse. The attack would be accompanied by thirteen German A7V tanks, divided into three groups. One group of three advanced north of the railway directly against the town; the second, with six tanks, moved south of the railway between the town and Cachy; and the third, with four tanks, attacked Cachy itself. Flamethrowers were also to be employed by the Germans in the assault.66