The cavalry question at Arras

During the German withdrawal to the Hindenburg Line our cavalry performed a useful role in operations against the enemy’s rearguard. It  successfully outflanked the village of Roisel, even though many of the garrison there were able to slip away. Elsewhere, there were similar small triumphs. At Equancourt, for example, the Fort Garry Horse showed flexibility, dismounting and setting up firing positions to the north and west of the village while Lord Strathcona’s Horse went in at a gallop from the south, albeit in the face of flimsy opposition. The operations carried out to secure a line through Guyencourt, Saulcourt and Villers-Faucon involved the same tactics of fire and movement, though on a somewhat larger scale. Again, these were carried out successfully.

However, it’s important to remember that all these operations were pitched against an enemy that had no intention of holding the positions in question. The garrisons left behind to defend these villages were there to cause a temporary delay to the British advance, not to halt it in its tracks. For the most part, the defenders were able to fall back unmolested while the cavalry set about securing a succession of half-empty villages. Roeux and Greenland Hill, on the other hand, were a very different proposition. Here, the Germans had no intention of falling back, in fact quite the opposite was true. Enemy action to delay the British advance on the 9th April was carried out in order to allow reinforcements to arrive, not to buy time for a further withdrawal. It’s an important distinction, particularly in relation to the cavalry question; that great ‘what if?’

Let’s begin with a couple of observations from people who were actually on the battlefield on the 9th April 1917. Both narratives relate to events immediately north of the River Scarpe where the 9th (Scottish) Division and the 4th Division were in action.

The first is part of a telephone conversation between Brigadier-General Frank Aylmer Maxwell, 27th Brigade, 9th (Scottish) Division, and Lieutenant-Colonel William Denman Croft, 11th Royal Scots. It’s a quotation that often appears in books and articles dealing with the opening day’s fighting.

“Are the Boches on the run?”
“Is cavalry good business?”
“Yes, ten thousand times yes, but it must be done now. Too late tomorrow…..why can’t we go on.”

The second observation has also found its way into many discussions and narratives relating to the opening day’s fighting at Arras. It was a response to events as they unfolded later that afternoon and comes to us courtesy of one of the war’s most colourful characters, Brigadier-General Adrian Carton de Wiart, 12th Brigade, 4th Division – 9th April 1917.

” We reached our objective, having suffered about two hundred casualties, which in these days was comparably light. We could have taken many more prisoners and much valuable ground if only cavalry had been available, but as it was we could see the guns being driven away into the distance, to be used against us, another day.”

Turning the clock back a few hours that day, the situation facing Croft and Maxwell at around lunchtime might well have been favourable for the deployment of cavalry. At that moment in time, Greenland Hill might have been captured along with embankment carrying the Arras-Douai railway line; the Germans might even have abandoned Roeux and the Chemical Works as a consequence, but we can never know for sure. The matter, of course, is completely academic because the cavalry was not readily to hand on that part of the battlefield at that precise moment in time.

What does seem fairly certain is that by the time Brigadier-General Carton de Wiart was lamenting the absence of cavalry, as the enemy’s remaining guns were being dragged off the battlefield in front of him, it was already too late.  If his own men were unable to advance further in the face of heavy rifle and machine-gun fire, what chance did the cavalry have?

Let’s suppose that sufficient cavalry had been assembled near Athies at around 4 p.m. on the afternoon of the 9th April. From that location there were only two possible approaches to Greenland Hill; the first was the road running along the valley of the Scarpe to Fampoux, the second was up over the Point du Jour Ridge. The first route, which offered some degree of cover, at least as far as Fampoux, would undoubtedly have been the safer option. After that, cavalry would have  to deploy across completely open ground between the sunken road, just north of Fampoux, and Greenland Hill. The second option merely led to an open skyline in full view of Gavrelle, which was still in enemy hands. Just beyond this village the Germans had a number of artillery pieces still able and willing to stand their ground.

Let’s assume that cavalry had been deployed from the Scarpe valley towards Greenland Hill. Any movement of mounted troops across the open ground between Fampoux and Greenland Hill at around 4 p.m. was almost certain to result in heavy casualties. Cavalry at full gallop would probably have been capable of reaching the ruined inn at the foot of Greenland Hill, no doubt overwhelming the small garrison there, but that wasn’t the problem. The real problem was the enfilade fire coming from the embankment and the area around the station. In terms of casualties, the damage inflicted from these positions would have been significant, though perhaps not catastrophic.

Reaching the embankment was one thing, but mounting and clearing it was a much more difficult task, and again potentially costly. Even if cavalry succeeded here, there was still Roeux to contend with. In order to appreciate the probable outcome we only have to look at what happened when a squadron of Canadian Light Horse approached the village of Willerval at 4.20 p.m. on the opening day. One patrol came under machine-gun fire and lost half of its men and horses, another was wiped out by rifle fire alone, and the rest of the squadron lost half of its horses when it came under shell fire. Unless the cavalry could dislodge the German detachments along the embankment, then deal with those around the station and the Chemical Works, it would prove impossible for the 4th Division to advance from its current position towards Greenland Hill.

By the latter part of the afternoon the Germans had begun to shell Fampoux and the Point du Jour Ridge. Had cavalry been deployed towards Greenland Hill and beyond, there was a real risk of it fragmenting and becoming cut off, unable to go forward alone and with no way back other than across open ground now swept by shelling, as well as heavy rifle and machine-gun fire. The only other option was to dig in and hope that other units could come forward in support, something that we know the infantry was unable to do, or else wait until nightfall before pulling back, which was exactly the situation faced by ‘B’ Squadron, Fort Garry Horse, at Cambrai when it was forced to abandon its mounts and make its way back on foot under cover of darkness, having advanced too far without support.

Cavalry, even if it manages to capture ground, needs to be relieved reasonably quickly by infantry. On the 9th April each trooper carried just under 200 rounds of rifle ammunition. The cavalry also had its own machine guns, but the re-supply of ammunition to these units was absolutely critical to their effectiveness. At Monchy, on the 11th April, the cavalry performed extremely well in a dismounted defensive capacity, but it didn’t actually face counter-attacks that day, just a terrific bombardment. It arrived there in full daylight, despite blizzard conditions, and was still largely intact when it got there. More importantly, it was never cut off from support, though immediate relief throughout the day was made difficult by enemy shelling, and, of course, infantry units were already occupying the village when cavalry sought refuge there. On the other hand, it seems likely that groups of cavalry in and around Greenland Hill would have found it very difficult to hold their ground without infantry and artillery support, especially if they found themselves isolated in small pockets. We have to remember that the fear of overnight counter-attacks was very real, though with hindsight we know that the Germans didn’t have the capacity to deliver them.

It’s also important to bear in mind that the area between the River Scarpe and the Point du Jour Ridge was already a very crowded place and any attempt to bring cavalry forward here during the afternoon of the 9th April would have resulted in even greater congestion. The decision to rely solely on infantry to press home the advance on this part of the battlefield meant that there simply wasn’t the space for the assembly of massed cavalry.

One voice that deserves to be heard is that of Brigadier-General Archibald Fraser Home, who at the time was serving with the Cavalry Corps on General Kavanagh’s staff. According to him, there was no real opportunity for the cavalry on the opening day at Arras, nor did he see this as a missed opportunity. In his diary he notes that there was still too much uncut wire on the battlefield, pointing out that this was always a significant obstacle for cavalry. That was certainly true south of the River Scarpe. The Brown Line, between Orange Hill and the main Arras-Cambrai road, was only cleared on the 10th April, along with the wire that protected it. Here, the wire was as broad as an average street and around five feet tall. Such an obstacle ruled out any deployment of the 2nd and 3rd Cavalry Divisions as they gathered around Tilloy on the opening day. When it comes to the cavalry question Home’s assessment of the situation must surely carry some weight.

On the other side of the river the wire covering the Brown Line was still largely intact when Croft reached it at lunchtime on the 9th April. Gaps could have been created, but the nearest cavalry brigade was still too far away to have any bearing on the day’s events. The 1st Cavalry Division was located north of the River Scarpe, but its role was to co-operate with the First Army around Vimy Ridge rather than the Scarpe valley. The transfer of cavalry from south of the river to the opposite bank would have been logistically difficult on an active battlefield that was already very crowded; besides, there simply wasn’t time to organise such a move on the opening day.

The advance along the Scarpe valley, as set out in the overall plan, involved one infantry division (4th Division) leapfrogging another (9th Division). Within that plan, cavalry deployment would only be considered once the infantry had reached the Green Line, and even that was conditional upon whether the enemy was judged to be intent on retiring behind the Drocourt-Quéant Line. As the afternoon of the 9th April wore on, increasing resistance around Roeux and Greenland Hill made it very difficult to assess the enemy’s real intentions.

The real priority on the opening day was to bring up sufficient guns and shells to support further attacks, attacks that could be properly co-ordinated and carried out at a later date, rather than introduce large numbers of cavalry to an already congested battlefield. That was the view of the man in charge of the 9th (Scottish) Division’s artillery, Brigadier-General Tudor, who on the 12th April made the following comment regarding the opening day:

“As at Neuve Chapelle and Loos, it was demonstrated (on 9th April) that it is vital to send up supporting troops immediately. For an hour or two anything can be done.” With regard to the following days he goes on to say: “Any further attack must be thoroughly prepared like an original attack.” As far as thinking was concerned, Tudor was cut from the same cloth as General Plumer.

Carton de Wiart was no fool, and neither were Maxwell and Croft, but if the cavalry couldn’t be deployed at around 1 p.m on the opening day, perhaps 2 p.m. at the very latest, then the opportunity was simply gone. If anyone managed to assess the situation correctly it was Croft when, in the absence of cavalry, he requested permission to push on beyond the Brown Line. As I said in a previous article on this website, the real missed opportunity on the opening day was probably the two hours it took for the 4th Division to relieve the 9th (Scottish) Division.

The cavalry debate will no doubt continue with regard to the opening day at Arras. Cavalry still had a part to play on the battlefield, but not on the opening day at Arras. It’s role on other occasions would always be limited and it had been that way on the Western Front since the opening days of the war. Others have argued differently, but I have to disagree.