Over the Pyrenees

In the first part of this mini-series, On the Plains of Spain, WDS designer and historian Andrew Bamford followed a March 2026 battlefield study with the Royal Dragoon Guards through Salamanca and the Portuguese border country. This time, he returns to the Peninsula with a more senior group from Headquarters Regional Command, examining the operational and strategic problems behind Wellington’s 1813 campaign.

From the victory at Vitoria to the fighting in the Pyrenees and the sieges of San Sebastian, Andrew combines battlefield photographs, historical narrative, and screenshots from Wellington’s Peninsular War to explore how an army advances, sustains itself, and exploits success across difficult ground.

Three months after my trip to Salamanca and the Portuguese border country, recounted in a previous blog instalment, I was back in the Peninsula conducting another battlefield study (staff ride, to our American friends) with elements of the British Army. Whereas my previous study had been with a group from a single regiment, looking at tactical considerations and sites associated with one of their antecedent units, this time I was with a group from Headquarters Regional Command – a rather more senior party in terms of rank and experience, and seeking a different perspective. Our goal this time was therefore to look at the high-level conduct of war, concentrating, though not exclusively so, on the operational sphere. Our focus would be Wellington’s 1813 campaign to drive the French out of Spain – accomplished in the Battle of Vitoria and subsequent exploitation – and the fighting that followed in the Pyrenees. As with the previous blog instalment, what follows is a selection of my photographs from the trip, paired where appropriate with screenshots that show how these actions are portrayed in the second of WDS’s two titles covering the conflict, Wellington’s Peninsular War (WPW for short), designed by Rich White with me providing the order-of-battle research. I’ve tried to provide enough history for the images to make sense even to someone not familiar with the ebb and flow of the conflict, but this is by no means intended to be an exhaustive account of these campaigns.

The Advance to Vitoria

To briefly pick up the story from where my previous study left off, after his victory at Salamanca in July 1812 Wellington went on to liberate Madrid and advance deep into northern Spain. He was eventually stopped by his inability to capture the fortress of Burgos due to a lack of heavy artillery, and then obliged to withdraw in order to escape destruction as the French field armies concentrated against him. The end of the year therefore saw the allies back on the Portuguese border where they had started the campaign, but by concentrating against Wellington the French had had to give up their hold on much of southern Spain and this they never regained. As he spent the next six months rebuilding his depleted army for a new campaign, Wellington could therefore incorporate considerable numbers of British, Portuguese, and Spanish troops previously engaged in the southern theatre as well as fresh British troops sent out as reinforcements. The French commanders, by contrast, found themselves called upon to detach men to help make up the losses from Napoleon’s Russian disaster. When hostilities resumed, therefore, the forces would be much more evenly matched.

So far as the Anglo-Portuguese element of Wellington’s command was concerned, the backbone was provided by the seven numbered divisions of infantry, generally formed with one brigade of British and one of Portuguese. There was also the Light Division of two mixed Anglo-Portuguese brigades, one all-Portuguese division of two brigades, and two independent Portuguese brigades. The cavalry was organised as a single large division of nine brigades but in effect this was an administrative formation and brigades functioned as the main operational unit. Anglo-Portuguese all-arms strength at the outset of the campaign totalled 81,000. An army that size required higher command above the divisional level. From the outset, Sir Thomas Graham and Sir Rowland Hill functioned as ‘corps’ commanders of the left and right wings respectively, but more commanders would in due course be needed for multi-divisional columns or detached commands and this became a problem as the fighting moved into the mountains and the army became dispersed. Appointments based on seniority did not always produce the best man for the job.

In addition to the Anglo-Portuguese elements, Wellington also had significant numbers of Spanish troops under direct command for the first time. Totalling some 46,000 men, for an allied total of 127,000, these mostly operated as part of their own corps-sized armies, with the army commanders reporting to Wellington, but Longa’s division was directly subordinated to Graham at Vitoria and Morillo’s formed part of long-term partnership with Hill’s command. The Spanish army had been substantially reformed by this stage of the war, including taking the more effective guerilla units into the line, but poor commissariat arrangements and desire for revenge over the French limited Wellington’s ability to take them over the Pyrenees and into France.

The situation of the French armies, meanwhile, was shaped by the fact that Napoleon had lost tens of thousands of troops in Russia and needed to rebuild his army in Germany. Few regiments in Spain were withdrawn in their entirety – just a few guard units – but most gave up elements, concentrating effective manpower into one or two battalions or squadrons and sending junior sub-unit cadres home to be built up with conscripts and used for the war in Germany.

From 1810 onwards, the French forces in Spain had been broken up into a series of regional armies all reporting – at least notionally – to Joseph Bonaparte as French-backed King of Spain (who had Marshal Jourdan as his chief-of-staff). By 1813, so far as the troops facing Wellington were concerned, these comprised:

• Armée du Centre, under d’Erlon; two divisions of infantry and one of cavalry, plus Joseph’s Spanish ‘army’ with a division apiece of line and guard troops = 17,500
• Armée du Midi, under Gazan, four and a half divisions of infantry, three divisions of cavalry = 34,500
• Armée du Nord, under Clausel, three divisions of infantry, a brigade of cavalry, garrisons = 21,000
• Armée de Portugal, under Reille, six divisions of infantry and two of cavalry = 32,000

Total = 105,000.

French forces in Aragon and Catalonia under Marshal Suchet were to all intents and purposes a separate command and not included in these totals – from Wellington’s perspective, they were neutralised by an equivalent Anglo-Spanish force.

French armies had not fully reoccupied Spain after Wellington’s retreat to Portugal at the end of the 1812 campaign. With Madrid held as an outpost, the concentration was on Valladolid and the focus was to be clearing the ‘brigands’ – i.e. guerillas – from northern Spain. This employed the whole Armée du Nord, reinforced with three infantry divisions of the Armée de Portugal, and was all against Joseph’s preference, which had been to hold Madrid in force for political reasons and prepare for an advance by Wellington. Napoleon was still, after five years, convinced that he could run the war by remote control. The French expectation was a repeat of 1812 – an advance on Salamanca then up the Great Road that ran via Valladolid, Burgos, and Vitoria to the French border. Wellington sought to encourage them in that view, placing himself with Hill’s corps when it moved from Ciudad Rodrigo in late May. Surely, the commanding general would be with his main effort? Well, no, he wasn’t. Graham in fact had the bulk of the Anglo-Portuguese troops up in the Tras os Montes, the mountainous north-east of Portugal, and, with French eyes on Wellington, began to move out covered by a cavalry screen. The French flank was completely turned, and for the rest of the campaign they would be forced to trade space for time as they sought to concentrate their forces and extricate the huge baggage train containing all the loot of their occupation.

By 3 June Wellington had concentrated at Toro and the French were falling back on Burgos. The move to outflank the French right continued, with the army advancing in four main columns. On 13 June, outflanked, the French blew up the fortifications of Burgos and fell back on Vitoria, and on the 17th Wellington gave orders for his final advance. This led to two actions the following day at San Millan and Osma, both allied victories.

Joseph finally called a halt at Vitoria. Time was needed to get the accumulated heavy baggage and loot away to France, and – hopefully – for Clausel’s detached command to come in. On the 20th a first convoy went north escorted by the erstwhile Vitoria garrison; at dawn on the 21st Maucune’s division, cut up at San Millan, set off with a second convoy. Nevertheless, much of the baggage still remained, along with the collection of wives, families, mistresses, courtiers, hangers-on, Spaniards who had backed the wrong side, and other such waifs and strays, leading one French officer to describe Joseph’s command as ‘un bordel ambulant’: a moving brothel! Allowing for losses prior to the main battle, there were probably around 63,000 troops available to Joseph on the field of Vitoria. However, Wellington had only two of the Spanish divisions with him on the battlefield, and was missing his own 6th Division and a brigade from the 1st, so fielded only around 80,000 – hardly a massive margin of superiority considering that the allies were on the offensive.

So far as the study was concerned, we reviewed this background on our first evening in Spain at an introductory stand at the village of La Puebla de Arganzon, south of Vitoria, having flown into Bilbao from the UK that morning. The location also helps contextualise both the wider strategic situation and the lead-up to the battle. One of the village’s older buildings was once used as a fortified blockhouse by the French during their occupation of the Peninsula, with a Napoleonic imperial eagle badge still visible where it was carved into the wall more than 200 years ago. The fact the French had to rely on posts like these to protect their supply lines emphasises how much of a drain on French resources the Spanish guerilla war was, with thousands of troops tied down like this or committed to often-futile anti-guerilla sweeps. So far as the upcoming battle of Vitoria was concerned, however, La Puebla de Arganzon gave Wellington a crossing point over the River Zadorra that was not contested by the French, enabling him to shift a portion of his forces to the eastern bank unopposed, something that was essential to the plan of operations that he was developing.

The Battle of Vitoria

For our study, as for the two contending armies in 1813, Vitoria would be a long hot day. On the plus-side, however, we at least had no-one shooting at us!

The city of Vitoria is situated in a bowl in the mountains, with the River Zadorra running in from the north and out to the south, generally hugging the west side of the bowl. At the south-east corner the Great Road from Madrid enters where the river exits, via a narrow pass between the Puebla Heights to the east and the extensive hills to the west. Assuming that Wellington would advance directly up the road, the French armies were deployed one behind the other – Midi, Centre, and Portugal – to resist this. Although the strongest of the three French armies, Gazan’s Armée du Midi was badly stretched to cover the necessary frontage – only a brigade at Subijana de Alava covering the Puebla Heights, and outposts along the riverbank.

Wellington, however, decided to repeat at a tactical level what he had done on an operational level at the start of the campaign. He would feint with his right under Hill, whose command had used the bridge at La Puebla de Arganzon to cross to the eastern bank of the Zadorra, but move the bulk of his forces around to the left and turn the French right. A centre-right column under Wellington’s own direction with the 4th and Light Divisions would move by Nanclares and Villodas, leaving the Great Road to their right; a centre-left column under Dalhousie, with his own 7th Division and Picton’s 3rd, would go over the high ground and emerge further to the left; and Graham with the 1st and 5th Divisions plus Longa’s Spaniards would form the left-most column acting semi-independently against the French right-rear. Other than Hill, who had the 2nd and Portuguese Divisions plus Morillo’s Spaniards, the columns would all have to force a crossing of the Zadorra to engage the enemy and debouch onto the plain. Cavalry was attached to all the columns except Dalhousie’s, but the initial work, either to clear the Puebla Heights or cross the river, would be infantry work.

Focussing our first stand on issues of command and control, we began our day at the bridge of Villodas, putting ourselves in Wellington’s shoes. To begin with, things had gone well on the morning of 21 June 1813. At around 8.00 a.m. Morillo’s Spanish division opened Hill’s attack on the Puebla Heights, backed up by Stewart’s 2nd Division and Silveira’s Portuguese. Colonel Cadogan’s British brigade supported the Spanish on the heights themselves, the others pushed into Subijana de Alava, which had been vacated as the French brigade stationed there moved to Heights. This action drew away more and more French forces to the eastern flank of the battlefield, weakening the first line along the river facing the centre-right column.

Meanwhile, Graham’s command was coming into action on the far left, drawing away most of the Armée de Portugal but not as yet pressing its advance as the other columns were not in action. Of Dalhousie’s there was as yet no sign, his troops were still negotiating the mountains to the west.

Wellington already had the Light Division closed up with the river crossing around Villodas, but this bridge was covered by a company of French infantry. However, the command team then received word from a local peasant that the bridge at Tres Puentes, a shot way upstream, was unguarded. Kempt’s Brigade was sent on the rush to exploit this opportunity and crossed the bridge but was unable to progress any further as the high ground beyond was dominated by French artillery, fire from which killed the unfortunate local guide (a key military lesson here for the taking – never volunteer!).

Meanwhile, Sir Thomas Picton was not-so-quietly fuming at Mendoza, west of the Zadorra, waiting for Dalhousie and the 7th Division to come up. Being placed under the orders of Dalhousie, his superior by seniority but lacking Picton’s years of experience as a commander, did not improve the fiery Welshman’s temper. Dalhousie’s march had been delayed as he waited for his artillery to negotiate the rough ground. At length, Picton intercepted an aide with orders for Dalhousie to cross the river to support the attack of the Light and 4th Divisions. Unwilling to wait any longer, but also realising that he was well-placed to carry out the spirit of Wellington’s intent if not its actual detail, Picton took the responsibility upon himself of ordering a crossing of the bridge at Mendoza and adjacent fords by his own 3rd Division. This relieved the pressure on Kempt and allowed the whole of the Light Division to cross; however, in response, the Armée du Centre’s two French infantry divisions deployed to extend the right of the Armée du Midi: with the Armée de Portugal deployed against Graham, this committed most of the French infantry to action, leaving them with very limited reserves.

Having walked from Villodas to Tres Puentes to examine that portion of the field in detail, we now relocated by road to continue the story from the Knoll of Arinez, where the French command post was situated for much of the battle. This offers a good view of the Puebla Heights and of the loop of the Zadorra with the various crossing points; only the view of the far northern sector of the field, where Graham was facing Reille’s Armée de Portugal, is obscured.

Picking up the story at around 3.00p.m., Picton’s lead brigade under Colville was now heavily engaged, but was in the process of being relieved by Vandeleur’s Brigade of the Light Division and Grant’s of 7th (the remainder of the 7th was still well to the rear). To their right, Picton’s other two brigades under Brisbane and Power (the latter a Portuguese formation) were moving against Arinez with Kempt in reserve, engaging the right flank of the Armée du Midi. Grant was checked by heavy French gunfire, but Vandeleur pushed on and seized Arinez village.

By this time, British artillery under Colonel Alexander Dickson was coming into action against the French artillery concentration covering the Great Road between Arinez and the Puebla Heights. The 4th Division, having crossed at Nanclares now that bridge had been uncovered thanks to the other British crossings, moved up in support of the guns, and an extensive artillery duel commenced. With 75 allied guns in action against 76 French, this was one of the largest concentrations of firepower in the whole Peninsular War. At length, their gunners having won the artillery duel, the allied infantry went forward in force. This attack cracked the hinge between the Armées du Midi and de Centre, and the former was also under threat from Hill’s advance along the heights. What was more, the French were by now aware that there was a major threat to their rear, with rumours spreading through the ranks that the road to France was already cut. The result was a sudden collapse, with the French infantry and cavalry breaking and abandoning their artillery. To fully understand what had precipitated this, and the consequences, we now need to turn to the actions of Graham’s column.

Continuing the study after a break for lunch in Vitoria, we now drove out to the west of the city to the sector where Graham faced Reille. The latter, in order to fight a delaying action, had pushed part of his force across the Zadorra to engage the allies on the western bank – some members of the study felt that this was an error; others, including myself, argued that it was a good way of buying some extra time to make the best of things considering that only a weak force was at the French commander’s disposal. The result was an encounter battle in the morning as the two sides probed each other’s positions. At this stage, over and above Reille’s delaying tactics, Graham did not want to become heavily engaged with no sign of Dalhousie’s column coming into action as even this, his closest support, was some way off). Longa’s attached Spanish division was flung out to form the allied far-left, engaging their turncoat countrymen from the Armée du Centre at Durana where their fire could command, if not block, the road to France.

Under pressure as more of Graham’s troops came up, Reille fell back to the Zadorra bridge at Puente de Arriaga. The allied 1st Division followed up but was unable to force a crossing as the far bank was covered by a battery of 20 cannon; meanwhile, the 5th Division moved against Gamarra Mayor and its bridge, which were covered by the rest of Reille’s troops. Taking this crossing was a far better option for cutting off the French retreat, whereas the 1st Division’s axis of advance, even if pressed in the face of the French guns, would have led directly to Vitoria.

However, Gamarra Mayor proved a tough objective to crack. Robinson’s brigade eventually took the village, 1/4th leading the attack and 2/47th and 2/59th exploiting. This was achieved by around 2.00 p.m., but there was still no indication of Dalhousie coming into action, and a strong French force was evident on the other bank. When at length Robinson attempted to take the bridge, he was repulsed and then a French counterattack drove back into the village. If further evidence is needed of the fierce nature of the fighting, it is here that Bernard Cornwell chose to insert his fictional hero Richard Sharpe into the story! Only when the other French armies broke did Reille’s men fall back; this was immediately exploited by Anson’s Brigade of light dragoons, who cut up the French infantry though not without loss to themselves.

The bulk of the French army was now in full retreat. French casualties a little shy of 8,000, but a staggering 151 cannon were lost. Allied casualties amounted to around 5,000. The French baggage also fell into allied hands, but mass looting severely inhibited the pursuit and the army was not back on the road until 10.00 a.m. the following day. The detached divisions of the Armée de Portugal, off to the north, and Armée du Nord, off to the south-east, extricated themselves and withdrew to France. Undoubtedly this was a great triumph for the allies, completing the removal of the French forces from Spain and winning for Wellington the baton of a field marshal. However, it was also marred by the difficulties experienced in coordinating a multi-pronged attack and by the post-battle misconduct of many of the troops. Both of these factors would recur in different forms before the end of the fighting covered in our study.

The Battle of Sorauren

Having spent the first full day of the study covering one large battle in detail, the next day would be spent on the road, covering a lot of ground to look at three actions of steadily diminishing scale. Since geography trumped chronology with respect to the order in which we visited these sites, it is worth taking a moment to recap the post-Vitoria situation. Wellington’s pursuit, once he got it moving, pushed up as far as the French border, where two cavalry officers in a rowing boat briefly crossed the River Bidassoa in order to claim bragging rights as the first British soldiers on French soil. Wellington, however, had no intention of making a crossing in force for some time to come. The military and political situation in Northern Europe was still confused, and the last thing that Wellington wanted was to pursue into France only to discover that Napoleon had made peace with the northern allies and was able to turn his attention south again. Furthermore, the Spanish fortresses of Pamplona and San Sebastian were still held by strong French garrisons and ideally needed to be taken before an advance over the Pyrenees could be considered. Wellington therefore tasked Graham to besiege San Sebastian and his Spanish contingent to blockade Pamplona, while the rest of the army provided a defensive screen holding the mountain passes against a potential French counterattack.

On the French side, meanwhile, King Joseph and Marshal Jourdan had both been recalled to France in disgrace and Marshal Soult, who until early 1813 had commanded the Armée du Midi, was sent to command a new consolidated Armée d’Espagne. This was organised into three ‘lieutenancies’ or wings – corps in all but name, denominated as left, right and centre – plus a Reserve of infantry and two divisions of cavalry. With this force, Soult intended to mount a swift counteroffensive to break the blockade of Pamplona. Three of the four erstwhile army commanders received these lieutenancy commands, with Gazan reverting to his former role as Soult’s chief of staff. The new army structure came out looking like this:

• Left Wing – Clausel, 17,000
• Centre – d’Erlon, 21,000
• Right Wing – Reille, 17,000
• Reserve – Villatte, 12,000 including all the foreign troops.
• Cavalry – Pierre Soult and Treillard, 6,000

Total 73,000 active troops, plus garrisons etc.

Confusingly, the lieutenancies were not always deployed relative to one another in the manner that their names would suggest. The fact that Soult could nail together a new army and mount an offensive within weeks of taking command says a lot about his abilities as an organiser – not for nothing was he known as Iron Hand – but also the resources that the Napoleonic state could call upon in terms of conscript manpower and reserves of materiel; especially cannon at Bayonne to replace those lost at Vitoria.

Having taken command on 11 July 1813, Soult attacked ten days later. Vilatte’s Reserve replaced Reille’s Left Wing along the Bidassoa, making a feint attack towards San Sebastian, allowing Reille to join Clausel in a thrust against the Roncesvalles pass on the left of Wellington’s defensive cordon while d’Erlon attacked that of Maya in the centre of the allied line. Since Roncesvalles was not on the study itinerary, suffice it to say for now that Reille and Clausel cleared the pass, which had been defended by Lowry Cole’s reinforced 4th Division. Cole checked the French attacks, but, feeling himself exposed, fell back to link up with Picton’s 3rd Division. Notwithstanding instructions to the contrary, the two divisions fell all the way back to Pamplona. On the 27th, Soult and Clausel arrived at the village of Zabaldica and began to probe the allied positions there, but Soult elected not to attack until Reille was up. This allowed Wellington time to come up and take command, the cheers marking his arrival being heard from the French lines.

On the 28th, Reille arrived and took over the left of the French line around Zabaldica, pushing Foy’s division past the allied right to threaten Pamplona. This move was checked by Picton plus cavalry around Huarte, but Foy’s failure did not interfere with the progress of the main French attack. Here, between Sorauren and Zabaldica, the French right and allied left faced each other across a valley between two opposing ridge lines with the western edge of the battlefield delineated by the River Ulzama running at right angles to the opposing lines. Soult had planned to launch his attack at 1.00 p.m., but at noon, with the 6th Division under Denis Pack arriving on the field to bolster the allied left, Clausel went forwards early. Conroux’s division on the French right was checked, but Vandermaesen on the left of Clausel’s line was able to take advantage of a saddle linking the heights occupied by the French and allies, and to cause some disruption amongst the Portuguese before also being driven back. Pack, once he was up, left his Portuguese brigade under Madden on the far side of the river to harass the French in Sorauren village, and took his two British brigades across to support the defence and then launched an unsuccessful counterattack to try and retake Sorauren. By 4:00 p.m. it was clear to Soult and his subordinates that their troops were exhausted, and that any further attacks would be futile. During all of this, the two Spanish divisions of Enrique O’Donnel’s Army of Andalucia maintained the blockade of Pamplona but would have been available as a further reinforcement in case of need.

Meanwhile, as the study party would review at its next stand, d’Erlon had pushed through the Maya pass and driven back the 2nd and Light Divisions. Soult decided to link with him for a renewed offensive that would cut Graham’s troops at San Sebastian off from the rest of the allies. However, for Clausel and – especially – Reille to support this meant disengaging at Sorauren by moving across Wellinton’s front. Daybreak on the 30th found Reille’s three divisions badly exposed, and in the Second Battle of Sorauren they were routed by the attack of the 6th, 4th, and 3rd Divisions. Meanwhile, the 7th Division attacked Clausel’s retreating corps, adding further casualties. Foy’s division, which had been on the French far-left, was cut off and made a circuitous independent retreat; the remainder tumbled back to link up with d’Erlon and eventually escape back into France.

The Defence of the Pass of Maya

While Clausel and Reille had been attacking the pass of Roncesvalles on 25 July, d’Erlon attacked that of Maya. It was held by two brigades of Sir William Stewart’s 2nd Division under Lieutenant Colonel Cameron (vice Cadogan killed at Vitoria) and Major General William Pringle. The other two brigades were absent; Byng’s was at Roncesvalles, and Ashworth’s Portuguese were covering a lesser pass seven miles to the east. The pass was approached from the north by the main road (now a track) and a track (now the main road) – this change in the road net, and extensive tree growth since 1813, makes this a very hard field on which to orientate oneself. In terms of deployments, Cameron covered the main road while Pringle had outposts watching the track from a feature known as the Gorospil Knoll but the rest of his brigade was down in the village of Maya, separated from the pass by both distance and altitude.

With Stewart and corps commander Hill both absent, respectively disbelieving of the French threat and dealing with wider responsibilities, Pringle was in charge as senior officer present. Warned of possible troop movements, the picquet watching the track was reinforced by the four light companies of the brigade (that is to say, one from each line battalion plus a rifle company from the 5/60th). Moyle Sherer, of the 2/34th’s light company, assured by a staff officer that the French were making only ‘a trifling movement of no consequence’ was doubtful and was proved right. At around 10.30 a.m., the light companies of Armagnac’s division, leading d’Erlon’s advance, stormed the Gorospil Knoll and eliminated the outpost: of around 400 men posted there, 140 were captured and the rest killed or wounded.

Belatedly called up as reinforcements, Pringle’s battalions came up the hill piecemeal – first the 2/34th, then the 1/39th, while the 1/50th from Cameron’s Brigade was also called across in support. They did little to check Armagnac’s advance, and a second division under Abbè was now moving up too. The 1/28th and elements of the 1/92nd and 1/71st from Cameron’s Brigade were now also committed, but at length the redcoats were driven back – Pringle’s men down the slope, Cameron’s falling back on the rest of his brigade, which was now being pressed by Maransin’s division coming up the main road. Stewart arrived around 2:00 p.m. and ordered a withdrawal, which took place in stages down the main road. Four Portuguese cannon covering the road could not be extricated and were overrun. By keeping firing rather than pulling back they did considerable damage, leading to an interesting discussion amongst our modern study party as to whether it was ever acceptable in terms of contemporary doctrine to accept the loss of a weapons system if said system was able to inflict heavy losses on the enemy before being neutralised.

Around 3.30 p.m. the French made a renewed attack, by which time the 1/82nd from the 7th Division had arrived as reinforcement. Nevertheless, numbers told and the British fell back. When a renewed attack was made around 4.30 p.m., two more 7th Division battalions had arrived and d’Erlon – whose troops had broken ranks to loot the British camp – called it a day after becoming concerned that the whole 7th Division was threatening his flank. Hill came up in the evening and eventually concentrated his forces at Lizaso, some way to the rear, but was forced to fall back from there as well before d’Erlon was caught up in Soult’s post-Sorauren retreat and Hill could go over to the counteroffensive.

The fight at Maya had been a nasty setback for the allies. Although 2nd Division commander William Stewart was an astute tactician at a small-scale level, he was not capable of the initiative and self-possession required for effective independent command and Wellington had previously said that ‘It is necessary that Stewart should be under the particular charge of somebody’. On this occasion, however, with Hill absent that ‘somebody’ was missing. Pringle, meanwhile, was newly appointed to his brigade and unfamiliar with his command and the terrain. If Stewart himself should not have been left unsupervised, no more should he have left Pringle unsupervised. Yet so far as the blame-game is concerned, Wellington himself had overstretched his forces trying to cover two sieges at once. Commanders both at Maya and at Roncesvalles were left to their own devices, whereas ordinarily they were used to exact and detailed instructions. Better, under such circumstances, to dodge a decision than to make the wrong one and bring down the anger of the great man. This lack of initiative, both here and at Roncesvalles, badly compromised the initial responses to Soult’s offensive.

Cadoux’s Bridge

Our final stand of the day took a look at a small action containing clear lessons about leadership, command, and the difference between moral and physical courage – looking at the doomed defence of the bridge at Bera by a company of the 95th Rifles.

After the defeat of his Pyrenees offensive, Soult fell back and regrouped. The three ‘lieutenancies’ were rebuilt by transfers from the Reserve, and the army was soon ready to take the offensive again. A relief of Pamplona was clearly not going to be possible (though the garrison there held out until the end of October) but it might be possible to break the siege of San Sebastian by means of an attack across the lower Bidassoa. This river flows out of the mountains and then turns hard left just north of the village of Bera and flows out to the sea. From its bend to the sea, it largely formed the frontline between the two armies. Soult’s plan for 31 August was for d’Erlon with two divisions to feint towards the passes again, while Clausel with four divisions attacked immediately west of the Bidassoa bend and Reille with three plus the Reserve attacked lower down where Spanish troops were holding the heights of San Marcial.

Alas for Soult’s hopes, he had planned his attack for the same day that the allies stormed and took San Sebastian so it was all for nothing. What was more, the Spanish under Manuel Freire neatly defeated Reille’s attack at the Battle of San Marcial. Clausel, meanwhile, had three divisions across the river and a fourth in reserve but by afternoon was facing stiffer and stiffer resistance. Notified by Soult that Reille’s attack had failed, he began to pull back but this coincided with a torrential downpour that caused the river to rise dramatically. Clausel and two of his divisions got safely away across the fords, but as night fell the remaining division – Vandermaesen’s – was trapped on the wrong side of the Bidassoa. The obvious way back was the bridge at Bera, but this was covered by outposts of the Light Division on the right (east) bank.

The allied officer commanding in this sector was Major General John Byne Skerrett. Skerrett was an interesting contradiction – undoubted physical courage, but no moral courage to make a decision and stick to it. At Tarragona in 1811 he had refused to land troops to aid the Spanish defence; at Tarifa in 1812 he had wanted to evacuate the fortress only for his subordinates and allies to carry out a successful defence. He was a recent arrival in the Light Division, replacing Vandeleur transferred to the cavalry, and not a good pick for the role.

The company of 95th Rifles at the bridge was under Captain Daniel Cadoux. He was deployed to block a French attack east to west, using a stone-built house on the east bank as a strong point but prepared if necessary to withdraw across the bridge where a barricade had been set up at the other end. Any attack on this axis would in any case expose its flank to the rest of Skerrett’s Brigade up on the heights of Santa Barbara a couple of miles back. However, this arrangement was not ideally placed to defend against a French force trying to escape back from the ‘friendly’ west bank of the river, and Vandermaesen’s vanguard quickly seized the barricade and took out the two sentries there. However, even in the dark the rifles took their toll and began inflicting losses on the French who could only form three-abreast on the narrow bridge. Vandermaesen himself was shot down trying to get his troops across.

If Skerrett had brought down his whole brigade, the entire French force would have been cut off, but instead he remained immobile, unwilling to act, and ordered Cadoux to withdraw. This Cadoux refused to do, and all night the position was held. Daylight, however, revealed to the French the limited nature of the force opposing them, and they began to inflict heavy losses with musketry and then light guns. Skerrett again ordered a retreat and this time Cadoux complied, only to be shot dead as he mounted his horse. His company was near annihilated – 17 dead, 42 wounded, two missing – and there were a further 34 casualties in the two companies belatedly sent to cover his retreat. The French lost 231 men but the rest of the division made it to safety.

Widely blamed for Cadoux’s death and the escape of the French troops, Skerrett left the Peninsula shortly afterwards. Reassigned to the Low Countries expedition he seems to have set out to regain his reputation but was thwarted by circumstances – his death in action at Bergen op Zoom in March 1814 may have been a case of redemption-or-die-trying, but his actions actually helped foul up that operation too.

The Bidassoa and La Rhune

The final full day of our study was taken up by consideration of the first stages of Wellington’s invasion of France, which involved forcing a crossing of the Bidassoa and evicting the French from their positions on the last major line of Pyrenean foothills before the land drops away to the flat country stretching towards Bayonne. For our purposes, we were primarily looking at the logistical, political, and operational challenges rather than getting into the tactical details, so our two stands were focussed on locations that provided a good commanding view of the relevant terrain. We began in the morning by driving to Ibardin, which sits at the top of a pass right on the Franco-Spanish border, and hiking up to the heights to the west to the site of the Bayonette Redoubt, one of a number of French defensive positions constructed to impede an allied advance into France.

Soult had now switched to a defensive stance, occupying a series of redoubts from the sea at Hendaye on the Bidassoa estuary to the upper reaches of the River Nivelle (which runs parallel to the Bidassoa and flows into the sea at St Jean de Luz). He expected an allied attack from the east, via the passes, which Wellington fostered by having Hill demonstrate against the French forces there. In fact, the main effort was to come at the other end of the line where Graham’s corps was to cross the lower Bidassoa which was only covered by a single French division. At 7.25 a.m. on 7 October (10 minutes late, but they were waiting for the tide to completely fall) the 5th Division crossed the mudflats near Hendaye while the 1st attacked via the fords a little upriver at Behobie. The attack was then extended by the two Spanish divisions from the San Marcial heights. It was a complete success – the 5th Divison’s picquets were across the river without a shot fired – and wrecked Maucune’s French division.

Turning to our own viewpoint, this sector was defended by Taupin’s French division of Clausel’s ‘lieutenancy’. Caught in a pincer between the two attacking brigades of the Light Division, supported by Longa’s Spaniards, the 9e Légère holding our redoubt abandoned the position and fell back. To the east, the two Spanish Andalucian divisions, now under Pedro Giron in place of O’Donnell, outflanked the Greater Rhune feature. With Freire’s Spanish exploiting the initial gains to occupy the Mandale hill to the west of our viewpoint, the whole western portion of the French line was rendered untenable and Soult had to withdraw.

In the afternoon, we moved to the even higher vantage point of the Greater Rhune. This is accessible during the holiday season by a rack-railway, which we took full advantage of. During an earlier trip to plan the study the train had been closed, and so we had been obliged to make a partial ascent on foot, as far as the Mouiz Redoubt about halfway up the feature. We passed this again by train on the way up, before reviewing the final phases of the war from the summit.

After his defeat at the Bidassoa, Soult had fallen back to his next defensive line, south of the Rover Nivelle with extensive works covering St Jean de Luz. Pamplona having finally fallen, the Spanish troops there moved up to hold the passes, leaving Hill free to form the right wing of the attack, concentrating against d’Erlon on the upper Nivelle. On the sea-front, Sir John Hope, who had replaced Graham, made a feint attack that kept Reille pinned in his works. The main business would be in the centre, with the 3rd, 4th, 7th and Light Divisions, plus Giron’s and Longa’s Spaniards. The date of the offensive was 10 November, set back a day because Hill had been delayed by snow in the passes.

The Greater Rhune feature that we were using as our viewpoint had already been evacuated by the French, but the Lesser Rhune and the Mouiz Redoubt were still held by elements of Marinsin’s French division. These were taken by the Light Division hooking around the flank of the Greater Rhune. Colonel Colborne, commanding Skerrett’s former brigade, would take the Mouiz and Kempt the Lesser Rhune itself. This was done in good order and time, allowing elements of the division to assist the main attack to the east that drove back the rest of Clausel’s corps. Harry Smith of the 95th, Colborne’s brigade major, called it ‘the most beautiful attack ever made in the history of war’

Badly demoralised, both by their own defeat and the news of Napoleon’s disaster at Leipzig, the French fell back again, this time to positions around Bayonne. During the second week of December, Wellington’s Anglo-Portuguese troops closed in on the town, resisting a series of aggressive counterattacks by Soult who attempted to exploit the fact that Wellington’s army was divided by the River Nive (across which he could pass troops via the bridges in Bayonne). Thanks to an impressive stand by Hill’s isolated corps at St Pierre on 13 December, this plan was thwarted and the armies went briefly into winter quarters.

Come the new year, Wellington was able to invest Bayonne by having a bridge of boats constructed across the Adour estuary. This allowed him to invest the town with Hope’s corps and take the rest of the army after Soult, who, having sent troops north to aid Napoleon’s defence of France and thrown a division into Bayonne, now had a much smaller field army. Beating Soult again at Orthez on 27 February, the main body of the allied army ended the war with a final, if bloody, victory at Toulouse on 10 April.

San Sebastian

Although our stand on the summit of the Greater Rhune had taken us to the end of the story so far as our chronological coverage was concerned, we ended the study by retracing our steps in both time and geography by using the morning before our flight home to review the events surrounding the fall of San Sebastian. There is not a lot that can be taken from Napoleonic siege operations that is directly relevant to modern warfare, and in any case the sites of the siege batteries are now obscured by modern urban sprawl. However, the response to the failure of the first attempt to capture the fortress, and the disorder that followed in the wake of the second, successful, attack gave a lot of food for thought in terms of leadership, values, and standards.

As discussed above, the fortress, which sits on the coast just south of the Franco-Spanish border, was cut off after the French retreat post-Vitoria, but was held by a strong and resolute garrison under Général de Brigade Louis-Emmanuel Rey with 3,000 men manning 76 guns and with the potential for reinforcement by sea. Directed by Graham, siege operations began on 12 July. The 5th Division under Major General John Oswald plus Bradford’s Portuguese Brigade manned the siege lines with the 1st Division acting as a covering force.

Initial attempts sought to clear the approaches on the south bank of the Urumea River, which flows into the Bay of Biscay immediately north of the old walled town. Here, discovery of an underground drainage channel allowed a mine to be placed beneath the counterscarp. Meanwhile, the main effort north of the river had developed two breaches in the walls facing the Urumea, but to storm them required low tide for the troops to attack and was set for 5;00 a.m. – 20 minutes before dawn – on 25 July. Oswald was tasked with the tactical direction of the assault, allocating elements of three British and two Portuguese battalions to attack the breaches, and the 8° Cacadores to support.

It was not enough. Intended as a diversion, the mine did more damage than expected but the handful of Cacadores assigned to exploit it could not make a lodgement. The 3/1st, attacking the main breach, was shot to bits and the other attacking battalions also suffered heavily. Casualties amounted to around a thousand, including over 300 men taken prisoner. On the same day, Soult’s Pyrenees offensive opened and the siege had to be abandoned.

Siege operations resumed on 6 August, but the heavy guns had been withdrawn in response to Soult’s offensive and were not ready to start firing until the 26th. Wellington, Graham, and chief engineer Sir Richard Fletcher all agreed on a repeat of the original plan but with more troops. However, there was a feeling that the 5th Division had not done its best in the first attack and needed stiffening. A general order, insensitive in its tone even by Wellington’s standards, called for volunteers from other divisions to show the 5th how to storm a breach.

On the night of 26 August a naval contingent established a battery on Santa Clara island to enfilade the defenders. A battery was also placed well forward where the mine had been exploded during the first storm. The main attack was to go in as before, however, and was set for 31 August. This time, it would go in in daylight, in part to match the tide but also to improve coordination. The forlorn hope under Lieutenant Maguire reached the top of the breach but progressed no further. The two British brigades of the 5th Division were fed into the attack, then the volunteers from the other divisions. Sir James Leith, wounded at Salamanca, had returned to replace Oswald at the head of the 5th Division, but was soon wounded and Oswald resumed command.

In order to restore momentum to a stalling attack, Graham ordered the heavy guns to reopen fire and sweep the defenders from the walls, and ordered up Bradford’s Portuguese to reinforce the redcoats. This combined effort, and the chance explosion of a French powder magazine, was enough to push the defenders from the walls. The allies surged into the town, but Rey fell back to the castle with around a thousand men and continued the fight until 8 September when he finally surrendered. Under circumstances that remain unclear, the town was set on fire and was also thoroughly sacked and looted by the enraged allied troops, with scenes reminiscent of the sack of Badajoz the previous year. Allied casualties were heavy: 867 dead including chief engineer Fletcher, 1,416 wounded including Leith and Oswald, and 44 missing.

As I said at the conclusion of the first of these two Peninsular blogs, it is a particular privilege to revisit these sites in the company of the men and women of today’s British Army. Their good humour and enthusiasm were constant throughout. Again, I was made to feel extremely welcome, and, again, I learned an awful lot – this time about how the modern army plans and conducts operations. The extent to which all those involved threw themselves into the exercise was also marked. Syndicates had been tasked with preparing activities in advance of the study to pick up on the themes that we planned to cover, and the effort that had gone into this was remarkable: from a strong field, top prize has to go to the group tasked with looking at logistical issues, who had gone to the extent of baking hard-tack biscuits to an original 19th century recipe – truly going above and beyond, even if the end product was revolting!

Suggested Further Reading

Bamford, Andrew, Sickness, Suffering, and the Sword: The British Regiment on Campaign 1808-1815 (University of Oklahoma)

Collins, Bruce, Wellington and the Siege of San Sebastian 1813 (Pen and Sword)

Divall, Carole, Wellington and the Vitoria Campaign 1813: ‘Never A Finer Army’ (Pen and Sword)

Esdaile, Charles, Battles and Battlefields of the Peninsular War (Helion)

Linch, Kevin, The British Army: 1783-1815 (Pen and Sword)

Lipscombe, Nick, The Peninsular War Atlas (Osprey)

Muir, Rory, Wellington: The Path to Victory 1769-1814 (Yale)

Robertson, Ian C., Wellington Invades France: The Final Phase of the Peninsular War 1813-1814 (Greenhill)

Thompson, Mark S., Wellington’s Engineers: Military Engineering in the Peninsular War 1808-1814 (Pen and Sword)