A history of the Peninsular War, Vol. 6, September 1, 1812-August 5, 1813

Having completed the depressing chronicle of the leaguer of the Castle of Burgos, it is necessary that we should turn back for a week or two, to examine the changing aspect of external affairs, which had affected the general strategical position in Spain while the siege lingered on. It will be remembered that the original scheme for the campaign had been that, after the Army of Portugal had been pushed back to the Ebro, a great part of Wellington’s field force should return to Madrid, to pick up the divisions left there, and the corps of Hill, before Soult and King Joseph should begin to give trouble on the side of Valencia. Only a containing force was to have been left behind, to aid Castaños and his Galicians in keeping Clausel (or his successor Souham) out of mischief, while more important movements were on foot in New Castile[65]. As to the command of this containing force there was a difficulty: Wellington was not at all satisfied with the way in which Clinton had handled the troops left on the Douro in August, and it seems doubtful whether he wished to give him a far more important commission in the first half of October, when he was the natural person to be chosen for it. For at this time none of the other senior officers who had served in the recent campaigns were available: Beresford, Stapleton Cotton, and Picton were all three invalided at the moment. Graham, the best of them all, had returned to England. Hill could not be removed from the charge of his old army in the South. Of the five divisions before Burgos four were commanded by brigadiers acting in the position of locum tenens[66]: most of the brigades were, in a similar fashion, being worked by colonels for want of a sufficient number of major-generals to go round. But after October 11th the formal difficulty was solved: there was now available an officer whom it was fitting to leave in charge against Souham, viz. Sir Edward Paget, who came up to the front on that day, after having been three years absent from the Peninsula; he had last served under Wellington at the Passage of the Douro in 1809, where he lost an arm. He had now come out nominally to succeed Graham at the head of the 1st Division, but also with the commission to act as second-in-command in the event of the illness or disabling of the general-in-chief. Paget was known as a good soldier, and had served with distinction during the Corunna retreat; but he had never been trusted with the management of a large force, and had little knowledge either of the army which Wellington had trained, or of the parts of Spain in which the campaign was now going on. Moreover, Wellington disliked on principle the idea of having a second-in-command, occupying to a certain extent a position independent of his own.

The separate containing force never came into existence, because Wellington never left the North or returned to Madrid, though rumours were afloat after Paget’s arrival that he was about to do so without delay. Why, contrary to his own expressed intentions did he never go South? The answer must certainly be that down to a very late period of the siege he continued to keep in his mind the idea of returning to Madrid[67], but that he was distracted from his purpose by several considerations. The first was the abominable weather in October, which made the prospect of a long forced march distasteful—the army was sickly and would suffer. The second was his insufficient realization of the nearness of the danger in the South: he thought that Hill was in a less perilous position than was actually the case. The third was that he was beginning to mark the growing strength of the Army of Portugal, which lay in his own front, and to see that he could not hope to ‘contain’ it by any mere detachment, if he departed for the South with his main body. Souham’s host was no longer a spent force, which could be ignored as a source of danger; if Wellington took away two or three divisions, the officer left in charge in the North would be at once assailed by very superior numbers. He did not like the idea of trusting either Clinton or Paget with the conduct of a retreat before an enemy who would certainly press him fiercely. In addition, there was the lingering hope that Burgos might fall: if it were captured the situation would be much improved, since the allies could use it as a sort of point d’appui—to use the terminology of the day—on which the containing army could rest; and the enemy would be forced to detach heavily for the investment of the place.

By the day of the last assault (October 18) Wellington had clearly lingered too long, for very large reinforcements had just come up to join Souham, who on that morning was in a position to evict from before Burgos not only any mere ‘covering force’ that might be left opposite him, but the whole 35,000 men which formed the total of the allied host. For by now the Army of Portugal was recruited up to a strength of 38,000 men present with the colours: it had also just received the disposable battalions of the Bayonne Reserve, a strong brigade of 3,500 men under General Aussenac, and what was more important still—Caffarelli had appeared at Briviesca with the main field-force of the Army of the North. The operations of Home Popham, Mendizabal, Mina and Longa, which had detained and confused the French troops in Biscay and Navarre for the whole summer, had at last reached their limits of success, and having patched up affairs on the coast, and left some 20,000 men to hold the garrison places, and to curb the further raids of the British commodore and the Spanish bands, Caffarelli had come southwards with the whole of his cavalry—1,600 sabres,—three batteries, and nearly 10,000 infantry, forming the greater part of the divisions of Vandermaesen and Dumoustier. There were something like 50,000 French concentrated between Pancorbo and Briviesca on October 18th[68], while Wellington—allowing for his losses in the siege—had not more than 24,000 Anglo-Portuguese and 11,000 Spaniards in hand. The arrival of the Army of the North not only made the further continuance of the siege of Burgos impossible, but placed Wellington in a condition of great danger. It is clear that his campaign in the North had only been able to continue for so long as it did because Popham, Mina, and the Cantabrian bands had kept Caffarelli employed for a time that exceeded all reasonable expectation. It was really surprising that a small squadron and 10,000 half-organized troops, of whom a large part were undisciplined guerrillero bands, and the rest not much better, should have held the 37,000 men of the Army of the North in play for three months. But the combination of a mobile naval force, and of local levies who knew every goat-path of their native mountains, had proved efficacious in the extreme. It had taken Caffarelli the whole summer and much of the autumn to vindicate his position, and recover the more important strategical points in his wide domain. Even when he was marching on Burgos there was fierce fighting going on round Pampeluna, on which Mina had pressed in more closely when he heard of the departure of the general-in-chief. The Cantabrians and Navarrese did marvels for Wellington, and their work has never been properly acknowledged by British writers.

It is clear that the whole situation would have been different if Wellington had brought up to the North in September the 16,000 veteran British troops left at Madrid. For want of them he was in a state of hopeless inferiority to his immediate opponents. And yet at the same time they were useless at Madrid, because Hill—even with their aid—was not nearly strong enough to keep Soult and King Joseph in check. Wellington had in all at this time some 55,000 Anglo-Portuguese troops under arms and at the front; but they were so dispersed that on both theatres of war he was inferior to his enemies. He had 24,000 himself in Old Castile—Hill some 31,000 in New Castile; each of the two halves of the army could count on the help of some 10,000 or 12,000 Spaniards[69]. But of what avail were the 35,000 men of all nations at Burgos against the 50,000 French under Souham and Caffarelli, or the 43,000 men of all nations near Madrid against the 60,000 of Soult and the King? The whole situation would have been different if a superiority or even an equality of numbers had been established against one of the two French armies; and it is clear that such a combination could have been contrived, if Wellington had adopted other plans.

There can be no doubt that the excessive tardiness of Soult’s evacuation of Andalusia was the fact which caused the unlucky distribution of Wellington’s forces in October. The Marshal, it will be remembered, only left Granada on September 16th, and did not come into touch with the outlying cavalry of the Valencian Army till September 29-30, or reach Almanza, where he was in full connexion with Suchet and King Joseph, till October 2. There was, therefore, no threatening combination of enemies on the Valencia side till thirteen days after the siege of Burgos had begun; and Wellington did not know that it had come into existence till October 9th, when he received a dispatch from Hill informing him that the long foreseen, but long deferred, junction had taken place. If it had occurred—as it well might have—three weeks earlier, there would have been no siege of Burgos, and Wellington would have been at Madrid, after having contented himself with driving the army of Portugal beyond the Douro. It is probable that he would have done well, even at so late a date as October 9th, if he had recognized that the danger in front of Madrid was now pressing, and had abandoned the siege of Burgos, in order to make new arrangements. Souham was not yet in a condition to press him, and Caffarelli’s 10,000 men had not arrived on the scene. But the last twelve days spent before Burgos ruined his chances.

On getting Hill’s dispatch Wellington pondered much—but came to an unhappy decision. In his reply he wrote, ‘I cannot believe that Soult and the King can venture to move forward to attack you in the position on the Tagus, without having possession of the fortresses in the province of Murcia [Cartagena and Chinchilla] and of Alicante;—unless indeed they propose to give up Valencia entirely. They would in that case[70] bring with them a most overwhelming force, and you would probably have to retire in the direction given to General Alten [i.e. by the Guadarrama Pass, Villa Castin, and Arevalo] and I should then join you on the Adaja. If you retire in that direction, destroy the new bridge at Almaraz.... I write this, as I always do, to provide for every event, not believing that these instructions are at all necessary[71].’

In a supplementary letter, dated two days later, Wellington tells Hill that he imagines that the autumn rains, which have made the siege of Burgos so difficult, will probably have rendered the rivers of the South impassable. ‘I should think that you will have the Tagus in such a state as to feel in no apprehension in regard to the enemy’s operations, be his numbers what they may[72].’ Yet though he does not consider the danger in the South immediate or pressing, he acknowledges that he ought to bring the siege of Burgos to an end, even though it be necessary to raise it, and to give up the hope of its capture. But the continual storms and rains induced him to delay his departure toward the South: ‘I shall do so as soon as the weather holds up a little.’ On the same day (October 12) he wrote to Popham to say that if he had to march towards Madrid, he expected that Souham would follow him, but that Caffarelli must on no account be allowed to accompany Souham. At all costs more trouble must be made in Cantabria and Biscay, to prevent the Army of the North from moving. Popham must not withdraw his squadron, or cease from stimulating the Northern insurgents[73]. Wellington was not aware that Caffarelli was already on the move, and that the diversion in the North—through no fault of the officers in charge of it—had reached its limit of success.

The moment had now arrived at which it was necessary at all costs to come to some decision as to the movements of the Army at Burgos, for Caffarelli (though Wellington knew it not) had started for Briviesca, while Soult and King Joseph were also getting ready for an immediate advance on Madrid, which must bring matters in the South to a head. But relying on letters from Hill, dated October 10th, which stated that there was still no signs of movement opposite him, Wellington resolved on October 14 to stay yet another week in the North, and try his final assault on Burgos: it came off with no success (as will be remembered) on the 18th of that month. He was not blind to the possible consequences to Hill of a prompt advance of the French armies from Valencia, but he persuaded himself that the weather would make it difficult, and that he had means of detaining Soult and Joseph, if they should, after all, begin the move which he doubted that they proposed to make.

The scheme for stopping any forward march on the part of the French had two sections. The first was to be executed by the Anglo-Spanish Army at Alicante. Maitland had fallen sick, and this force of some 16,000 men was now under the charge of General John Mackenzie. To this officer Wellington wrote on October 13th: ‘In case the enemy should advance from Valencia into La Mancha, with a view to attack our troops on the Tagus, you must endeavour to obtain possession of the town and kingdom of Valencia.’ He was given the option of marching by land from Alicante, or of putting his men on shipboard and making a descent from the sea on Valencia[74]. The second diversion was to be executed by Ballasteros and the Army of Andalusia. Wellington wrote to him that he should advance from Granada, cross the Sierra Morena with all his available strength, and place himself at Alcaraz in La Mancha[75]. He ought to have at least 12,000 men, as some of the Cadiz troops were coming up to join him. Such a force, if placed on the flank of Soult and the King, when they should move forward for Almanza and Albacete, could not be ignored. And Wellington hoped to reinforce Ballasteros with that part of the Murcian Army, under Elio and Freire, which had got separated from the Alicante force, and had fallen back into the inland. There were also irregular bodies, such as the division of the Empecinado, which could be called in to join, and if 20,000 men lay at Alcaraz, Soult could not go in full force to assail Hill in front of Madrid. He must make such a large detachment to watch the Spaniards that he would not have more than 30,000 or 35,000 men to make the frontal attack on the Anglo-Portuguese force, which would defend the line of the Tagus. Against such numbers Hill could easily hold his ground.

It may be objected to these schemes that they did not allow sufficient consideration to the power of Suchet in Valencia. Mackenzie’s force, of a very heterogeneous kind, was not capable of driving in the detachments of the Army of Aragon, which still lay cantoned along the Xucar, facing the Allies at Alicante. Suchet was able to hold his own, without detaining any of the troops of Soult or of King Joseph from their advance upon Madrid. Mackenzie’s projected expedition against the city of Valencia had no chance of putting a check upon the main manœuvre that the French had in hand.

But the use of Ballasteros, whose movement must certainly have exercised an immense restraining power upon Soult and King Joseph, if only it had been carried out, was an experiment of a kind which Wellington had tried before, nearly always with disappointing results. To entrust to a Spanish general an essential part of a wide strategical plan had proved ere now a doubtful expedient. And Ballasteros had always shown himself self-willed if energetic; it was dangerous to reckon upon him as a loyal and intelligent assistant in the great game. And at this moment the captain-general of Andalusia was in the most perverse of moods. His ill temper was caused by the recent appointment of Wellington as commander-in-chief of all the Spanish armies, a measure to which the Cortes had at last consented, after the consequences of the battle of Salamanca and the occupation of Madrid had prepared public opinion for this momentous step. It had been bitterly opposed at the secret session when it was brought forward, but there could no longer be any valid excuse for putting off the obviously necessary policy of combining all military effort in the Peninsula, by placing the control of the whole of the Spanish armies in the hands of one who had shown himself such a master of the art of war.

Wellington’s nomination as Generalissimo of the Spanish armies had been voted by the Cortes on September 22nd, and conveyed to him before Burgos on October 2nd. He had every intention of accepting the offer, though (as he wrote to his brother Henry[76]) the change in his position would not be so great in reality as in form, since Castaños and most of the other Spanish generals had of late been wont to consult him on their movements, and generally to fall in with his views. There had been exceptions, even of late, such as Joseph O’Donnell’s gratuitous forcing on of battle at Castalla, when he had been specially asked to hold back till the Anglo-Sicilian expedition began to work upon the East Coast. But it would certainly be advantageous that, for the future, he should be able to issue orders instead of advice. Meanwhile he could not formally accept the post of general-in-chief without the official leave of the Prince Regent, and prompt information of the offer and a request for permission to accept it, were sent to London. The Cortes, foreseeing the necessary delay, had refrained from publishing the decree till it should be certain that Wellington was prepared to assume the position that was offered him. But the fact soon became known in Cadiz, and was openly spoken of in the public press. Wellington continued to write only letters of advice to the Spanish generals, and did not assume the tone of a commander-in-chief as yet, but his advice had already a more binding force. There was no opposition made to him, save in one quarter—but that was the most important one. Ballasteros, as Commander of the 4th Army and Captain-General of Andalusia, burst out into open revolt against the Cortes, the Regency, and the commander-in-chief elect.

This busy and ambitious man had taken up his abode in Granada, after Soult’s departure on September 23rd, and since then had not stirred, though Wellington had repeatedly asked him to advance into La Mancha with his available force, and to put himself in touch with Elio, Penne Villemur, and Hill. But Ballasteros was suffering from an acute attack of megalomania: after years of skulking in the mountains and forced marches, he was now in the position of a viceroy commanding the resources of a great province. Though Andalusia had been cleared of the French by no merit of his, but as a side-effect of the battle of Salamanca, he gave himself the airs of a conqueror and deliverer, and fully believed himself to be the most important person in Spain. As an acute observer remarked, ‘Ballasteros wanted to begin where Bonaparte ended, by seizing supreme authority, though he had performed no such services for his fatherland as the French general. He spared no effort to attach his army to his person, and to win its favour sacrificed without hesitation the whole civil population. The land was drained of all resources, but he found new means of extorting money. His officers went from town to town exacting a so-called “voluntary contribution” for the army. Those who gave much were put in his white book as patriots—those who did not were enrolled in his black book as supporters of King Joseph. He wanted to collect all the troops in Andalusia into a great army, with himself as commander. But the Regency directed him, in accordance with Wellington’s advice, to march at once with the troops immediately available and take post in La Mancha[77].’

Then came the news of Wellington’s nomination as commander-in-chief. Ballasteros thought the moment favourable for an open bid for the dictatorship. Instead of obeying the orders sent, he issued on October 23rd a manifesto directed to the Regency, in which he declared that Wellington’s appointment to supreme power was an insult to the Spanish nation, and especially to the Spanish Army. He openly contemned the decree, as dishonourable and debasing—Spaniards should never become like Portuguese the servants of the foreigner. If—what he could not believe—the national army and the nation itself should ratify such an appointment, he himself would throw up his post and retire to his home. At the same time there was an outburst of pamphlets and newspaper articles inspired by him, stating that England was to be feared as an oppressor no less than France; and all sorts of absurd rumours were put about as to the intentions of the British Government and the servility of the Cortes.

Ballasteros, however, had altogether mistaken his own importance and popularity: budding Bonapartes must be able to rely on their own troops, and he—as the event showed—could not. The Cortes took prompt measures for his suppression: one Colonel Rivera, a well-known Liberal, was sent secretly to Granada to bear to Virues, the second in command of the 4th Army, orders to arrest his chief, and place him in confinement. And this was done without any difficulty, so unfounded was the confidence which Ballasteros had placed in his omnipotence with the army. Virues and the Prince of Anglona, his two divisional generals, carried out the arrest in the simplest fashion. On the morning of October 30th they ordered out their troops for a field day on the Alcalá road, save a battalion of the Spanish Guards, who had not belonged to Ballasteros’s old army, and had no affection for him. This regiment surrounded his residence, and when he issued out the picquet refused him passage, and Colonel Rivera presented him with the warrant for his arrest. There was some little stir in Granada among the civil population, but the army made no movement when Virues read the decree of the Cortes to them. Before he well understood what had happened, Ballasteros was on his way under a guard to the African fortress of Ceuta. He was afterwards given Fregenal in Estremadura as the place of his detention.

But all this happened on October 30th, and meanwhile the Army of Andalusia had remained motionless, though Wellington had believed that it had marched for La Mancha as early as October 5th[78]. Nearly a month had been lost by Ballasteros’s perversity, and when he was finally arrested and his troops became available, it was too late; for Hill had been forced to retreat, and Madrid was about to fall into the hands of the French, who had advanced through La Mancha unmolested. The Fourth Army was put under the command of the Duke del Parque, the victor of Tamames, but the vanquished of Alba de Tormes in the campaign of 1809. It marched early in November to its appointed position at Alcaraz; but it was of no use to have it there when the enemy had changed all his positions, and had driven Hill beyond the Guadarrama into the valley of the Douro.

Wellington’s arrangements for the defence of Madrid were therefore insufficient as the event proved, for three reasons. The diversion from Alicante was too weak—Suchet alone was able to keep Maitland in check. The main Spanish force which was to have co-operated with Hill never put in an appearance, thanks to the disloyal perversity of its general. And, thirdly, the autumn rains, which had so incommoded Wellington before Burgos, turned out to be late and scanty in New Castile. The Tagus and its affluents remained low and fordable almost everywhere.

As late as the 17th October Wellington had no fears about Hill’s position. The probability that the enemy from Valencia would advance upon Madrid, as he wrote on that morning, seemed diminishing day by day—reinforcements were coming up from the South to Hill, and Ballasteros was believed to be already posted in La Mancha, ‘which renders the enemy’s movement upon the Tagus very improbable[79].’ On the 19th, when his last assault on the Castle of Burgos had just been beaten off, and he was thinking of retreat, came much less satisfactory news from Hill. The enemy seemed to be drawing together: they had laid siege to the Castle of Chinchilla, and it was not impossible that they were aiming at Madrid, perhaps only with a view of forcing the abandonment of the siege of Burgos, perhaps with serious offensive intention. If so, Wellington might be caught with not one but both of the halves of his army exposed to immediate attack by a superior enemy, who was just assuming the offensive. And the total force of the French was appalling—50,000 men on the Northern field of operations, 60,000 on the other—while the united strength of all the available troops under Wellington’s orders was about 80,000, of whom 25,000 were Spaniards, whose previous record was not in many cases reassuring. He had often warned the ministers at home that if the French evacuated outlying provinces, and collected in great masses, they were too many for him[80]. But that ever-possible event had been so long delayed, that Wellington had gone on, with a healthy and cheerful opportunism, facing the actual situation of affairs without taking too much thought for the morrow. Now the conjunction least to be desired appeared to be at length coming into practical existence and it remained to be seen what could be done.

As for his own army, there was no doubt that prompt retreat, first beyond the Pisuerga, then beyond the Douro, was necessary. Souham had already made his first preliminary forward movement. On the evening of October 18th a strong advance guard, consisting of a brigade of Maucune’s division, pushed in upon the front of the allied army at the village of Santa Olalla, and surprised there the outlying picket of an officer and thirty men of the Brunswick-Oels, who were nearly all taken prisoners. The heights above this place commanded the town of Monasterio, wherefore Wellington directed it to be evacuated, and drew in its line to a position slightly nearer Burgos, for it was clear that there were heavy forces behind the brigade which had formed the French attacking force[81].

On the 19th he was in order of battle with all his army except the troops left before Burgos—Pack’s Portuguese, a brigade of the 6th Division, and three weak battalions from the 1st Division. The position extended from Ibeas on the Arlanzon, through Riobena to Soto Palacios, and was manned by the 1st, 5th, and 7th Divisions, two brigades of the 6th, Bradford’s Portuguese, Castaños’s Galicians, and all the cavalry—about 30,000 men. Wellington expected to be attacked on the 20th, and was not prepared to give back till he had ascertained what force was in his front. He knew that the Army of Portugal had been reinforced, but was not sure whether the report that Caffarelli had come up with a large portion of the Army of the North was correct or no.

Souham’s movement on the 18th had been intended as the commencement of a general advance: he brought up his whole force to Monasterio, and he would have attacked on the 20th if he had not received, before daybreak, and long ere his troops were under way, a much-delayed dispatch from King Joseph. It informed him that the whole mass of troops from Valencia was on the march for Madrid, that this movement would force Wellington to abandon the siege of Burgos and to fall back, and that he was therefore not to risk a general action, but to advance with caution, and pursue the allied force in front of him so soon as it should begin to retreat, when it might be pressed with advantage.[82] He was to be prepared to link his advance with that of the armies from the South, which would have columns in the direction of Cuenca. Souham was discontented with the gist of this dispatch, since he had his whole army assembled, while Caffarelli was close behind, within supporting distance, at Briviesca. He thought that he could have fought with advantage, and was probably right, since he and Caffarelli had a marked superiority of numbers, though Wellington’s positions were strong.

Balked of the battle that he desired, he resolved to make a very strong reconnaissance against the allied centre, to see whether the enemy might already be contemplating retreat, and might consent to be pushed back, and to abandon Burgos. This reconnaissance was conducted by Maucune’s and Chauvel’s (late Bonnet’s) divisions and a brigade of light cavalry: it was directed against Wellington’s centre, in front of the villages of Quintana Palla and Olmos, where the 7th division was posted. Maucune’s troops, forming the front line of the French, were hotly engaged in this direction, when Wellington, seeing that the main body of the enemy was very remote, and that the two vanguard divisions had ventured far forward into his ground, directed Sir Edward Paget with the 1st and 5th divisions—forming his own left wing—to swing forward by a diagonal movement and take Maucune in flank. The French general discovered the approaching force only just in time, and retreated in great haste across the fields, without calling in his tirailleurs or forming any regular order of march. Each regiment made off by its own way, just early enough to escape Paget’s turning movement. The British horse artillery arrived only in time to shell the last battalions; the infantry was too far off to reach them. Dusk fell at this moment, and Paget halted: if there had been one hour more of daylight Maucune and Chauvel would have been in an evil case, for they were still some distance from their own main body on the Monasterio position, and the British cavalry was coming up in haste. The reconnaissance had been pushed recklessly and too far—in accordance with the usual conduct of Maucune, who was (as his conduct at Salamanca had proved) a gallant but a very rash leader. The losses on both sides were trifling—apparently about 80 in the French ranks[83], and 47 in the British 7th division. Paget’s troops did not come under fire.

On the morning of the 21st Wellington received a letter of Hill’s, written on the 17th, which made it clear to him that he must delay no longer. It reported that the enemy were very clearly on the move—the whole of Drouet’s corps was coming forward from Albacete, the castle of Chinchilla had fallen into the hands of the French on Oct. 9th, and—what was the most discomposing to learn—Ballasteros had failed to advance into La Mancha: by the last accounts he was still at Granada. He would certainly come too late, if he came at all. The only compensating piece of good news was that Skerrett’s brigade from Cadiz was at last near at hand, and would reach Talavera on the 20th. Elio, from his forward position at Villares, reported that Suchet’s troops opposite Alicante were drawing back towards Valencia; possibly they were about to join Soult and King Joseph, who were evidently coming forward. Hill was inclined to sum up the news as follows: ‘The King, Soult, and Suchet having united their armies are on the frontiers of Murcia and Valencia, and appear to be moving this way. It is certain that a considerable force is advancing toward Madrid, but I think it doubtful whether they will attempt to force their way to the capital.’ Meanwhile he had, as a precautionary measure, brought up the troops in cantonments round Madrid to Aranjuez and its neighbourhood, and had thrown forward a cavalry screen beyond the Tagus, to watch alike the roads from Albacete and those from Cuenca.

The mere possibility that Soult and the King were advancing on Madrid was enough to move Wellington to instant retreat: if he waited for certainty, he might be too late. And his own position in front of Souham and Caffarelli was dangerous enough, even if no more bad news from the South should come to hand. Accordingly on the afternoon of the 21st orders were sent to Pack to prepare to raise the siege of Burgos, when the main army in its backward movement should have passed him. The baggage was to be sent off, the stores removed or burned in the following night. The divisions in line opposite Souham were to move away, after lighting camp fires to delude the enemy, in two main columns, one on each side of the Arlanzon. The northern column, composed of the 5th Division, two-thirds of the Galician infantry, the heavy dragoons of Ponsonby (late Le Marchant’s brigade) and the handful of Spanish regular cavalry, was to skirt the northern side of the city of Burgos, using the cross-roads by Bivar and Quintana Dueñas, and to retire to Tardajos, beside the Urbel river. The southern and larger column, consisting of the 1st, 7th, and 6th Divisions (marching in that order), of Bradford’s Portuguese and the remainder of the Galicians, was to move by the high road through Villa Fria, to cross the town-bridge of Burgos (with special caution that silence must be observed, so as not to alarm the garrison of the Castle) to Frandovinez and Villa de Buniel. Here they would be in touch with the other column, which was to be only two miles away at Tardajos. Anson’s cavalry brigade was to cover the rear of this column, only abandoning the outposts when the infantry should have got far forward. They were to keep the old picquet-line till three in the morning. Bock’s German dragoons and the cavalry of Julian Sanchez formed a flank guard for the southern column: they were to cross the Arlanzon at Ibeas, five miles east of Burgos, and to retire parallel to the infantry line of march, by a circuit on cross-roads south of Burgos, finally joining the main body at Villa de Buniel[84].

All this complicated set of movements was carried out with complete success: the army got away without arousing the attention of the French of Souham, and the column which crossed the bridge of Burgos, under the cannon of the Castle[85], was never detected, till a Spanish guerrilla party galloped noisily across the stones and drew some harmless shots. Since the enemy was alarmed, the rear of the column had to take a side-path. Pack filed off the troops in the trenches, and got away quietly from the investment line. In the morning the two infantry columns were safely concentrated within two miles of each other at Tardajos and Villa de Buniel. There being no sign of pursuit, they were allowed some hours’ rest in these positions, and then resumed their retreat, the northern column leading, with the 5th Division at its head, the southern column following, and falling into the main road by crossing the bridge of Buniel. The 7th Division formed the rearguard of the infantry; it was followed by Bock’s dragoons, Anson and Julian Sanchez bringing up the rear. On the night of the 22nd the army bivouacked along the road about Celada, Villapequeña, and Hornillos; the light cavalry, far behind, were still observing the passages of the Arlanzon at Buniel and of the Urbel at Tardajos.

The French pursuit was not urged with any vigour on this day. The departure of Wellington had only been discovered in the early morning hours by the dying down of the camp fires along his old position[86]. At dawn Souham advanced with caution; finding nothing in front of it, his vanguard under Maucune entered Burgos about 10 o’clock in the morning, and exchanged congratulations with Dubreton and his gallant garrison. The light cavalry of the Army of Portugal pushed out along the roads north and south of the Arlanzon: those on the right bank found the three broken 18-pounders of Wellington’s battering train a few miles outside of Burgos, and went as far as the Urbel, where they came on the rear vedettes of the Allies. Those on the left bank obtained touch with Anson’s outlying picquet at San Mames, and drove it back to the Buniel bridge.

The 23rd October, however, was to be a day of a much more lively sort. Souham and Caffarelli, having debouched from Burgos, brought all their numerous cavalry to the front, and at dawn it was coming up for the pursuit, in immense strength. There were present of the Army of Portugal Curto’s light horse, Boyer’s dragoons, and the brigade formerly commanded by Chauvel now under Colonel Merlin of the 1st Hussars. These, by the morning state of October 15th, made up 4,300 sabres. The Army of the North contributed the bulk of its one—but very powerful—cavalry brigade, that of Laferrière, 1,650 strong.[87] Thus there were nearly 6,000 veteran horsemen hurrying on to molest Wellington’s rear, where the covering force only consisted of Anson’s and Bock’s two brigades and of the lancers of Julian Sanchez—1,300 dragoons British and German[88], and 1,000 Spaniards admirable for raids and ambushes but not fit to be placed in battle line. Ponsonby’s dragoons and the regular squadrons of the Galician army were far away with the head of the column. To support the cavalry screen were the two horse-artillery batteries of Downman and Bull, and the two Light Battalions of the German Legion from the 7th Division, under Colonel Halkett, which were left behind as the extreme rearguard of the infantry.

All day the main marching column of the allied army laboured forward unmolested along the muddy high road from Celada del Camino to Torquemada, a very long stage of some 26 miles: Wellington seldom asked his infantry to make such an effort. What would have happened to Clausel’s army if the British had marched at this rate in the week that followed the battle of Salamanca? Behind the infantry column, however, the squadrons of the rearguard were fighting hard all day long, to secure an unmolested retreat for their comrades. This was the most harassing and one of the most costly efforts that the British cavalry was called upon to make during the Peninsular War. The nearest parallel to it in earlier years was the rearguard action of El Bodon in September 1811[89]. But there the forces engaged on both sides had been much smaller.

The French advanced guard consisted of Curto’s light cavalry division of the Army of Portugal, supported by Maucune’s infantry division. It had started at dawn, and found the British cavalry vedettes where they had been marked down on the preceding night, at the bridge of Villa Buniel. They retired without giving trouble, being in no strength, and the morning was wearing on when the French came upon Wellington’s rearguard, the horsemen of Anson and Julian Sanchez, and Halkett’s two German battalions, a mile or two east of Celada, holding the line of the Hormaza stream. Anson was in position behind its ravine, with a battalion of the light infantry dispersed along the bushes above the water, and the cavalry in support. Julian Sanchez was visible on the other side of the Arlanzon, which is not here fordable, beyond Anson’s right: on his left, hovering on hills above the road, was a small irregular force, the guerrilla band of Marquinez[90]. The ground along the Hormaza was strong, and the infantry fire surprised the enemy, who were stopped for some time. They came on presently in force, and engaged in a bickering fight: several attempts to cross the ravine were foiled by partial charges of some of Anson’s squadrons. Wellington says in his dispatch that the skirmishing lasted nearly three hours, and that Stapleton Cotton, who had recently come up from the Salamanca hospital cured of his wound, made excellent dispositions. But the odds—more than two to one—were far too great to permit the contest to be maintained for an indefinite period; and when, after much bickering, the French (O’Shee’s brigade of Curto’s division) at last succeeded in gaining a footing beyond the Hormaza, Anson’s squadrons went off in good order. The ground for the next five or six miles was very unfavourable for a small detaining force, as the valley of the Arlanzon widened out for a time, and allowed the enemy space to deploy his superior numbers. Nevertheless the light dragoons turned again and again, and charged with more or less success the pursuing Chasseurs. While Curto’s division was pressing back the British brigade in front, Merlin’s brigade, pushing out on the French right, ascended the hills beyond Anson’s left flank, and there found and drove off the partida of Marquinez. The fugitive guerrilleros, falling back for support towards the British, came in upon the flank of the 16th Light Dragoons with French hussars in close pursuit, and mixed with them. The left squadron of the 16th was thrown into confusion, and suffered heavily, having some thirty casualties; and the commander of the regiment, Colonel Pelly, and seven of his men were taken prisoners. As Curto was pushing on at the same time, Anson’s brigade was much troubled, and was greatly relieved when it at last came in sight of its support, Bock’s German Heavy Dragoons and Bull’s battery, posted behind a bridge spanning a watercourse near the lonely house called the Venta del Pozo. The two battalions of Halkett, who had been retiring under cover of Anson’s stubborn resistance, were now some little way behind Bock, near the village of Villadrigo. On seeing the obstacle of the watercourse in front of them, and the British reserves behind it, the cavalry of the Army of Portugal halted, and began to re-form. Anson’s harassed brigade was permitted to cross the bridge and to join Bock.

At this moment there came on the scene the French cavalry reserves, Boyer’s dragoons and the three regiments of the Army of the North, which were commanded that day not by their brigadier Laferrière (who had been injured by a fall from his horse) but by Colonel Faverot of the 15th Chasseurs. Souham, who had arrived in their company as had Caffarelli also, gave orders that the pursuit was not to slacken for a minute. While the cavalry of Curto and Merlin were re-forming and resting, the newly arrived squadrons were directed to drive in the British rearguard from its new position. The brigade of Faverot was to attack in front, the dragoons of Boyer were to turn the hostile line, by crossing the watercourse some way to their right, and to fall on its flank and rear. So great was the superiority in numbers of the attacking party that they came on with supreme confidence, ignoring all disadvantages of ground. Faverot’s brigade undertook to pass the bridge in column, in face of Bock’s two regiments drawn up at the head of the slope, and of Bull’s battery placed on the high road so as to command the crossing. Boyer’s dragoons trotted off to the right, to look for the first available place where the water should allow them a practicable ford.

What followed was not in accordance with the designs of either party. Souham had intended Boyer and Faverot to act simultaneously; but the former found the obstacle less and less inviting the farther that he went on: he rode for more than a mile without discovering a passage, and finally got out of sight of Souham and the high road[91]. Meanwhile Faverot had advanced straight for the bridge, and had begun to cross it. Of his three regiments the Berg lancers led, the 15th Chasseurs came next, the Legion of Gendarmerie brought up the rear. They were ten squadrons in all, or some 1,200 sabres. As each squadron got quit of the bridge it formed up in line, the first to pass to the right of the road, while those which followed successively took ground to the left, between the chaussée and the bank of the Arlanzon. Eight squadrons had got into position before the British line gave any signs of movement on the hillside above.

It is clear that to cross a bridge in this fashion was a reckless manœuvre. If Bock had charged when two or three squadrons only had passed[92], he must have crushed them in the act of deployment, and have jammed the rest of the French column at the bridgehead in a position of helpless immobility. That no such charge was made resulted from a curious chance. Stapleton Cotton, who was still conducting the retreat, had placed Bull’s battery at a point on the slope where he judged that it fully commanded the bridge, with Bock’s four squadrons on its right. Then Anson’s brigade, not in the best of order after its long pursuit by the enemy, trotted over the bridge and came up the slope. Cotton intended that it should deploy on the left of the guns, but the retreating regiments, before the directions reached them, turned to the right and began to form up behind Bock’s line. The general at once sent them orders to cross to the left and prolong the line on the other side of Bull’s battery. By some extraordinary blunder the leading regiment (perhaps to shorten its route) passed to the left in front of the guns, instead of moving behind them. It was thus masking the battery just at the moment that the French, to the surprise of every one, began to cross the bridge. There was considerable confusion in changing the march of the light dragoons, and before they were cleared off and the front of the battery was free, several French squadrons were across the ravine and forming up. The artillerymen then opened, but in their hurry they misjudged the elevation, and not one shot of the first discharge told. The second did hardly better; more French were passing every moment, and Cotton then directed the cavalry to charge, since the enemy was becoming stronger and stronger on the near side of the bridge.

The charge was delivered in échelon of brigades; Bock’s two regiments, who had been awaiting orders for some time, got away at once, and fell upon the left of the French—the gendarmes and the left squadrons of the Chasseurs. Anson’s six squadrons started somewhat later, and having tired horses—after their long morning’s fight—came on at a much slower pace. They found themselves opposed to the Lancers and the right squadrons of the Chasseurs. There was such a perceptible space between the two attacks, that some of the French narratives speak of the British charge as being made by a front line and a reserve. The numbers engaged were not very unequal[93]—probably 1,200 French to 1,000 British—but the former were all fresh, while the larger half of the latter (Anson’s brigade) were wearied out—man and horse—by long previous fighting.

All accounts—English and French—agree that this was one of the most furious cavalry mêlées ever seen: the two sides broke each other’s lines at various points—the 1st K.G.L. Dragoons rode down their immediate opponents, the 2nd got to a standing fight with theirs: Anson’s regiments, coming on more slowly on their jaded mounts, made no impression on the Lancers and Chasseurs opposite them. The whole mass fell into a heaving crowd ‘so completely mixed that friend could hardly be distinguished from foe—the contest man to man lasted probably a long minute, during which the ground was strewn with French, and our own loss was heavy in the extreme.’ General Bock was at one time seen defending himself against six Frenchmen, and was barely saved by his men. Beteille, the Colonel of the French Gendarme Legion, received, it is said, twelve separate wounds, and was left for dead. The combat was ended by the intervention of the two rear squadrons of the Gendarmes, who had not crossed the bridge when the charge began; they came up late, and fell upon the flank and rear of the 2nd Dragoons of the K.G.L. This last push settled the matter, and the British brigades broke and fell back[94].

They were not so disheartened, however, but that they rallied half a mile to the rear, and were showing fight when Boyer’s dragoons came tardily upon the scene—they had at last found a passage over the ravine, and appeared in full strength on the flank. It was hopeless to oppose them, and the wrecks of the two British brigades fell back hastily towards their infantry support, Halkett’s two Light Battalions of the German Legion, which had been retreating meanwhile towards Villadrigo, marching in column at quarter distance and prepared to form square when necessary. The need now came—the leading regiment of Boyer’s Dragoons turned upon the rear battalion—the 1st—and charged it: the square was formed in good time, and the attack was beaten off. The battalion then retired, falling back upon the 2nd, which had nearly reached Villadrigo. Both then retreated together, covering the broken cavalry, and had gone a short distance farther when the French renewed the attack. Both battalions formed square, both were charged, and both repulsed the attack of the dragoons with loss. The enemy, who had still plenty of intact squadrons, seemed to be contemplating a third charge, but hesitated, and finally Bock’s and Anson’s men having got back into some sort of order, and showing a front once more, the squares retired unmolested, and marched off in company with the cavalry. The French followed cautiously and gave no further trouble. It was now dark, and when the pursuit ceased the rearguard halted for two hours, and resuming its march about ten o’clock, finally reached the bridge of Quintana del Puente, outside Torquemada, at two o’clock next morning.

Thus ended a costly affair, which might have proved much more perilous had the French been better managed. But though their troopers fought well enough, their generals failed to get such advantage as they should out of their superior numbers. It is to be noted that both Curto and Boyer have very poor records in the memoirs of Marmont, Foy, and other contemporary writers, who speak on other occasions of the ‘inertie coupable’ and ‘faute de décision’ of the one, and call the other ‘bon manœuvrier, mais n’ayant pas la réputation qui attire la confiance aveugle du soldat’: Napoleon once summed them up as ‘mauvais ou médiocres[95], along with several other generals. It seems clear, at any rate, that they should have accomplished more, with the means at their disposal. The damage inflicted on the British rearguard was much smaller than might have been expected—only 230 in all, including 5 officers and 60 men taken prisoners[96]. It is probable indeed that the pursuers suffered no less; for, though their official report spoke of no more than about a hundred and fifty casualties, Faverot’s Brigade alone had almost as many by itself, showing a list of 7 killed and 18 officers and 116 men wounded. But the regimental lists give 35 French cavalry officers in all disabled that day, 5 in Curto’s division, 5 in Merlin’s Brigade, 5 in Boyer’s Dragoons, and 2 on the staff, beside the 18 in Faverot’s Brigade. This must indicate a total loss of nearly 300, applying the very moderate percentage of casualties of men to officers that prevailed in Faverot’s regiments to the other corps. The engagement had been most honourable to the two cavalry brigades of Wellington’s rearguard, and not less so to the German Legion light infantry. An intelligent observer—not a soldier—who saw the whole of the fighting that day remarked, ‘I twice thought Anson’s Brigade (which was weak in numbers and much exhausted by constant service) would have been annihilated; and I believe we owe the preservation of that and of the heavy German brigade to the admirable steadiness of Halkett’s two Light Battalions.... We literally had to fight our way for four miles, retiring, halting, charging, and again retiring. I say we because I was in the thick of it, and never witnessed such a scene of anxiety, uproar, and confusion. Throughout the whole of this trying occasion Stapleton Cotton behaved with great coolness, judgement, and gallantry. I was close to him the whole time, and did not observe him for an instant disturbed or confused[97].’ At the end of the fight one officer in a jaded regiment observed to another that it had been a bad day. His friend replied that it had been a most honourable day for the troops, for at nightfall every unwounded man on an efficient horse was still in line and ready to charge again. There had been no rout, no straggling, and no loss of morale.

While the combat of Venta del Pozo was in progress the main body of the army had been resting in and about Torquemada, undisturbed by any pursuit, so well had the rearguard played its part. Next morning (October 24th) it marched off and crossed the Carrion river by the bridges of Palencia, Villamuriel, and Dueñas. Behind this stream Wellington was proposing to check the enemy for at least some days. He chose it, rather than the Pisuerga, as his defensive line for tactical reasons. If he had stayed behind the Pisuerga, he would have had the Carrion and its bridges in his immediate rear, and he was determined not to fight with bridge-defiles behind him. In the rear of the Carrion there were no such dangerous passages, and the way was clear to Valladolid. It was fortunate that, after the lively day on the 23rd, the French made no serious attempt at pursuit upon the 24th, for there were many stragglers that morning from the infantry. Torquemada was the centre of a wine-district and full of barrels, vats, and even cisterns of the heady new vintage of 1812; many of the men had repaid themselves for a twenty-six mile march by over-deep potations, and it was very hard to get the battalions started next morning.[98] Napier says that ‘twelve thousand men at one time were in a state of helpless inebriety:’ this is no doubt an exaggeration, but the diaries of eye-witnesses make it clear that there was much drunkenness that day.

Wellington’s new position extended from Palencia on the Carrion to Dueñas on the Pisuerga, below its junction with the first-named river. He intended to have all the bridges destroyed, viz. those of Palencia, Villa Muriel, and San Isidro on the Carrion, and those of Dueñas and Tariego on the Pisuerga. He placed one of the Galician divisions (Cabrera’s) in Palencia and south of it, supported by the 3/1st from the 5th Division. Another Galician division (Losada’s) continued the line southward to Villa Muriel, where it was taken up by the 5th Division—now under General Oswald who had recently joined. The 1st, 6th, and 7th Divisions prolonged the line down the Pisuerga. Behind the rivers, all the way, there was a second line of defence, formed by the dry bed of the Canal of Castile, which runs for many miles through the valley: it was out of order and waterless, but the depression of its course made a deep trench running parallel with the Carrion and Pisuerga, and very tenable. It must be noted that Palencia lay to the wrong side of the Carrion for Wellington’s purpose, being on its left or eastern bank. It had, therefore, either to be held as a sort of tête de pont, or evacuated before the enemy should attack. The Spanish officer in command resolved to take the former course, though the town was indefensible, with a ruinous mediaeval wall: he occupied it, and made arrangements for the blowing up of the bridge only when his advanced guard should have been driven out of Palencia and should have recrossed the river.

Souham, who had pushed his head-quarters forward to Magaz this day, resolved to try to force the line of the Carrion. Two divisions under Foy—his own and that of Bonté (late Thomières)—were to endeavour to carry Palencia and its bridge. Maucune, with the old advanced guard, his own division and that of Gauthier[99], was to see if anything could be done at the bridges of Villa Muriel and San Isidro. Foy had Laferrière’s cavalry brigade given him, Maucune retained Curto’s light horse. The main body of the army remained in a mass near Magaz, ready to support either of the advanced columns, if it should succeed in forcing a passage.

Unfortunately for Wellington, everything went wrong at Palencia. Foy, marching rapidly on the city, drove away the few squadrons of Galician cavalry which were observing his front, and then burst open one of its rickety gates with artillery fire, and stormed the entry with Chemineau’s brigade. The Galician battalions in the town, beaten in a short street fight, were evicted with such a furious rush that the British engineer officer with the party of the 3/1st, who were working at the bridge, failed[100] to fire the mine, whose fuse went wrong. Most of them were taken prisoners. Colonel Campbell of the 3/1st, who had been sent to support the Spaniards, fell back towards Villa Muriel, judging the enemy too strong to be resisted, and the routed Galicians had to be covered by Ponsonby’s heavy dragoons from the pursuit of the French cavalry, who spread over the environs of Palencia and captured some baggage trains, British and Spanish, with their escorts.

Thus Souham had secured a safe passage over the Carrion and it mattered little to him that his other attack farther south had less decisive success. Here Maucune, always a very enterprising—not to say reckless—commander had marched against the two bridges of Villa Muriel and San Isidro. He led his own division towards the former, and sent that of Gauthier against the latter. The 5th Division had an infantry screen of light troops beyond the river—this was pushed in, with some loss, especially to the 8th Caçadores. Both bridges were then attacked, but each was blown up when the heads of the French columns approached, and a heavy fire, both of artillery and of musketry, from the other bank checked their further advance. Gauthier then turned east in search of a passage, and went coasting along the bank of the Pisuerga for some miles, as far as the bridge of Tariego (or Baños, as the British accounts call it). This had been prepared for destruction like the other bridges, but the mine when exploded only broke the parapets and part of the flooring. The French charged across, on the uninjured crown of the arch, and captured some 40 of the working party[101]: they established themselves on the opposite bank, but advanced no farther; here the British had still the lower course of the Pisuerga to protect them, while Gauthier had no supports near, and halted, content to have captured the bridge, which he began to repair.

Meanwhile, the bridge of Villa Muriel having been more efficiently blown up, Maucune dispersed his voltigeur companies along the river-bank, brought his artillery up to a favourable position, and entered on a long desultory skirmish with Oswald, which lasted for some hours. During this time he was searching for fords—several were found[102], but all were deep and dangerous. About three o’clock in the afternoon a squadron of cavalry forded one of them, apparently unobserved, reached the western bank, and rolled up a company of the 1/9th which it caught strung out along the river bickering with the skirmishers on the other side. An officer and 33 men were taken prisoners[103]. This passage was followed by that of eight voltigeur companies, who established themselves on the other side, but clung to the bank of the river, the British light troops having only retired a few score of yards from it. Having thus got a lodgement beyond the Carrion, Maucune resolved to cross in force, in support of his voltigeurs. Arnauld’s brigade passed at the ford which had first been used, Montfort’s at another, to the left of the broken bridge, farther down the stream. The French seized the village of Villa Muriel, and established themselves in force along the dry bed of the canal, a little to its front. Wellington had expected that Maucune would come on farther, and intended to charge him with the whole 5th Division when he should begin to ascend the slope above the canal. But when, after an hour or more, the enemy made no advance, it became necessary to assume the offensive against him, since Foy was now across the Carrion at Palencia, and, if he commenced to push forward, he might drive the Spanish troops between him and Villa Muriel into Maucune’s arms. The latter must be expelled before this danger should arise: wherefore at about 4 o’clock a brigade of Losada’s Spaniards was sent against the right of the French division, while Pringle’s brigade attacked its front along the bed of the canal. The Spaniards made no progress—they were indeed driven back, and rallied with difficulty by Wellington’s friend General Alava, who was severely wounded. To replace them the British brigade of Barnes was brought up, and told to storm Villa Muriel—Spry’s Portuguese supporting. After a stiff fight Pringle and Barnes swept all before them: the enemy yielded first on the right, but held out for some time in the village, where he lost a certain number of prisoners. He finally repassed the fords, and sought safety in the plain on the eastern bank, where the French reserves were now beginning to appear in immense force. ‘Their numbers were so great that the fields appeared to our view almost black[104].’ Wellington wrote home that he had hitherto under-estimated Souham’s force, and only now realized that the Army of the North had brought up its infantry as well as its cavalry. He, naturally, made no attempt to pursue the enemy beyond the Carrion, and the fight died down into a distant cannonade across the water[105].

The combat of Villa Muriel cost the 5th Division some 500 casualties—not including 43 men of the 3/1st, hurt or taken in the separate fight at Palencia, or a few casualties in Ponsonby’s dragoons in that same quarter. Maucune’s losses were probably not far different—they must certainly have exceeded the ‘250 killed or wounded, 30 prisoners and 6 drowned’ of the official report: for one of Maucune’s regiments—the 15th Line—had 200 casualties by itself[106]. Foy at Palencia lost very few men—though his statement of ‘three or four’ as their total can hardly be correct. He had taken 100 prisoners (27 of them British), and killed or wounded 60 more, mostly Galicians[107]. The balance of the day’s losses was certainly against the Allies.

Moreover, despite of Maucune’s repulse, the result of the day’s fighting was much to Wellington’s detriment, since (though the loss of the bridge of Tariego mattered little) the capture of Palencia ruined his scheme for defending the line of the Carrion. Souham’s whole army could follow Foy, and turn the left flank of the Allies behind that river. Wherefore Wellington first threw back the left flank of the 5th Division en potence, to make a containing line against Foy, while its right flank still held the line of the river about Villa Muriel. Under cover of this rearguard, the remainder of the army evacuated its positions on each side of Dueñas at dawn on the 26th, and marched down the Pisuerga to Cabezon, where it passed the river at the bridge of that place and retired to the opposite side. The 5th Division followed (unmolested by the enemy) in the early morning: at night the entire allied army had retired across the broad stream, and was lining its left bank from Valladolid upward, with that city (held by the 7th Division) as the supporting point of its southern flank, and its northern resting on the Cubillas river. This was one of the most surprising and ingenious movements that Wellington ever carried out. On the 23rd he was defending the right bank of the Pisuerga—on the 24th he was on the other side and defending its left bank in a much stronger position. He was in no way sacrificing his communications with Portugal or Madrid, since he had behind him two good bridges over the Douro, those of Tudela and Puente Duero, with excellent roads southward and westward to Medina del Campo, Arevalo, and Olmedo. The enemy might refuse to attack him, and march down the west bank of the Pisuerga towards the Douro, but Wellington had already provided for the destruction of the three bridges of Simancas, Tordesillas, and Toro, and the Douro was now a very fierce and broad barrier to further progress, twice as difficult to cross as the Pisuerga. On the other hand, if Souham should make up his mind to come down the Pisuerga on its eastern bank, by the restored bridge of Tariego, there was an excellent fighting position along the Cubillas river. But the roads on this side of the Pisuerga were bad, and the country rough, while on the western bank the ground was flat and fertile, and the line of march easy.