SECTION III: CHAPTER IV

DUPONT IN ANDALUSIA: THE CAPITULATION OF BAYLEN

We left General Dupont at Andujar, on the upper course of the Guadalquivir, whither he had retired on June 19 after evacuating Cordova. Deeply troubled by the interruption of his communications with Madrid, and by the growing strength displayed by the Spanish army in his front, he had resolved that it was necessary to draw back to the foot of the Sierra Morena, and to recover at all costs his touch with the main French army in the capital. He kept sending to Murat (or rather to Savary, who had now superseded the Grand-Duke) persistent demands for new orders and for large reinforcements. Most of his messengers were cut off on the way by the insurgents, but his situation had become known at head quarters, and was engrossing much of Savary’s attention—more of it indeed than Napoleon approved. The Emperor wrote on July 13 that the decisive point was for the moment in Castile, and not in Andalusia, and that the best way to strengthen Dupont was to reinforce Bessières[141].

Such had not been Savary’s opinion: frightened at the isolation in which Dupont now lay, he sent to his assistance the second division of his corps, 6,000 men under General Vedel, all recruits of the ‘legions of reserve,’ save one single battalion of Swiss troops. The division was accompanied by Boussard’s cavalry, the 6th Provisional Dragoons, some 600 strong. Vedel made his way through La Mancha without difficulty, but on entering the Despe?a Perros defiles found his passage disputed by a body of insurgents—2,000 peasants with four antique cannon—who had stockaded themselves in the midst of the pass. A resolute attack scattered them in a few minutes, and on reaching La Carolina on the southern slope of[p. 177] the mountains Vedel got in touch with Dupont, who had hitherto no notice of his approach [June 27].

Instead of leaving the newly arrived division to guard the passes, Dupont called it down to join him in the valley of the Guadalquivir. With the assistance of Vedel’s troops he considered himself strong enough to make head against the Spanish army under Casta?os, which was commencing to draw near to Andujar. Keeping his original force at that town—a great centre of roads, but a malarious spot whose hospitals were already crowded with 600 sick,—he placed Vedel at Baylen, a place sixteen miles further east, but still in the plain, though the foot-hills of the Sierra Morena begin to rise just behind it. To assert himself and strike terror into the insurgents, Dupont ordered one of Vedel’s brigades to make a forced march to Jaen, the capital of a province and a considerable focus of rebellion. This expedition scattered the local levies, took and sacked Jaen, and then returned in safety to Baylen [July 2-3].

Meanwhile Casta?os was drawing near: he had now had a month in which to organize his army. Like Blake in Galicia, he had used the recruits of Andalusia to fill up the gaps in the depleted battalions of the regular army. But less fortunate than his colleague in the north, he had not been able to prevent the Juntas of Seville and Granada from creating a number of new volunteer corps, and had been obliged to incorporate them in his field army, where they were a source of weakness rather than of strength. His total force was some 33,000 or 34,000 men, of whom 2,600 were cavalry, for in this arm he was far better provided than was the army of the North. The whole was organized in four divisions, under Generals Reding, Coupigny, Felix Jones (an Irish officer, in spite of his Welsh name), and La Pe?a. In addition there was a flying brigade of new levies under Colonel Cruz-Murgeon, which was pushed forward along the roots of the mountains, at a considerable distance in front of the main body: it was ordered to harass Dupont’s northern flank and to cut his communications with Baylen and La Carolina.

With 16,000 or 17,000 men, including nearly 3,500 cavalry, Dupont ought to have been able to contain Casta?os, if not to beat him. The proportion of his forces to those of the enemy was not much less than that which Bessières had possessed at Medina de Rio Seco. But, unfortunately for himself and his[p. 178] master, Dupont was far from possessing the boldness and the skill of the marshal. By assuming not a vigorous offensive but a timid defensive along a protracted front, he threw away his chances. The line which he had resolved to hold was that of the Upper Guadalquivir, from Andujar to the next passage up the river, the ferry of Mengibar, eight miles from Baylen. This gave a front of some fifteen miles to hold: but unfortunately even when drawn out to this length the two divisions of Barbou and Vedel did not cover all the possible lines of attack which Casta?os might adopt. He might still march past them and cut them off from the defiles of the Morena, by going a little higher up the river and crossing it near Baeza and Ubeda. Dupont was wrong to take this line of defence at all: unless he was prepared to attack the army of Andalusia in the open, he should have retired to Baylen or to La Carolina, where he would have been able to cover the passes for as long as he might choose, since he could not have had either of his flanks turned.

Meanwhile he was gratified to hear that further reinforcements were being sent to him. Unreasonably disquieted about Andalusia, as Napoleon thought, Savary proceeded to send a third division to aid Dupont. This was Gobert’s, the second of Moncey’s corps: it started from Madrid not quite complete, and left strong detachments at the more important towns along the road through La Mancha. Though originally seventeen battalions strong, it reached the northern slope of the Sierra Morena with only ten. Savary had not intended it to go any further: he had told Dupont that it was to be used to cover his retreat, if a retreat became necessary, but not for active operations in Andalusia. But disregarding these directions Dupont commanded Gobert to cross the Morena and come down to join Vedel: this he did, bringing with him nine ‘provisional battalions[142]’ and the second provisional regiment of cuirassiers, perhaps 5,000 men in all. There were now over 20,000 French on the south side of the mountain, a force amply sufficient to deal with Casta?os and his 33,000 Andalusians [July 7]. But[p. 179] they were still widely scattered. Dupont lay at Andujar with 9,000 or 10,000 sabres and bayonets: Vedel was sixteen miles away at Baylen, with 6,000 men, of whom 2,000 under General Liger-Belair were pushed forward to the ferry of Mengibar. Gobert was at La Carolina, at the foot of the passes, with five battalions about him, and a sixth encamped on the summit of the defile. He had sent forward the remainder of his division (the four battalions of the sixth provisional regiment, and half the second provisional cuirassiers) to join Dupont at Andujar, so that he had not more than 2,800 bayonets and 350 cavalry with him.

Casta?os, meanwhile, had brought up his whole army, with the exception of the flying corps of Cruz-Murgeon, to a line close in front of Andujar: the heads of his columns were at Arjona and Arjonilla, only five miles from Dupont. On July 11 the Spanish generals held a council of war at Porcu?a, and drew out their plan of operations. Since the enemy seemed to be still quiescent, they resolved to attack him in his chosen position behind the river. Casta?os, in person—with the divisions of Jones and La Pe?a, 12,000 strong—undertook to keep Dupont employed, by delivering an attack on Andujar, which he did not intend to press home unless he got good news from his second and third columns. Meanwhile, six miles up the river, Coupigny with the second division, nearly 8,000 strong, was to attempt to cross the Guadalquivir by the ford of Villa Nueva. Lastly, Reding with the first division, the best and most numerous of the whole army, 10,000 strong, was to seize the ferry of Mengibar and march on Baylen. Here he was to be joined by Coupigny, and the two corps were then to fall upon the rear of Dupont’s position at Andujar, while Casta?os was besetting it in front. It was their aim to surround and capture the whole of the French division, if its general did not move away before the encircling movement was complete. Meanwhile the flying column of Cruz-Murgeon, about 3,000 strong, was to cross the Guadalquivir below Andujar, throw itself into the mountains in the north, and join hands with Reding and Coupigny behind the back of Dupont.

This plan, though ultimately crowned with success, was perilous in the highest degree. But Casta?os had seriously underestimated the total force of Dupont, as well as misconceived his exact position. He was under the impression that the main body of the French, which he did not calculate at more than 12,000 or 14,000 men,[p. 180] was concentrated at Andujar, and that there were nothing more than weak detachments at Mengibar, Baylen, or La Carolina. These, he imagined, could not stand before Reding, and when the latter had once got to the northern bank of the river, he would easily clear the way for Coupigny to cross. But as a matter of fact Vedel had 6,000 men at Mengibar and Baylen, with 3,000 more under Gobert within a short march of him. If the Spanish plan had been punctually carried out, Reding should have suffered a severe check at the hands of these two divisions, while Dupont could easily have dealt with Casta?os at Andujar. Coupigny, if he got across at Villa Nueva, while the divisions on each side of him were beaten off, would have been in a very compromised position, and could not have dared to push forward. But in this curious campaign the probable never happened, and everything went in the most unforeseen fashion.

On July 13 the Spanish plan began to be carried out, Reding marching for Mengibar and Coupigny for Villa Nueva. Casta?os kept quiet at Arjonilla, till his lieutenants should have reached the points which they were to attack. On the same day Dupont received the news of Moncey’s repulse before Valencia, and made up his mind that he must persevere in his defensive attitude, without making any attempt to mass his troops and fall upon the enemy in his front[143]. Just at the moment when his enemies were putting the game into his hands, by dividing themselves into three columns separated from each other by considerable gaps, he relinquished every intention of taking advantage of their fault.

On July 14 Reding appeared in front of the ferry of Mengibar, and pushed back beyond the river the outlying pickets of Liger-Belair’s detachment. He made no further attempt to press the French, but Dupont, disquieted about an attack on this point, ordered Gobert to bring down the remains of his division to Baylen, to join Vedel. Next morning the Spaniards began to develop their whole plan: Casta?os appeared on a long front opposite Andujar, and made a great demonstration against the position of Dupont, using all his artillery and showing heads of columns at several points. Coupigny came down to the river[p. 181] at Villa Nueva, and got engaged with a detachment which was sent out from Andujar to hold the ford. Reding, making a serious attempt to push forward, crossed the Guadalquivir at Mengibar and attacked Liger-Belair. But Vedel came up to the support of his lieutenant, and when the Swiss general found, quite contrary to his expectation, a whole division deployed against him, he ceased to press his advance, and retired once more beyond the river.

Nothing decisive had yet happened: but the next day was to be far more important. The operations opened with two gross faults made by the French: Dupont had been so much impressed with the demonstration made against him by Casta?os, that he judged himself hopelessly outnumbered at Andujar, and sent to Vedel for reinforcements. He bade him send a battalion or two, or even a whole brigade, if the force that he had fought at Mengibar seemed weak and unenterprising[144]. This was an error, for Casta?os only outnumbered the French at Andujar by two or three thousand men, and was not really to be feared. But Vedel made a worse slip: despising Reding overmuch, he marched on Baylen, not with one brigade, but with his whole division, save the original detachment of two battalions under Liger-Belair which remained to watch Mengibar. Starting at midnight, he reached Andujar at two on the afternoon of the sixteenth, to find that Casta?os had done no more than repeat his demonstration of the previous day, and had been easily held back. Cruz-Murgeon’s levies, which the Spanish general had pushed over the river below Andujar, had received a sharp repulse when they tried to molest Dupont’s flank. Coupigny had made an even feebler show than his chief at the ford of Villa Nueva, and had not passed the Guadalquivir.

But Reding, on the morning of the sixteenth, had woken up to unexpected vigour. He had forded the river near Mengibar, and fallen on Liger-Belair’s detachment for the second time. Hard pressed, the French brigadier had sent for succour to Baylen, whither Gobert had moved down when Vedel marched for Andujar. The newly arrived general came quickly to the aid of the compromised detachment, but he was very weak, for he had left a battalion at La Carolina and sent another with a squadron of cuirassiers to Li?ares, to guard against a rumoured movement of the Spaniards along the Upper Guadalquivir. He only brought with him three battalions and 200 cavalry, and this was not[p. 182] enough to contain Reding. The 4,000 men of the two French detachments were outnumbered by more than two to one; they suffered a thorough defeat, and Gobert was mortally wounded. His brigadier, Dufour, who took over the command, fell back on Baylen, eight miles to the rear. Next morning, though not pressed by Reding, he retired towards La Carolina, to prevent himself being cut off from the passes, for he credited a false rumour that the Spaniards were detaching troops by way of Li?ares to seize the Despe?a Perros.

Dupont heard of Gobert’s defeat on the evening of the sixteenth. It deranged all his plans, for it showed him that the enemy were not massed in front of Andujar, as he supposed, but had a large force far up the river. Two courses were open to him—either to march on Baylen with his whole army in order to attack Reding, and to reopen the communications with La Carolina and the passes, or to fall upon Casta?os and the troops in his immediate front. An enterprising officer would probably have taken the latter alternative, and could not have failed of success, for the whole French army in Andalusia save the troops of Belair and Dufour was now concentrated at Andujar, and not less than 15,000 bayonets and 3,000 sabres were available for an attack on Casta?os’ 12,000 men[145]. Even if Coupigny joined his chief, the French would have almost an equality in numbers and a great superiority in cavalry and guns. There cannot be the slightest doubt that the Spaniards[p. 183] would have suffered a defeat, and then it would have been possible to expel Reding from Baylen without any danger of interference from other quarters.

But, in a moment of evil inspiration, Dupont chose to deprive himself of the advantage of having practically his whole army concentrated on one spot, and determined to copy the error of the Spaniards by splitting his force into two equal halves. He resolved to retain his defensive position in front of Andujar, and to keep there his original force—Barbou’s infantry and Frésia’s horse. But Vedel with his own men, the four battalions from Gobert’s division which were at Andujar, and 600 cavalry, was sent off to Baylen, where he was directed to rally the beaten troops of Dufour and Liger-Belair, and then to fall upon Reding and chase him back beyond the Guadalquivir[146].

On the morning, therefore, of July 17 Vedel set out with some 6,000 men and marched to Baylen. Arriving there he found that Dufour had evacuated the place, and had hurried on to La Carolina, on the false hypothesis that Reding had pushed past him to seize the passes. As a matter of fact the Spaniard had done nothing of the kind: after his success at Mengibar, he had simply retired to his camp by the river, and given his men twenty-four hours’ rest. It was a strange way to employ the day after a victory—but his quiescence chanced to have the most fortunate[p. 184] effect. Vedel, on hearing that Dufour had hastened away to defend La Carolina and the passes, resolved to follow him. He was so inexcusably negligent that he did not even send a cavalry reconnaissance towards Mengibar, to find out whether any Spanish force remained there. Had he done so, he would have found Reding’s whole division enjoying their well-earned siesta! In the direction of La Carolina and the passes there was no enemy save a small flanking column of 1,800 raw levies under the Count of Valdeca?as, which lay somewhere near Li?ares.
Map of the battle of Baylen

Enlarge  Battle of Baylen July 19, 1808, at the moment of Dupont’s third attack.
Map of Andalusia

Enlarge  Part of Andalusia, between Andujar and the Passes. July 19, 1808.

On the night of the seventeenth, Vedel and his men, tired out by a long march of over twenty miles, slept at Guarroman, halfway between Baylen and La Carolina. Dufour and Liger-Belair had reached the last-named place and Santa Elena, and had found no Spaniards near them. On the morning of the eighteenth Vedel followed them, and united his troops to theirs. He had then some 10,000 or 11,000 men concentrated in and about La Carolina, with one single battalion left at Guarroman to keep up his touch with Dupont. The latter had been entirely deceived by the false news which Vedel had sent him from Baylen—to the effect that Reding and his corps had marched for the passes, in order to cut the French communications with Madrid. Believing the story, he forwarded to his subordinate an approval of his disastrous movement[147], and bade him ‘instantly attack and crush the Spanish force before him, and after disposing of it return as quickly as possible to Andujar, to deal with the troops of the enemy in that direction.’ Unfortunately, as we have seen, there was no Spanish corps at all in front of Vedel; but by the time that he discovered the fact it was too late for him to rejoin Dupont without a battle[148]. His[p. 185] troops were tired out with two night marches: there were no supplies of food to be got anywhere but at La Carolina, and he decided that he must halt for at least twelve hours before returning to join Dupont.

Meanwhile, on the morning of the eighteenth, Reding’s 9,500 men, of whom 750 were cavalry, had been joined by Coupigny and the second Andalusian division, which amounted to 7,300 foot and 500 horse. Advancing from Mengibar to attack Baylen, they found to their surprise that the place was unoccupied: Vedel’s rearguard had left it on the previous afternoon. Reding intended to march on Andujar from the rear on the next day, being under the full belief that Vedel was still with Dupont, and that the troops which had retired on La Carolina were only the fragments of Gobert’s force. For Casta?os and his colleagues had drawn up their plan of operations on the hypothesis that the enemy were still concentrated at Andujar.

Reding therefore, with some 17,000 men, encamped in and about Baylen, intending to start at daybreak on July 19, and to fall on Dupont from behind, while his chief assailed him in front. But already before the sun was up, musket-shots from his pickets to the west announced that the French were approaching from that direction. It was with the head and not with the rear of Dupont’s column that Casta?os’ first and second divisions were to be engaged, for the enemy had evacuated Andujar, and was in full march for Baylen.

On the night of the seventeenth Dupont had received the news that Vedel had evacuated Baylen and gone off to the north-east, so that a gap of thirty miles or more now separated him from his lieutenant. He had at first been pleased with the move, as we have seen: but presently he gathered, from the fact that Casta?os did not press him, but only assailed him with a distant and ineffective cannonade, that the main stress of the campaign was not at Andujar but elsewhere. The Spanish army was shifting[p. 186] itself eastward, and he therefore resolved that he must do the same, though he would have to abandon his cherished offensive position, his entrenchments, and such part of his supplies as he could not carry with him. Having made up his mind to depart, Dupont would have done wisely to start at once: if he had gone off early on the morning of the eighteenth, he would have found Reding and Coupigny not established in position at Baylen, but only just approaching from the south. Probably he might have brushed by their front, or even have given them a serious check, if he had fallen on them without hesitation.

But two considerations induced the French general to wait for the darkness, and to waste fourteen invaluable hours at Andujar. The first was that he hoped by moving at night to escape the notice of Casta?os, who might have attacked him if his retreat was open and undisguised. The second was that he wished to carry off his heavy baggage train: not only had he between 600 and 800 sick to load on his wagons, but there was an enormous mass of other impedimenta, mainly consisting of the plunder of Cordova. French and Spanish witnesses unite in stating that the interminable file of 500 vehicles which clogged Dupont’s march was to a very great extent laden with stolen goods[149]. And it was the officers rather than the men who were responsible for this mass of slow-moving transport.

It was not therefore till nine in the evening of the eighteenth that the French general thought fit to move. After barricading and blocking up the bridge of Andujar—he dared not use gunpowder to destroy it for fear of rousing Casta?os—he started on his night march. He had with him thirteen battalions of infantry and four and a half regiments of cavalry, with twenty-four guns, in all about 8,500 foot soldiers and 2,500 horse, allowing for the losses which he had sustained in sick and wounded during the earlier phases of the campaign[150]. His march was arranged as[p. 187] follows:—Chabert’s infantry brigade led the van: then came the great convoy: behind it were the four Swiss battalions under Colonel Schramm, which had lately been incorporated with the French army. These again were followed by Pannetier’s infantry brigade and Dupré’s two regiments of chasseurs à cheval. The rearguard followed at some distance: it was composed of two and a half regiments of heavy cavalry, placed under the command of General Privé, with the one veteran infantry battalion which the army possessed, the 500 Marines of the Guard, as also six compagnies d’élite picked from the ‘legions of reserve.’ From the fact that Dupont placed his best troops in this quarter, it is evident that he expected to be fighting a rearguard action, with Casta?os in pursuit, rather than to come into contact with Spanish troops drawn up across his line of march. He was ignorant that Reding and Coupigny had occupied Baylen on the previous day—a fact which speaks badly for his cavalry: with 2,500 horsemen about him, he ought to have known all that was going on in his neighbourhood. Probably the provisional regiments, which formed his whole mounted force, were incapable of good work in the way of scouting and reconnaissances.

The little town of Baylen is situated in a slight depression of a saddle-backed range of hills which runs southward out from the Sierra Morena. The road which leads through it passes over the lowest point in the watershed, as is but natural: to the north and south of the town the heights are better marked: they project somewhat on each flank, so that the place is situated in a sort of amphitheatre. The hill to the south of Baylen is called the Cerrajon: those to the north the Cerro del Zumacar Chico, and the Cerro del Zumacar Grande. All three are bare and bald, without a shrub or tree: none of them are steep, their lower slopes are quite suitable for cavalry work, and even their rounded summits are not inaccessible to a horseman. The ground to the west of them, over which the French had to advance, is open and level for a mile and a half: then it grows more irregular, and is thickly covered with olive groves and other vegetation, so that a force advancing over it is hidden from the view of a spectator on the hills above Baylen till it comes out into the open. The wooded ground is about two and a half miles broad: its western limit is the ravine of a mountain torrent, the Rumblar (or Herrumblar, as the aspirate-loving Andalusians sometimes call[p. 188] it). The road from Andujar to Baylen crosses this stream by a bridge, the only place where artillery can pass the rocky but not very deep depression.

It is necessary to say a few words about the ground eastward from Baylen, as this too was not unimportant in the later phases of the battle. Here the road passes through a broad defile rather than a plain. It is entirely commanded by the heights on its northern side, where lies the highest ground of the neighbourhood, the Cerro de San Cristobal, crowned by a ruined hermitage. The difference between the approach to Baylen from the west and from the east, is that on the former side the traveller reaches the town through a semicircular amphitheatre of upland, while by the latter he comes up a V-shaped valley cut through the hills.

Reding and Coupigny were somewhat surprised by the bicker of musketry which told them that the French had fallen upon their outposts. But fortunately for them their troops were already getting under arms, and were bivouacking over the lower slopes of the hills in a position which made it possible to extemporize without much difficulty a line of battle, covering the main road and the approaches to Baylen. They hastily occupied the low amphitheatre of hills north and south of the town. Reding deployed to the right of the road, on the heights of the Cerro del Zumacar Chico, Coupigny to its left on the Cerrajon. Their force was of a very composite sort—seventeen battalions of regulars, six of embodied militia, five of new Andalusian levies. The units varied hopelessly in size, some having as few as 350 men, others as many as 1,000. They could also dispose of 1,200 cavalry and sixteen guns. The greater part of the latter were placed in battery on the central and lowest part of the position, north and south of the high road and not far in front of Baylen. The infantry formed a semicircular double line: in front were deployed battalions near the foot of the amphitheatre of hills; in rear, higher up the slope or concealed behind the crest, was a second line in columns of battalions. The cavalry were drawn up still further to the rear. Finally, as a necessary precaution against the possible arrival of Vedel on the scene from La Carolina, Reding placed seven battalions far away to the east, on the other side of Baylen, with cavalry pickets out in front to give timely notice of any signs of the enemy in this quarter. These 3,500 men were quite out of the battle as long as Dupont was the only enemy in sight.

[p. 189]

Before it was fully daylight General Chabert and his brigade had thrust back the Spanish outposts. But the strength of the insurgent army was quite unknown to him: the morning dusk still lay in the folds of the hills, and he thought that he might possibly have in front of him nothing but some flying column of insignificant strength. Accordingly, after allowing the whole of his brigade to come up, Chabert formed a small line of attack, brought up his battery along the high road to the middle of the amphitheatre, between the horns of the Spanish position, and made a vigorous push forward. He operated almost entirely to the south of the road, where, opposite Coupigny’s division, the hill was lower and the slope gentler than further north.

To dislodge 14,000 men and twenty guns in position with 3,000 men and six guns was of course a military impossibility. But Chabert had the excuse that he did not, and could not, know what he was doing. His attempt was of course doomed to failure: his battery was blown to pieces by the Spanish guns, acting from a concentric position, the moment that it opened. His four battalions, after pushing back Coupigny’s skirmishing line for a few hundred yards, were presently checked by the reserves which the Spaniard sent forward. Having come to a stand they soon had to retire, and with heavy loss. The brigade drew back to the cover of the olive groves behind it, leaving two dismounted guns out in the open.

Behind Chabert the enormous convoy was blocking the way as far back as the bridge of the Rumblar. Five hundred wagons with their two or four oxen apiece, took up, when strung along the road, more than two and a half miles. Dupont, who rode up at the sound of the cannon, and now clearly saw the Spanish line drawn up on a front of two miles north and south of the road, realized that this was no skirmish but a pitched battle. His action was governed by the fact that he every moment expected to hear the guns of Casta?os thundering behind him, and to find that he was attacked in rear as well as in front. He accordingly resolved to deliver a second assault as quickly as possible, before this evil chance might come upon him. With some difficulty the Swiss battalions, Dupré’s brigade of light cavalry, and Privé’s dragoons pushed their way past the convoy and got into the open. They were terribly tired, having marched all night and covered fifteen miles of bad road, but their general threw them at once into the[p. 190] fight: Pannetier’s brigade and the Marines of the Guard were still far to the rear, at or near the bridge of the Rumblar.

Dupont’s second attack was a fearful mistake: he should at all costs have concentrated his whole army for one desperate stroke, for there was no more chance that 6,000 men could break the Spanish line than there had been that Chabert’s 3,000 could do so. But without waiting for Pannetier to come up, he delivered his second attack. The four Swiss battalions advanced to the north of the road, Chabert’s rallied brigade to the south of it: to the right of the latter were Privé’s heavy cavalry, two and a half regiments strong, with whom Dupont intended to deliver his main blow. They charged with admirable vigour and precision, cut up two Spanish battalions which failed to form square in time, and cleared the summit of the Cerrajon. But when, disordered with their first success, they rode up against Coupigny’s reserves, they failed to break through. Their own infantry was too far to the rear to help them, and after a gallant struggle to hold their ground, the dragoons and cuirassiers fell back to their old position. When they were already checked, Chabert and Schramm pushed forward to try their fortune: beaten off by the central battery of the Spanish line and its infantry supports, they recoiled to the edge of the olive wood, and there reformed.

The French were now growing disheartened, and Dupont saw disaster impending over him so closely that he seems to have lost his head, and to have retained no other idea save that of hurling every man that he could bring up in fruitless attacks on the Spanish centre. He hurried up from the rear Pannetier’s brigade of infantry, leaving at the bridge of the Rumblar only the single battalion of the Marines of the Guard. At eight o’clock the reinforcements had come up, and the attack was renewed. This time the main stress was at the northern end of the line, where Pannetier was thrown forward, with orders to drive Reding’s right wing off the Cerro del Zumacar Grande, while the other battalions renewed their assault against the Spanish centre and left. But the exhausted troops on the right of the line, who had been fighting since daybreak, made little impression on Coupigny’s front, and Reding’s last reserves were brought forward to check and hold off the one fresh brigade of which Dupont could dispose.

The fourth attack had failed. The French general had now but one intact battalion, that of the Marines of the Guard, which[p. 191] had been left with the baggage at the bridge over the Rumblar, to protect the rear against the possible advent of Casta?os. As there were still no signs of an attack from that side, Dupont brought up this corps, ranged it across the road in the centre of the line, and drew up behind it all that could be rallied of Chabert’s and Pannetier’s men. The whole formed a sort of wedge, with which he hoped to break through the Spanish centre by one last effort. The cavalry advanced on the flanks, Privé’s brigade to the south, Dupré’s to the north of the road. Dupont himself, with all his staff around him, placed himself at the head of the marines, and rode in front of the line, waving his sword and calling to the men that this time they must cut their way through [12.30 P.M.].

All was in vain: the attack was pressed home, the marines pushed up to the very muzzles of the Spanish cannon placed across the high road, and Dupré’s chasseurs drove in two battalions in Reding’s right centre. But the column could get no further forward: the marines were almost exterminated: Dupré was shot dead: Dupont received a painful (but not dangerous) wound in the hip, and rode to the rear. Then the whole attack collapsed, and the French rolled back in utter disorder to the olive groves which sheltered their rear. The majority of the rank and file of the two Swiss regiments in the centre threw up the butts of their muskets in the air and surrendered—or rather deserted—to the enemy[151].

At this moment, just as the firing died down at the front, a lively fusillade was heard from another quarter. Cruz-Murgeon’s light column, from the side of the mountains, had come down upon the Rumblar bridge, and had begun to attack the small baggage-guard[152] which remained with the convoy. All was up. Cruz-Murgeon was the forerunner of La Pe?a, and Dupont had not a man left to send to protect his rear. The battalions were all broken up, the wearied infantry had cast themselves down in the shade of the olive groves, and could not be induced even to rise to their feet. Most of them were gasping for water, which could not be got, for[p. 192] the stream-beds which cross the field were all dried up, and only at the Rumblar could a drink be obtained. Not 2,000 men out of the original 11,000 who had started from Andujar could be got together to oppose a feeble front to Reding and Coupigny. It was only by keeping up a slow artillery fire, from the few pieces that had not been silenced or dismounted, that any show of resistance could be made. When the attack from the rear, which was obviously impending, should be delivered, the whole force must clearly be destroyed.

Wishing at least to get some sort of terms for the men whom he had led into such a desperate position, Dupont at two o’clock sent his aide-de-camp, Captain Villoutreys, one of the Emperor’s equerries, to ask for a suspension of hostilities from Reding. He offered to evacuate Andalusia, not only with his own troops but with those of Vedel and Dufour, in return for a free passage to Madrid. This was asking too much, and if the Spanish general had been aware of the desperate state of his adversary, he would not have listened to the proposal for a minute. But he did not know that La Pe?a was now close in Dupont’s rear, while he was fully aware that Vedel, returning too late from the passes, was now drawing near to the field from the north. His men were almost as exhausted as those of Dupont, many had died from sunstroke in the ranks, and he did not refuse to negotiate. He merely replied that he had no power to treat, and that all communications should be made to his chief, who must be somewhere in the direction of Andujar. He would grant a suspension of arms for a few hours, while a French and a Spanish officer should ride off together to seek for Casta?os.

Dupont accepted these terms gladly, all the more so because La Pe?a’s division had at last reached the Rumblar bridge, and had announced its approach by four cannon-shots, fired at regular intervals, as a signal to catch Reding’s ear. It was with the greatest difficulty that the commander of the fourth Andalusian division could be got to recognize the armistice granted by his colleague; he saw the French at his mercy, and wanted to fall upon them while they were still in disorder. But after some argument he consented to halt. Captain Villoutreys, accompanied by the Spanish Colonel Copons, rode through his lines to look for Casta?os.

The Spanish commander-in-chief had displayed most blameworthy torpidity on this day. He had let Dupont slip away from[p. 193] Andujar, and did not discover that he was gone till dawn had arrived. Then, instead of pursuing at full speed with all his forces, he had sent on La Pe?a’s division, while he lingered behind with that of Felix Jones, surveying the enemy’s empty lines. The fourth division must have marched late and moved slowly, as it only reached the Rumblar bridge—twelve miles from Andujar—at about 2 p.m. It could easily have been there by 8 or 9 a.m., and might have fallen upon Dupont while he was delivering one of his earlier attacks on the Baylen position.

At much the same moment that Villoutreys and Copons reached Casta?os at Andujar, at about five o’clock in the afternoon, the second half of the French army at last appeared upon the scene. General Vedel had discovered on the eighteenth that he had nothing to fear from the side of the passes. He therefore called down all Dufour’s troops, save two battalions left at Santa Elena, united the two divisions at La Carolina, and gave orders for their return to Baylen on the following morning. Leaving the bivouac at five o’clock Vedel, with some 9,000 or 9,500 men, marched down the defile for ten miles as far as the village of Guarroman, which he reached about 9.30 or 10 a.m.[153] The day was hot, the men were tired, and though the noise of a distant cannonade could be distinctly heard in the direction of Baylen, the general told his officers to allow their battalions two hours to cook, and to rest themselves. By some inexplicable carelessness the two hours swelled to four, and it was not till 2 p.m. that the column started out again, to drop down to Baylen. An hour before the French marched, the cannonade, which had been growling in the distance all through the mid-day rest, suddenly died down. Vedel was in nowise disturbed, and is said to have remarked that his chief had probably made an end of the Spanish corps which had been blocking the road between them.

After this astonishing display of sloth and slackness, Vedel proceeded along the road for ten miles, till he came in sight of the rear of the Spanish position at Baylen. His cavalry soon brought him the news that the troops visible upon the hillsides were enemies: they consisted of the brigade which Reding had told off[p. 194] at the beginning of the day to hold the height of San Cristobal and the Cerro del Ahorcado against a possible attack from the rear. It was at last clear to Vedel that things had not gone well at Baylen, and that it was his duty to press in upon the Spaniards, and endeavour to cut his way through to his chief. He had begun to deploy his troops across the defile, with the object of attacking both the flanking hills, when two officers with a white flag rode out towards him. They announced to him that Dupont had been beaten, and had asked for a suspension of hostilities, which had been granted. La Pe?a’s troops had stayed their advance, and he was asked to do the same.

Either because he doubted the truth of these statements, or because he thought that his appearance would improve Dupont’s position, Vedel refused to halt, and sent back the Spanish officers to tell Reding that he should attack him. This he did with small delay, falling on the brigade opposed to him with great fury. Boussard’s dragoons charged the troops on the lower slopes of the Cerro del Ahorcado, and rode into two battalions who were so much relying on the armistice that they were surprised with their arms still piled, cooking their evening meal. A thousand men were taken prisoners almost without firing a shot[154]. Cassagnes’ infantry attacked the steep height of San Cristobal with less good fortune: his first assault was beaten off, and Vedel was preparing to succour him, when a second white flag came out of Baylen. It was carried by a Spanish officer, who brought with him De Barbarin, one of Dupont’s aides-de-camp. The general had sent a written communication ordering Vedel to cease firing and remain quiet, as an armistice had been concluded, and it was hoped that Casta?os would consent to a convention. The moment that his answer was received it should be passed on; meanwhile the attack must be stopped and the troops withdrawn.

Vedel obeyed: clearly he could do nothing else, for Dupont was his hierarchical superior, and, as far as he could see, was still a free agent. Moreover, De Barbarin told him of the very easy terms which the commander-in-chief hoped to get from Casta?os. If they could be secured it would be unnecessary, as well as risky, to continue the attack. For La Pe?a might very possibly have anni[p. 195]hilated the beaten division before Vedel could force his way to its aid, since horse and foot were both ‘fought out,’ and there was neither strength nor spirit for resistance left among them. Vedel therefore was justified in his obedience to his superior, and in his withdrawal to a point two miles up the La Carolina road.

Meanwhile Villoutreys, the emissary of Dupont, had reached the camp of Casta?os at Andujar[155] late in the afternoon, and laid his chief’s proposals before the Spaniard. As might have been expected, they were declined—Dupont was in the trap, and it would have been absurd to let him off so easily. No great objection was made to the retreat of Vedel, but Casta?os said that the corps caught between La Pe?a and Reding must lay down its arms. Early next morning (July 20) Villoutreys returned with this reply to the French camp.

Dupont meanwhile had spent a restless night. He had gone round the miserable bivouac of his men, to see if they would be in a condition to fight next morning, in the event of the negotiations failing. The result was most discouraging: the soldiers were in dire straits for want of water, they had little to eat, and were so worn out that they could not be roused even to gather in the wounded. The brigadiers and colonels reported that they could hold out no prospect of a rally on the morrow[156]. Only Privé, the commander of the heavy-cavalry brigade, spoke in favour of fighting: the others doubted whether even 2,000 men could be got together for a rush at the Spanish lines. When an aide-de-camp, whom Vedel had been allowed to send to his chief, asked whether it would not be possible to make a concerted attack on Reding next morning, with the object of disengaging the surrounded division, Dupont told him that it was no use to dream of any such thing. Vedel must prepare for a prompt retreat, in order to save himself; no more could be done.

At dawn, nothing having been yet settled, La Pe?a wrote to Dupont threatening that if the 1,000 men who had been captured[p. 196] by Vedel on the previous day were not at once released, he should consider the armistice at an end, and order his division to advance. The request was reasonable, as they had been surprised and taken while relying on the suspension of arms. Dupont ordered his subordinate to send them back to Reding’s camp. Casta?os meanwhile was pressing for a reply to his demand for surrender: he had brought up Felix Jones’s division to join La Pe?a’s in the early morning, so that he had over 14,000 men massed on the right bank of the Rumblar and ready to attack[157]. Dupont was well aware of this, and had made up his mind to surrender when he realized the hopeless demoralization of his troops. Early in the morning he called a council of war; the officers present, after a short discussion, drew up and signed a document in which they declared that ‘the honour of the French arms had been sufficiently vindicated by the battle of the previous day: that in accepting the enemy’s terms the commander-in-chief was yielding to evident military necessity: that, surrounded by 40,000 enemies, he was justified in averting by an honourable treaty the destruction of his corps.’ Only the cavalry brigadier Privé, refused to put his name to the paper, on which appear the signatures of three generals of division, of the officers commanding the artillery and engineers, of two brigadiers, and of three commanders of regiments.

After this formality was ended Generals Chabert and Marescot rode out from the French camp and met Casta?os. They had orders to make the best terms they could: in a general way it was recognized that the compromised division could not escape surrender, and that Vedel and Dufour would probably have to evacuate Andalusia and stipulate for a free passage to Madrid. The Spaniards were not, as it seems, intending to ask for much more. But while they were haggling on such petty points as the forms of surrender, and the exemption of officers’ baggage from search, a new factor was introduced into the discussion. Some irregulars from the Sierra Morena came to Casta?os, bringing with them as a prisoner an aide-de-camp of Savary[158]. They had secured his dispatch, which was a peremptory order to Dupont to evacuate Andalusia with all his three divisions, and fall back towards Madrid. This put a new face on affairs, for Casta?os saw that[p. 197] if he conceded a free retreat to Vedel and Dufour, he would be enabling them to carry out exactly the movement which Savary intended. To do so would clearly be undesirable: he therefore interposed in the negotiations, and declared that the troops of these two generals should not be allowed to quit Andalusia by the road which had been hitherto proposed. They must be sent round by sea to some port of France not immediately contiguous with the Spanish frontier.

Chabert and Marescot, as was natural, declaimed vehemently against this projected change in the capitulation, and declared that it was inadmissible. But they were answered in even more violent terms by the turbulent Conde de Tilly, who attended as representative of the Junta of Seville. He taunted them with their atrocities at the sack of Cordova, and threatened that if the negotiations fell through no quarter should be given to the French army. At last Casta?os suggested a compromise: he offered to let Dupont’s troops, no less than those of Vedel, return to France by sea, if the claim that the latter should be allowed to retreat on Madrid were withdrawn. This was conceding much, and the French generals accepted the proposal.

Accordingly Casta?os and Tilly, representing the Spaniards, and Chabert and Marescot, on behalf of Dupont, signed preliminaries, by which it was agreed that the surrounded divisions should formally lay down their arms and become prisoners of war, while Vedel’s men should not be considered to have capitulated, nor make any act of surrender. Both bodies of men should leave Andalusia by sea, and be taken to Rochefort on Spanish vessels. ‘The Spanish army,’ so ran the curiously worded seventh article of the capitulation, ‘guarantees them against all hostile aggression during their passage.’ The other clauses contain nothing striking, save some rather liberal permissions to the French officers to take away their baggage—each general was to be allowed two wheeled vehicles, each field officer or staff officer one—without its being examined. This article caught the eye of Napoleon, and has been noted by many subsequent critics, who have maintained that Dupont and his colleagues, gorged with the plunder of Cordova, surrendered before they needed, in order to preserve their booty intact. That they yielded before it was inevitable we do not believe: but far more anxiety than was becoming seems to have been shown regarding the baggage. This anxiety finds[p. 198] easy explanation if the Spanish official statement, that more than ￡40,000 in hard cash, and a great quantity of jewellery and silver plate was afterwards found in the fourgons of the staff and the superior officers, be accepted as correct[159].

The fifteenth clause of the capitulation had contents of still more doubtful propriety: it was to the effect that as many pieces of church plate had been stolen at the sack of Cordova, Dupont undertook to make a search for them and restore them to the sanctuaries to which they belonged, if they could be found in existence. The confession was so scandalous, that we share Napoleon’s wonder that such a clause could ever have been passed by the two French negotiators; if they were aware that the charge of theft was true (as it no doubt was), shame should have prevented them from putting it on paper: if they thought it false, they were permitting a gratuitous insult to the French army to be inserted in the capitulation.

While the negotiations were going on, Dupont sent secret orders to Vedel to abscond during the night, and to retreat on Madrid as fast as he was able. Chabert and Marescot had of course no knowledge of this, or they would hardly have consented to include that general’s troops in the convention. In accordance with his superior’s orders, and with the obvious necessities of the case, Vedel made off on the night of July 20-21, leaving only a screen of pickets in front of his position, to conceal his departure from the Spaniards as long as was possible. On the return of his plenipotentiaries to his camp on the morning of the twenty-first, Dupont learnt, to his surprise and discontent, that they had included Vedel’s division in their bargain with Casta?os. But as that officer was now far away—he had reached La Carolina at daybreak and Santa Elena by noon—the commander-in-chief hoped that his troops were saved.

The anger of the Spaniards at discovering the evasion of the second French division may easily be imagined. Reding, who was the first to become aware of it, sent down an officer into Dupont’s camp, with the message that if Vedel did not instantly return, he should regard the convention as broken, and fall upon the surrounded troops: he should give no quarter, as he considered that[p. 199] treachery had been shown, and that the armistice had been abused. Dupont could not hope to make a stand, and was at the enemy’s mercy. He directed his chief of the staff to write an order bidding Vedel to halt, and sent it to him by one of his aides-de-camp, accompanied by a Spanish officer. This did not satisfy Reding, who insisted that Dupont should write an autograph letter of his own in stronger terms. His demand could not be refused, and the two dispatches reached Vedel almost at the same hour, as he was resting his troops at Santa Elena before plunging into the passes.

Vedel, as all his previous conduct had shown, was weak and wanting in initiative. Some of his officers tried to persuade him to push on, and to leave Dupont to make the best terms for himself that he could. Much was to be said in favour of this resolve: he might have argued that since he had never been without the power of retreating, it was wrong of his superior to include him in the capitulation. His duty to the Emperor would be to save his men, whatever might be the consequences to Dupont. The latter, surrounded as he was, could hardly be considered a free agent, and his orders might be disregarded. But such views were far from Vedel’s mind: he automatically obeyed his chief’s dispatch and halted. Next day he marched his troops back to Baylen, in consequence of a third communication from Dupont.

On July 23 Dupont’s troops laid down their arms with full formalities, defiling to the sound of military music before the divisions of La Pe?a and Jones, who were drawn up by the Rumblar bridge. On the twenty-fourth Vedel’s and Dufour’s troops, without any such humiliating ceremony, stacked their muskets and cannon on the hillsides east of Baylen and marched for the coast. When the two corps were numbered it was found that 8,242 unwounded men had surrendered with Dupont: nearly 2,000 more, dead or wounded, were left on the battle-field; seven or eight hundred of the Swiss battalions had deserted and disappeared. With Vedel 9,393 men laid down their arms[160]. Not only did he deliver up his[p. 200] own column, but he called down the battalion guarding the Despe?a Perros pass. Even the troops left beyond the defiles in La Mancha were summoned to surrender by the Spaniards, and some of them did so, though they were not really included in the capitulation, which was by its wording confined to French troops in Andalusia. But the commanders of three battalions allowed themselves to be intimidated by Colonel Cruz-Murgeon, who went to seek them at the head of a few cavalry, and tamely laid down their arms[161].

The Spaniards had won their success at very small cost. Reding’s division returned a casualty list of 117 dead and 403 wounded, in which were included the losses of the skirmish of July 16 as well as those of the battle of the nineteenth. Coupigny lost 100 dead and 894 wounded. La Pe?a’s and Cruz-Murgeon’s columns, which had barely got into touch with the French when the armistice was granted, cannot have lost more than a score or two of men. The total is no more than 954. There were in addition 998 prisoners captured by Vedel when he attacked from the rear, but these were, of course, restored on the twentieth, in consequence of the orders sent by Dupont, along with two guns and two regimental standards.

Casta?os, a man of untarnished honour, had every intention of carrying out the capitulation. The French troops, divided into small columns, were sent down to the coast, or to the small towns of the Lower Guadalquivir under Spanish escorts, which had some difficulty in preserving them from the fury of the peasantry. It was necessary to avoid the large towns like Cordova and Seville, where the passage of the unarmed prisoners would certainly have led to riots and massacres. At Ecija the mob actually succeeded in murdering sixty unfortunate Frenchmen. But when the troops had been conducted to their temporary destinations, it was found that difficulties had arisen. The amount of Spanish shipping available would not have carried 20,000 men. This was a comparatively small hindrance, as the troops could have been sent off in detachments. But it was[p. 201] more serious that Lord Collingwood, the commander of the British squadron off Cadiz, refused his permission for the embarkation of the French. He observed that Casta?os had promised to send Dupont’s army home by water, without considering whether he had the power to do so. The British fleet commanded the sea, and was blockading Rochefort, the port which the capitulation assigned for the landing of the captive army. No representative of Great Britain had signed the convention[162], and she was not bound by it. He must find out, by consulting his government, whether the transference of the troops of Dupont to France was to be allowed.

On hearing of the difficulties raised by Collingwood, Casta?os got into communication with Dupont, and drew up six supplementary articles to the convention, in which it was stipulated that if the British Government objected to Rochefort as the port at which the French troops were to be landed, some other place should be selected. If all passage by sea was denied, a way by land should be granted by the Spaniards. This agreement was signed at Seville on August 6, but meanwhile the Junta was being incited to break the convention. Several of its more reckless and fanatical members openly broached the idea that no faith need be kept with those who had invaded Spain under such treacherous pretences. The newspapers were full of tales of French outrages, and protests against the liberation of the spoilers of Cordova and Jaen.

Matters came to a head when Dupont wrote to Morla, the Captain-General of Andalusia, to protest against further delays, and to require that the first division of his army should be allowed to sail at once [August 8]. He received in reply a most shameless and cynical letter[163]. The Captain-General began by declaring that there were no ships available. But he then went on to state that no more had been promised than that the Junta would request the British to allow the French troops to sail. He supposed that it was probable that a blank refusal would be sent to this demand. Why should Britain allow the passage by sea of troops who were destined to be used against her on some other point of the theatre of[p. 202] war? Morla next insinuated that Dupont himself must have been well aware that the capitulation could not be carried out. ‘Your Excellency’s object in inserting these conditions was merely to obtain terms which, impossible as they were to execute, might yet give a show of honour to the inevitable surrender.... What right have you to require the performance of these impossible conditions on behalf of an army which entered Spain under a pretence of alliance, and then imprisoned our King and princes, sacked his palaces, slew and robbed his subjects, wasted his provinces, and tore away his crown?’

After a delay of some weeks Lord Collingwood sent in to the Junta the reply of his government. It was far from being of the kind that Morla and his friends had hoped. Canning had answered that no stipulations made at Baylen could bind Great Britain, but that to oblige her allies, and to avoid compromising their honour, she consented to allow the French army to be sent back to France, and to be landed in successive detachments of 4,000 men at some port between Brest and Rochefort (i.e. at Nantes or L’Orient). It is painful to have to add that neither the Junta of Seville nor the Supreme Central Junta, which superseded that body, took any steps to carry out this project. Dupont himself, his generals, and his staff, were sent home to France, but their unfortunate troops were kept for a time in cantonments in Andalusia, then sent on board pontoons in the Bay of Cadiz, where they were subjected to all manner of ill usage and half-starved, and finally dispatched to the desolate rock of Cabrera, in the Balearic Islands, where more than half of them perished of cold, disease, and insufficient nourishment[164]. Vedel’s men were imprisoned no less than Dupont’s, and the survivors were only released at the conclusion of the general peace of 1814.

So ended the strange and ill-fought campaign of Baylen. It is clear that Dupont’s misfortunes were of his own creation. He ought never to have lingered at Andujar till July was far spent, but should either have massed his three divisions and fallen upon Casta?os, or have retired to a safe defensive position at Baylen or La Carolina and have waited to be attacked. He might have united something over 20,000 men, and could have defied every[p. 203] effort of the 35,000 Spaniards to drive him back over the Sierra Morena. By dividing his army into fractions and persisting in holding Andujar, he brought ruin upon himself. But the precise form in which the ruin came about was due less to Dupont than to Vedel. That officer’s blind and irrational march on La Carolina and abandonment of Baylen on July 17-18 gave the Spaniards the chance of interposing between the two halves of the French army. If Vedel had made a proper reconnaissance on the seventeenth, he would have found that Reding had not marched for the passes, but was still lingering at Mengibar. Instead, however, of sweeping the country-side for traces of the enemy, he credited a wild rumour, and hurried off to La Carolina, leaving the fatal gap behind him. All that followed was his fault: not only did he compromise the campaign by his march back to the passes, but when he had discovered his mistake he returned with a slowness that was inexcusable. If he had used ordinary diligence he might yet have saved Dupont on the nineteenth: it was his halt at Guarroman, while the cannon of Baylen were thundering in his ears, that gave the last finishing touch to the disaster. If he had come upon the battle-field at ten in the morning, instead of at five in the afternoon, he could have aided his chief to cut his way through, and even have inflicted a heavy blow on Reding and Coupigny. A careful study of Vedel’s actions, from his first passage of the Sierra Morena to his surrender, shows that on every possible occasion he took the wrong course.

But even if we grant that Vedel made every possible mistake, it is nevertheless true that Dupont fought his battle most unskilfully. If he had marched on the morning instead of the night of July 18, he probably might have brushed past the front of Reding and Coupigny without suffering any greater disaster than the loss of his baggage. Even as things actually fell out, it is not certain that he need have been forced to surrender. He had 10,000 men, the two Spanish generals had 17,000, but had been forced to detach some 3,500 bayonets to guard against the possible reappearance of Vedel. If Dupont had refused to waste his men in partial and successive attacks, and had massed them for a vigorous assault on the left wing of the Spaniards, where Coupigny’s position on the slopes of the Cerrajon was neither very strong nor very well defined, he might yet have cut his way through, though probably his immense baggage-train would have been lost. It is fair,[p. 204] however, to remember that this chance was only granted him because Casta?os, in front of Andujar, was slow to discover his retreat and still slower to pursue him. If that officer had shown real energy, ten thousand men might have been pressing Dupont from the rear before eight o’clock in the morning.

As it was Dupont mismanaged all the details of his attack. He made four assaults with fractions of his army, and on a long front. The leading brigades were completely worn out and demoralized before the reserves were sent into action. The fifth assault, in which every man was at last brought forward, failed because the majority of the troops were already convinced that the day was lost, and were no longer capable of any great exertions. It is absurd to accuse Dupont of cowardice—he exposed his person freely and was wounded—and still more absurd to charge him (as did the Emperor) with treason. He did not surrender till he saw that there was no possible hope of salvation remaining. But there can be no doubt that he showed great incapacity to grasp the situation, lost his head, and threw away all his chances.

As to the Spaniards, it can truly be said that they were extremely fortunate, and that even their mistakes helped them. Casta?os framed his plan for surrounding Dupont on the hypothesis that the main French army was concentrated at Andujar. If this had indeed been the case, and Dupont had retained at that place some 15,000 or 17,000 men, the turning movement of Reding and Coupigny would have been hazardous in the extreme. But the French general was obliging enough to divide his force into two equal parts, and his subordinate led away one of the halves on a wild march back to the passes. Again Reding acted in the most strange and unskilful way on July 17; after defeating Liger-Belair and Dufour he ought to have seized Baylen. Instead, he remained torpid in his camp for a day and a half: this mistake led to the far more inexcusable error of Vedel, who failed to see his adversary, and marched off to La Carolina. But Vedel’s blindness does not excuse Reding’s sloth. On the actual day of battle, on the other hand, Reding behaved very well: he showed considerable tenacity, and his troops deserve great credit. It was no mean achievement for 13,000 or 14,000[165] Spaniards, their ranks full of raw recruits[p. 205] and interspersed with battalions levied only five weeks before, to withstand the attack of 10,000 French, even if the latter were badly handled by their general. The Andalusians had good reason to be proud of their victory, though they might have refrained from calling Dupont’s Legions of Reserve and provisional regiments the ‘invincible troops of Austerlitz and Friedland,’ as they were too prone to do. They had at least succeeded in beating in the open field and capturing a whole French army, a thing which no continental nation had accomplished since the wars of the Revolution began.

NOTE

Sir Charles Vaughan, always in search of first-hand information, called on Casta?os and had a long conversation with him concerning the Convention. I find among his papers the following notes:—

‘Among other particulars of the surrender, General Casta?os stated that the French General Marescot had the greatest influence in bringing it about. The great difficulty was to persuade them [Marescot and Chabert] to capitulate for Vedel’s army as well as Dupont’s. A letter had been intercepted ordering Vedel back to Madrid, and another ordering Dupont to retire. This letter had considerable effect with the French: but the offer of carrying away their baggage and the plunder of the country was no sooner made, than the two generals desired to be permitted to retire and deliberate alone. After a few minutes they accepted the proposal. But General Casta?os, to make the article of as little value as possible, got them to insert the clause that the French officers should be allowed to embark all their baggage, &c., according to the laws of Spain. He well knew that those laws forbid the exportation of gold and silver. The consequence was that the French lost all their more valuable plunder when embarking at Puerto Santa Maria.’