SOULT’S EXPEDITION INTO ESTREMADURA. JANUARY-MARCH 1811. THE BATTLE OF THE GEBORA AND THE FALL OF BADAJOZ

In his original scheme for the invasion of Portugal, Napoleon had given no part to the Army of Andalusia, judging that Masséna, supported by the 9th Corps, would be amply strong enough to drive the English into the sea. It is not till the 29th of September that the imperial correspondence begins to show signs of a desire that Soult should do something to help the Army of Portugal. But the assistance which was to be given is defined, in the dispatch of that date, as no more than a diversion to be made against Estremadura by Mortier and the 5th Corps, with the object of preventing La Romana from giving any aid to Wellington[35]. Soult is directed to see that Mortier keeps the Spanish Army of Estremadura in check: he is always to be on its heels, so that it will have no opportunity of sending troops towards the Tagus. Nothing is said about making a serious attack on Estremadura, or of threatening Badajoz with a siege. On the 26th of October comes the next allusion to this subject[36]: the Emperor had learnt from the English newspapers—always his best source of intelligence—that La Romana with a large part of his forces has marched on Lisbon to join Wellington, and that he has been able to do so without molestation. That this should have happened was, he thought, due to direct disobedience on Soult’s part: the Marshal cannot have kept in touch with the enemy. And he is directed in vague terms to ‘faire pousser sur La Romana,’ whatever exactly that may mean. An interpretation for the phrase,[p. 24] however, turns up in the next imperial dispatch[37]—Mortier and the 5th Corps ought to have followed the Spanish general march for march, and to have presented themselves on the Lower Tagus in face of Lisbon shortly after the arrival of La Romana in the Portuguese capital.

Soult had little difficulty in proving that this scheme was absolutely impossible. It argued, indeed, a complete misconception of the situation in Estremadura and Andalusia. To talk lightly of pushing Mortier and the 5th Corps, which comprised at this moment just 13,000 men, right across Estremadura to the mouth of the Tagus, ‘en talonnant La Romana,’ was futile. The Spanish General had gone off to join Wellington with some 7,000 or 8,000 men. But he had left behind him in Estremadura two strong infantry divisions, those of Mendizabal and Ballasteros, with 12,000 bayonets, 6,000 more infantry in garrison at Badajoz, Olivenza, and Albuquerque, and the whole of his cavalry, 2,500 sabres. In addition there were interposed between Mortier and the Tagus about 8,500 Portuguese—a cavalry brigade under Madden which had been lent to the Spaniards, and, near Badajoz, a regular infantry brigade at Elvas, and four militia regiments, forming the garrisons of the last-named place, of Campo Mayor, and of Jerumenha. That is to say, there lay before Mortier, after La Romana’s departure, a field army which, if concentrated, would make up 18,000 men, and in addition six fortresses containing garrisons amounting to 11,000 men more, and covering all the main strategical points of the country. How could he have pursued La Romana? If he had followed, he would have found himself at once involved in a campaign against superior forces in a region studded with hostile strongholds. ‘On this frontier,’ as Soult wrote to Berthier, ‘there are six fortified places—Badajoz, Olivenza, Jerumenha, Elvas, Campo Mayor, Albuquerque, in which there are at least 20,000 infantry and 2,500 cavalry. It is clear to me that if I thrust a body of 10,000 men forward to the Tagus, as his majesty has directed, that body would never reach its[p. 25] destination, and would be cut off and surrounded before I could get up to its aid[38].’ This was indisputably correct: Mortier might have beaten Mendizabal and Ballasteros in the open field, if they chose to offer him battle; but if they preferred to concentrate on Badajoz or Elvas, and defied him from under the shadow of those great fortresses, he could not ignore them and march by in pursuit of La Romana. The moment that he was past their positions, they would cut him off from Andalusia, and he would find himself with their whole force at his back, and in front of him anything that Wellington might have sent to the south bank of the Tagus. In December this would have meant 14,000 men under Hill, at the New Year about 16,000 under Beresford. As Soult truly said, the expedition would have been encompassed, and probably destroyed, long before Masséna heard of its having got anywhere near him[39]. Nevertheless the[p. 26] Commander-in-Chief of the Army of Portugal was told to expect this diversion as a matter of certainty: in the dispatch that Foy took back to him (dated Dec. 4) he was given the precise statement that the 5th Corps was to be looked for somewhere in the direction of Montalv?o and Villaflor, on the Tagus above Abrantes, at no distant date.

After having, as he thought, demonstrated to his master that it would be useless to send Mortier alone, with 10,000 or 12,000 men, to make an impossible dash at the Tagus, Soult made another proposal. He would undertake not a mere raid, but the capture of Badajoz, the conquest of Estremadura, and the destruction of the army of that province; but he must take with him a force much greater than the mere 5th Corps. ‘The enterprise is a big one, but ought to succeed—at least it will produce a happy diversion in favour of the imperial army in Portugal[40].’ It would call back La Romana from Lisbon, and possibly cause Wellington to detach troops in his aid, and Masséna would have less in front of him in consequence, and might resume the offensive.

There was of course another course possible: Soult might have marched for Estremadura not with the 13,000 men of the 5th Corps alone, nor yet with the 20,000 men whom he actually took thither in January, but with the greater part of the French Army of Andalusia, 35,000 or 40,000 men. To do so he would have had to abandon Granada, Malaga, and Jaen on the one side, and his hold on the Condado de Niebla and the west upon the other. He might even possibly have had to raise the siege of Cadiz, though this is not quite certain. Many months after, in the end of March, when all chance of the conquest of Portugal was over, the Emperor told him that this would have been his proper course, and read him an ex-post-facto lecture on the advantages that might have followed, if he had evacuated two-thirds of his viceroyalty and taken an imposing force to sweep across the Alemtejo and assail Lisbon from the southern side[41].[p. 27] But it must be remembered that in December and January all the orders that were sent him directed him to move no more than a small corps—in one dispatch the Emperor calls it only 10,000 men. A supreme commander-in-chief present on the spot might have seen his way to make the temporary sacrifice of the provinces which had cost so many men to conquer and to hold, in order that every available man might be sent against Lisbon, and the English might at last be expelled from the Peninsula. But Soult was not such a commander-in-chief; he was only one of the many viceroys whom Napoleon preferred to a single omnipotent lieutenant. Was it likely that he would sacrifice half his own territory, when no order to do so lay before him, in order that a colleague, sent on a separate task with forces no less than his own, might have every possible advantage? Soult is often blamed for not having seen that the crushing of Wellington’s army was the end to which all others should have been subordinated, and that it would have been cheap in the end to surrender half or three-quarters of Andalusia, and even to raise the siege of Cadiz, in order to secure that point. He might have replied that his master was no less blind than himself: dispatch after dispatch had ordered him to send a trifling detachment towards the Tagus, not to mass every available man and march on Lisbon, leaving Seville and the Cadiz lines exposed to all manner of dangers. He was primarily responsible for the retention of the Andalusia that he had conquered; it was for the Emperor, not for himself, to order the evacuation of much or all of that great realm. The Emperor gave no such directions: from September to January his whole series of dispatches spoke of nothing more than the movement of a moderate force; those of January 25th and February 6th approved of the course which Soult had actually chosen, and took it for granted that he would not move towards the Tagus till he should have captured Badajoz[42]. It was not till March, when he began to[p. 28] see that all his arrangements were going wrong, and that his scheme of times was erroneous, that the Emperor began suddenly to launch out into criticisms of Soult, and to complain that ‘to try to hold every point at a moment of crisis leads to possible disaster,’ that ‘Seville, Badajoz, and the Cadiz lines were the only necessary things,’ and that the Marshal ought to have 30,000 men or more with him at Badajoz instead of the 20,000 men whom he had actually taken thither[43]. If so, why had the orders not been given to that effect early in December, when Napoleon had just learnt from Foy the state and position of the Army of Portugal, which had so long been hidden from him behind Wellington’s screen of Ordenan?a?

If we seek deep enough, we find the cause of all misdirections in the fact that the Emperor persisted in guiding the movements of all his army from Paris, and would not appoint an independent commander-in-chief of all the Spanish armies, who should be able to issue orders that would be promptly obeyed by every separate marshal or general in each province. A moment’s reflection shows that the data as to the details of the situation in the Peninsula, from which Napoleon had to construct his scheme of operations, always came to him a month late. And when he had issued the dispatch which dealt with the situation, it reached its destination after the interval of another month, and had long ceased to have any bearing on the actual position of affairs. A single example of how the system worked may suffice. Masséna started Foy for Paris, with his great report on the state of the Army of Portugal, on October 29. Foy reached Paris and saw the Emperor on November 22 and the succeeding days. The detailed dispatches to Masséna and Soult, consequent on Foy’s report, were not sent off till December 4.[p. 29] On January 22nd Soult acknowledges the receipt of the dispatch of that date, along with that of two others dated November 28 and December 10, all of which arrived together, because the guerrilleros of La Mancha had stopped the posts between Madrid and Seville for a full fortnight after the New Year[44]. Of what value to Soult on January 22 could be orders based on the condition and projects of Masséna on October 29? The data at the base of the orders were three months old—while Soult had been already for more than a month engaged on a campaign undertaken on his own responsibility, without any knowledge of the exact requirements of Masséna, or of the intentions of the Emperor.

The Estremaduran expedition of January-March 1811, therefore, must be looked upon as the private scheme of the Duke of Dalmatia[45], undertaken with the general object of giving indirect assistance to Masséna, because the last orders that he had received from Paris (those of October 26), telling him to give direct assistance, by sending Mortier to the Tagus, were impossible of execution[46]. Soult had two leading ideas in his mind when he planned out his campaign. The first was that he was going into a country thickly set with fortresses; the second was that, when once the skirts of the Sierra Morena have been passed, Estremadura is a ‘cavalry country,’ a land of heaths and of unenclosed tillage-fields of vast area. Accordingly he intended to march with a very large force of cavalry, and with a heavy siege-train. At Seville he had at his disposition the greatest arsenal of Spain; but for many months all that it produced had been going forward to Cadiz: no less than 290 pieces had been sent to arm the vast lines in front of the blockaded city. Accordingly it took some time to get ready the heavy guns, and to manufacture the ammunition required for such a big business as the siege of the six fortresses,[p. 30] small and great, into whose midst he was about to thrust himself. The personnel for the siege-train had also to be collected: requisitions were sent, both to Victor at Cadiz and to Sebastiani at Granada, to detach and send into Seville nearly all their sappers, and the men of several companies of artillery. They were also to send to the expeditionary force many regiments of cavalry. Mortier had only two (10th Hussars and 21st Chasseurs), which had sufficed when he was engaged in the heights of the Sierra Morena, but were insufficient when he was about to descend into the plain of the Guadiana. Accordingly half the cavalry of Victor’s 1st Corps was called up—four regiments (4th, 14th, 26th Dragoons, 2nd Hussars), while Sebastiani gave up one (27th Chasseurs); to these was added an experimental Spanish cavalry regiment of ‘Juramentados’ recently organized at Seville. Only one infantry regiment was requisitioned, the 63rd Line, from Victor’s 3rd Division. The putting together of these resources gave a force in which the proportions of the arms were very peculiar—4,000 cavalry, 2,000 artillery and sappers, to only 13,500 infantry; the last, all save the above-mentioned 63rd regiment, drawn from Mortier’s 5th Corps. The orders for the concentration of the troops were issued early in December, but owing to the time required for drawing in units from Granada and Cadiz, and for the preparation of the siege-train, it was not till the last day of the old year that the Marshal took his departure from Seville.

The collection of a field army of 20,000 men, which was to cut itself loose from Andalusia for a time, imposed some tiresome problems on Soult. Since he had resolved not to evacuate Granada or Malaga on the one hand, nor the posts west of the Guadalquivir on the other, and since he was drawing off the 5th Corps, which had hitherto provided for the safety of Seville and found detachments for the Condado de Niebla, he had to make provision for the filling of the gap left behind him. Hence we find him calling upon Victor to spare men from in front of Cadiz—a demand which the Duke of Belluno took very ill—since he truly declared that he had no more troops in the 1st Corps than sufficed to man the lines and to keep posts of observation in his rear. The garrison of Cadiz was always increasing, and included a strong nucleus of British troops.[p. 31] How could he face sorties, or disembarkations in his rear, if he was cut down to a mere 18,000 men in place of the 24,000 on which he had hitherto reckoned? Nevertheless, he was forced to provide a detachment to hold Xeres, as a half-way house to Seville, and to send out a cavalry regiment (9th Dragoons) and one battalion west of the Guadalquivir. Similarly, the brigade of Godinot in the kingdom of Cordova[47] was required to find a skeleton garrison for Seville, which was raised to a somewhat higher figure, in appearance, by the doubtful aid of some ‘juramentado’ companies of Spaniards, and of the dép?ts and convalescents of the 5th Corps. The great city, with a turbulent population of 100,000 souls, which formed the centre of his viceroyalty, became at this time Soult’s weakest point—he left it so inadequately held that it was at the mercy of any considerable hostile force which might approach it—and such a force was ere long, as we shall see, to make its appearance. Godinot had also to look after the insurgent bands of the central Sierra Morena, who often blocked the post road to Madrid. Sebastiani (save for the cavalry and artillery borrowed from him) was left with his 4th Corps intact, and his duty was unchanged—to watch the Spanish army of Murcia, and to suppress the guerrilleros of the Sierra de Ronda and the eastern coast—an unending task from which Soult thought that he ought not to be distracted. Napoleon, wise after the event, wrote in March that Soult should have left no more than the Polish division of the 4th Corps in the direction of Granada, and have brought the remainder of it to strengthen or support the troops at Seville and in the lines before Cadiz. In that case the Poles would certainly have had to move westward also ere long, since there were but 5,000 of them, and all Eastern Andalusia would have had to be evacuated. But this idea had never struck Soult as practicable, and Sebastiani’s whole corps was left in its old posts in the kingdom of Granada.

The invasion of Estremadura was carried out in two columns of about equal strength, which used the two main passes between Western Andalusia and the valley of the Guadiana.[p. 32] The right column under Latour-Maubourg took the route by Guadalcanal, Llerena, and Usagre; it was composed of his own regiments of dragoons from the 1st Corps, and of Girard’s infantry division of the 5th Corps, which latter had been cantoned in Llerena since the autumn, and was now picked up and taken forward by the cavalry. The left column, which was accompanied both by Soult and by Mortier, was composed of Briche’s light cavalry and Gazan’s division of the 5th Corps. It had to escort the slowly moving siege-train of 34 guns, which (with the 60,000 kilos of powder belonging to it) was drawn by 2,500 draught oxen, requisitioned along with their drivers from the province of Seville. This column took the route Ronquillo, Sta Olalla, Monasterio, which, if less steep and better made than the Llerena-Guadalcanal road, is longer, and passes through an even more desolate and resourceless country. It was intended that the two columns should join at Los Santos or Almendralejo, in the Estremaduran plain, and lay siege at once to Badajoz, the enemy’s most formidable stronghold. Its fall, so Soult hoped, would lead to the easy conquest of the minor fortresses.

But the two columns did not meet with equal fortune. That commanded by Latour-Maubourg met practically no resistance in its first stages. On arriving at Usagre on January 3, it found in its front almost the whole of the allied cavalry in Estremadura—Butron with 1,500 Spaniards, Madden with nearly 1,000 Portuguese. But this was merely a screen thrown out to cover the retreat beyond the Guadiana of Mendizabal and the division of Spanish infantry which had been cantoned in this region. That officer had been ordered by his chief, La Romana, to break the bridge of Merida, after retiring over it, and then to attempt to hold the line of the Guadiana. He did neither; precipitately marching on Merida, he passed through it in great haste, but forgot to see that the bridge was duly destroyed, and then retired along the north bank of the Guadiana to Badajoz. Latour-Maubourg, according to his directions, did not cross the river, but halted near Almendralejo, to wait for the other column, which was not forthcoming. Only Soult himself and Briche’s light cavalry appeared at Zafra on the 5th, and joined Latour-Maubourg on the 6th of January.[p. 33] Gazan’s infantry and the siege-train were far away, and unavailable for many a day. The plans of the left invading column had miscarried. For when its head reached Monasterio, at the summit of the long pass, its tail, the siege-train, was dragging far behind. In the desolate stages about Ronquillo and Sta Olalla it had met with tempestuous rains, as might have been expected at the season. Many of the oxen died, the unwilling Spanish drivers deserted wholesale, and there was much delay and considerable loss of vehicles. The train and its small escort got completely separated from Gazan’s infantry. At this moment Soult’s cavalry reported to him that a formidable column of hostile infantry was lying a few miles to the west of Monasterio, on the bad cross-road to Calera, and was apparently moving round his flank, either to fall upon the belated convoy or perhaps to make a dash at Seville.

This column was the 5,000 infantry of Ballasteros, who, as it chanced, had begun to march southward at the same moment that Soult had started northward. The Spanish general had just received orders from Cadiz bidding him cut himself loose from the Estremaduran army, and move into the Condado de Niebla, where he was to unite with the local levies under Copons, drive out the weak French detachment there stationed, and threaten Seville from the west if it should be practicable. These orders had been given, of course, before Soult’s plan for invading Estremadura was suspected at Cadiz. But though unwise in themselves—it was not the time to deplete Estremadura of troops—they had the effect of bringing Soult’s great man?uvre to a standstill for some weeks. The Marshal determined that he must free his flank from this threatening force before continuing his march, and ordered Mortier to attack Ballasteros without delay. This was done, but the Spaniard, after a running fight of two hours, retired to Fregenal, fifteen miles further west, without suffering any serious harm (January 4th). He was still in a position to threaten the rear of the convoy, or to slip round the flank of the French column towards Seville. Soult therefore resolved to go on with his cavalry and join Latour-Maubourg, but to drop Gazan’s infantry in the passes, with the order to head off Ballasteros at all costs, and to cover the siege-train in its journey across the mountains.[p. 34] Gazan therefore took post at Fuentes de Leon, but soon heard that Ballasteros had moved south again from Fregenal towards the Chanza river, and was apparently trying to get round his flank. Leaving a detachment to help the convoy on its slow and toilsome route, Gazan resolved to pursue the Spanish column and destroy it at all costs. This determination led him into three weeks of desperate mountain-marching and semi-starvation, at the worst season of the year. For Ballasteros, who showed considerable skill in drawing his enemy on, moved ever south and west towards the lower Guadiana, and picked up Copons’s levies by the way. He at last turned to fight at Villanueva de los Castillejos on January 24th. Gazan, who had been joined meanwhile by the small French detachment in the Condado de Niebla, brought his enemy to action on the 25th. The Asturian battalions which formed Ballasteros’s division made a creditable resistance, and when evicted from their position retired across the Guadiana to Alcoutim in Portugal, without having suffered any overwhelming loss[48]. Gazan therefore resolved to pursue them no further—indeed he had been drawn down into one of the remotest corners of Spain to little profit, and realized that Soult must be brought to a standstill one hundred miles away, for want of the 6,000 infantry who had now been executing their toilsome excursion in the mountains for three weeks.

Accordingly, the French general bade Remond, the commander of the Niebla detachment, watch Ballasteros, and himself returned to Estremadura by a most painful march through Puebla de Guzman, El Cerro, Fregenal, and Xeres de los Caballeros. He reported his return to Soult at Valverde on February 3rd. His services had been lost to his chief for a month all but two days, a fact which had the gravest results on the general course of the campaign of Estremadura[49].

[p. 35]For the Duke of Dalmatia, when he had joined Latour-Maubourg on January 6th, found that he had at his disposition 4,000 cavalry but only the 6,000 infantry of Girard, while the siege-train was still blocked in the passes by Monasterio. With such a force he did not like to beleaguer a place so large and so heavily garrisoned as Badajoz. Accordingly, he was forced to abandon his original intention of forming its siege, and to think of some lesser enterprise, more suited to his strength. After some hesitation, he determined to attack the weak and old-fashioned fortress of Olivenza, the southernmost of all the fortified places on the Spanish-Portuguese frontier. To cover his movement he sent Briche’s cavalry to Merida, which they occupied on January 7th, almost without resistance, finding the bridge intact. From thence they sought for Mendizabal on the north side of the Guadiana, and discovered that he had withdrawn to Albuquerque, twenty miles north of Badajoz. Meanwhile Latour-Maubourg with four dragoon regiments took post at Albuera to watch the garrison of Badajoz, while Soult marched with Girard’s infantry and one cavalry regiment to attack Olivenza, before whose walls he appeared on January 11th, 1811.

Olivenza ought never to have been defended. For since its cession by Portugal to Spain after Godoy’s futile ‘War of the Oranges’ in 1801, it had been systematically neglected. The breach made by the Spaniards at its siege ten years before had never been properly repaired—only one-third of the masonry had been replaced, and the rest of the gap had been merely stopped with earth. Its one outlying work, a lunette 300 yards only from its southern point, was lying in ruins and unoccupied. The circuit of its walls was about a mile, but there were only eighteen guns[50] to guard them. The garrison down to the 5th of January had consisted of a single battalion left there by La Romana, when he marched for Portugal in October. But Mendizabal, apparently in inexcusable ignorance of the condition of the place, had ordered a whole brigade of his infantry to throw themselves into it when Soult began to press forward. He sacrificed, in fact, 2,400 out of the 6,000 bayonets of his division[p. 36] by bidding them shut themselves up in an utterly untenable fortress[51]. The governor, General Manuel Herck—an old Swiss officer—was ailing and quite incapable; a man of resources might have done something with the heavy garrison placed under his orders, even though the walls were weak and artillery almost non-existent; but Herck disgraced himself.

When Soult arrived in front of Olivenza on January 11th, his engineers informed him that the place, weak as it was, was too strong to take by escalade, but that a very few days of regular battering would suffice to ruin it. Unfortunately for him, there was as yet no heavy artillery at his disposition, but only the divisional batteries of Girard’s two brigades; the siege-train was still stuck in the passes. However, the outlying lunette opposite the south front was at once seized, and turned into a battery for four field-guns, which opened their fire on the next day. The old Spanish breach of 1801, obviously ready to fall in on account of its rickety repairs, was visible in the north-west front, the bastion of San Pedro. Opposite this sites for two more batteries were planned, and a first parallel opened. The trench-work went on almost unhindered by the Spaniards, who showed but few guns and shot very badly, but under considerable difficulties from the rainy weather, which was perpetually flooding the lower parts of the lines. But in ten days approaches were pushed right up to the edge of the counterscarp, and mines prepared to blow it in. The siege artillery began to arrive on the 19th, in detachments, and continued to drop in for several days. On the 21st the batteries, being completed, were armed with the first 12-pounders that came up. On the 22nd the fire began, and at once proved most effective: the bastion of San Pedro began to crumble in, and the old breach of 1801 revealed itself, by the falling away of the rammed earth which alone stopped it up. The arrangements for a storm had not yet been commenced when the garrison hoisted the white[p. 37] flag. Mortier refused all negotiations and demanded a surrender at discretion. This the old governor hastened to concede, coming out in person at one of the gates, and putting the place at the disposition of the French without further argument.[52] Soult and Mortier entered next day, and 4,161 Spanish troops marched out and laid down their arms before the 6,000 infantry of Girard, who had formed the sole besieging force. The total loss of the French during the siege was 15 killed and 40 wounded—that of the besieged about 200. The figures are a sufficient evidence of the disgraceful weakness of the defence.

When one reflects what was done to hold the unfortified town of Saragossa, and the mediaeval enceinte of Gerona, it is impossible not to reflect what a determined governor might have accomplished at Olivenza. The place was short of guns, no doubt—but the enemy was worse off till the last days of the siege, since he had nothing but twelve light field-pieces until the siege-train began to arrive. General Herck made no sorties to disturb the works—though he had a superabundant garrison; he made no serious attempt to retrench the breach, and he surrendered actually ere the first summons had been sent in before the storm. At the worst he might have tried to cut his way out between the French camps, which were scattered far from each other, owing to the extremely small numbers of the besieging army, who only counted three men to the defenders’ two. Altogether it was a disgraceful business. The place, no doubt, ought never to have been held; but if held it might at least have been defended—which it practically was not.

Soult was placed in a new difficulty by the surrender of Olivenza. Though his siege-train had begun to come up, he had no news of Gazan, and his infantry was still no more than a single division. He had to spare two battalions to escort the 4,000 prisoners to Seville, and to put another in Olivenza as garrison[53]. This left him only eleven battalions—5,435 bayonets,[p. 38] to continue the campaign, though he had the enormous force of 4,000 cavalry at his disposition, and a siege-train that was growing every day, as more belated pieces came up from the rear. He might probably have waited for Gazan, for whom messages had been vainly sent, if he had not received, on the day that Olivenza fell, one more of Berthier’s peremptory letters, dated 22 December, in which he was told (as usual) to send the 5th Corps to join Masséna on the Tagus without delay. This letter came at an even more inappropriate moment than usual, as Gazan, with half that corps, was lost to sight in the mountain of the Condado de Niebla, more than a hundred and twenty miles away. But it was clear that something immediate must be done, or the Emperor would be more discontented than before; accordingly Soult resolved to take the very hazardous step of laying siege to Badajoz at once with the small infantry force at his disposition. For this move would undoubtedly provoke alarm at Lisbon, and lead Wellington to send off La Romana’s army to succour it, and perhaps some Anglo-Portuguese troops also, so that the mass opposed to Masséna would be more or less weakened.

Accordingly on the 26th of January Soult marched against Badajoz, which is only twelve miles north-west of Olivenza, with under 6,000 infantry, ten companies of artillery, and seven of sappers, to invest the southern side of Badajoz, while Latour-Maubourg, with six regiments of cavalry, crossed the Guadiana by a ford, and went to blockade the place on its northern front.

Badajoz, though owning some defects, was still a stronghold of the first class, in far better order than most of the Peninsular fortresses. It belonged to that order of places whose topography forces a besieger to divide his army by a dangerous obstacle, since it lies on a broad river, with the town on one side and a formidable outwork on the other. Indeed the most striking feature of Badajoz, whether the traveller approaches it from the east or the west, is the towering height of San Cristobal, crowned by its fort, lying above the transpontine suburb and dominating the whole city. Any enemy who begins operations against the place must take measures to blockade or to attack this high-lying fort, which completely covers the bridge and its tête-du-pont, and effectively protects ingress or egress to or from[p. 39] the place. But San Cristobal is not easy to blockade, since it is the end-bluff of a very steep narrow range of hills, which run for many miles to the north, and divide the country-side beyond the Guadiana into two separate valleys, those of the Gebora and the Caya, which are completely invisible from each other.

The city of Badajoz is built on an inclined plane, sloping down from the Castle, which stands on a lofty hill with almost precipitous grass slopes, at the north-east end of the place, down to the river on the north and the plain on the south and west. The castle-hill and San Cristobal between them form a sort of gorge, through which the Guadiana, narrowed for a space, forces its way, to broaden out again at the immensely long bridge with its thirty-two arches and 640 yards of roadway. Below the castle the Rivillas, a stagnant brook with hardly any current,—the home of frogs and the hunting-ground of the city storks,—coasts around the walls, and finally dribbles into the river. The front of the place from the river to the castle was composed of eight regular bastions; along the river edge there lies nothing more than a single solid wall without relief or indentations: but this side of the place is wholly inaccessible owing to the water. There are two outlying works, which cover heights so close into the place that it is necessary to hold them, lest the enemy should establish himself too near the enceinte. These are the Picurina lunette beyond the Rivillas, and the much larger Pardaleras fort, a ‘half-crown-work,’ opposite the south point of the city, which covers a well-marked hill that commands that low-lying part of the place, and is a position impossible to concede to the besieger, since it is only 250 yards from the nearest bastion. It was ill-designed, having a very shallow ditch, and being incompletely closed at its gorge by a mere palisade.

The eight bastions which form the attackable part of the enceinte of Badajoz have (they remain to-day just as they were in 1811, for the place has never been modernized) a height of about thirty feet[54] from the bottom of the ditch to the rampart, while the curtains between them are somewhat lower, about twenty-two feet only. The ditch was broad, with a good[p. 40] counterscarp in masonry seven feet high; beyond it each bastion was protected in front by a rather low and weak demi-lune.

The garrison, not more than enough for such an extensive place, consisted at the New Year of 4,100 men; but Mendizabal threw in two battalions more (1st and 2nd of the Second Regiment of Seville) before he retired to the borders of Portugal, so that the figure had risen to 5,000 before Soult appeared in front of the walls. The governor was a very distinguished soldier, General Rafael Menacho, who had served through the old French war of 1792-5, and had commanded a regiment at Baylen. He was in the full vigour of middle age (forty-four years old) and abounding in spirit, resolution, and initiative, as all his movements showed down to the unhappy day of his death.

Soult’s engineers, after surveying the situation of Badajoz, reported that under ordinary circumstances the most profitable front to attack would certainly be the western—that between the Pardaleras fort and the river; but at the same time they decided that it had better be left alone. For the army was so weak that it could not properly invest the whole city, and if the north bank of the Guadiana were left practically unoccupied, as must necessarily be the case, the Spaniards would be able to seize the ground beyond the bridge-head, and establish batteries there, which would effectually enfilade the trenches which would have to be constructed for approaching the west side of the place. The castle and the north-east angle of the town were too high-lying to be chosen as the point of attack, and the Rivillas and its boggy banks were better avoided. They therefore advised that the south front should be chosen as the objective, and that the first operation taken in hand should be the capture of the Pardaleras fort, for that work appeared weak and ill-planned, while its site would make the most advantageous of starting-points for breaching the enceinte of the town itself. It was the most commanding ground close in to the walls which could be discovered. Soult and Mortier concurred, and placed the army in the best position for utilizing this method of attack. The camps of Girard’s division were placed on and around two low hills, the Cerro de San Miguel on the right of the Rivillas, and the Cerro del Viento on its left. On[p. 41] the former height, about 1,800 yards from the town, nothing was done save the construction of a rough entrenchment—to face the Picurina and restrict possible sallies—in which three small batteries were afterwards inserted. The Cerro del Viento, which is about 1,200 yards from the Pardaleras, was to be the real starting-point of the attack, and under its side the siege-park and engineers’ camp were established. Two batteries in front of it were marked out and begun on the first night of ‘open trenches’ (January 28-9), but it was not till the third night (January 30-1) that the first parallel was commenced, on the undulating ground to the west of the Rivillas. When the work became visible next day, the governor directed a vigorous sortie against it, composed of 800 men. The trenches were occupied for a moment, but soon recovered by the French supports. A small body of Spanish cavalry which had taken part in the sally rode right round the rear of the camp, and sabred the chef-de-bataillon Cazin, the chief engineer, and a dozen of his sappers on the Cerro del Viento. But the total loss of the besiegers was only about seventy killed and wounded, while the men of the sortie suffered much more heavily, while they were being driven back across the open ground towards the city. Their commander, a Colonel Bassecourt of the 1st Regiment of Seville—the corps which furnished the sallying force—was killed. Next day the siege-works were so little injured that the artillery was able to put guns into the first batteries that had been marked out. On the first three days of February incessant and torrential rains stopped further work—the whole of the first parallel was inundated, and the flying bridge by which alone Soult could communicate with Latour-Maubourg on the other side of the Guadiana was washed away.

But despite of the rain February 3 was a day of joy for the French, for on its morning Gazan reported his arrival at Valverde, ten miles away, and at 3 o’clock his division of 6,000 men marched into camp and doubled the force of the besieging army. Their arrival was a piece of cruel ill-luck for the Spaniards, for on that same afternoon, at dusk, Menacho sent out a formidable sortie of 1,500 men—all that he could safely spare from the ramparts—who came out of the river-side gate (Puerto de las Palmas) and stormed the first parallel, driving[p. 42] out the workers and the three companies of their covering party. The Spaniards had already filled up a considerable section of the trench, when they were charged by two battalions of Gazan’s newly-arrived troops, and driven out again, before they had finished their task. The serious nature of the attack may be judged from the fact that the French lost 188 killed and wounded—including eight officers—in repelling it. If only one brigade of Girard had been in the Cerro del Viento camps, instead of Gazan’s entire division, it is probable that the whole first parallel and the batteries behind it would have been destroyed. While the damage was being repaired, on February 4, Soult began to bombard the town from these batteries, but with no good effect. The result, indeed, was rather to the profit of the Spaniards, for a great portion of the civil population fled at the first sign of bombardment, and escaped by night down the Guadiana bank towards Elvas. The provisions left in their deserted houses added appreciably to Menacho’s stores.

The work of extending the first parallel diagonally toward the Pardaleras was still going on, when, on February 5th, the whole situation before Badajoz was changed by the appearance in the neighbourhood of a Spanish army of succour. Even before Soult had started from Seville at the New Year, Wellington had been aware of the imminence of the invasion of Estremadura, and had been consulting with his colleague La Romana as to the measures that it would be necessary to take[55]. As early as the 2nd of January La Romana had sent orders to Mendizabal, to tell him that if the French should cross the Sierra Morena in force, he was to evacuate Southern Estremadura, break the bridges of Medellin and Merida, and endeavour to defend the line of the Guadiana[56]. By later instructions (January 8) Mendizabal was directed to retire into the Sierra de San Mamed if the enemy crossed the river above Badajoz,[p. 43] but to throw himself upon their rear, and to hang on to them, if they crossed below, and seemed to be making for Elvas and Portugal. On the 12th, Wellington, hearing that Soult seemed to be heading towards Olivenza rather than Merida, conceived doubts as to whether he might not be intending to abandon his communications with Seville, to leave the fortresses behind him, and to march to the Tagus to co-operate with Masséna upon the Alemtejo bank of that river. On the 14th arrived the more comfortable news that the French had sat down to beleaguer Olivenza, a sure sign that they did not propose to cut themselves loose from their base and to join Masséna as a flying column. As a matter of fact, as we have seen, Soult, having been deprived of Gazan’s assistance, was too weak at this moment to dream of an incursion into Portugal, and had attacked Olivenza because he could find nothing else to do for the present.

Acc............