The co-operation of the Spaniards was the crucial point. Unless it were assured, Wellington considered that Beresford must assume the more cautious and defensive attitude. If it were secured, the bolder policy might be pursued. The lines which Wellington laid down were in the main those which had already been suggested by Beresford: (1) Castaños must undertake to keep the horse of Villemur in the Sierra Morena, closely observing Latour-Maubourg, but forbid him to engage in any fighting; he must retire if pressed; the infantry of the 5th army must stay at Merida, as at present, but be ready to join Beresford if Soult invaded Estremadura. (2) Ballasteros was to take a similar position on the other flank, with his head quarters at Burguillos (near Zafra)[351] and his advanced posts at Fregenal and Monasterio; if Soult moved forward, he was to join Beresford without attempting to fight. (3) When Blake’s army had landed, it was to pass up the Guadiana, and take post at Xeres de los Caballeros; on any alarm from Soult, it was (like the other Spanish troops) to join Beresford at once. If these arrangements worked, at least 15,000 Spaniards would be in line at Albuera, the chosen position, to assist in holding back Soult. And, as we shall see, the scheme did work exactly as Wellington had designed, and the whole force was collected. (4) Lastly, and this was all-important, when the allied forces were concentrated, they must be placed under a single commander, and not worked with divided authority and divided responsibility, as had been the case in the Talavera campaign of 1809. Concerning Blake there could be no difficulty, as he was junior to Castaños, and the latter had consented to place himself at Beresford’s disposition when they met at Jerumenha on March 30th. Too much praise cannot be given to his reasonable and conciliatory conduct, which alone rendered possible the co-operation of all the allied forces during the ensuing campaign.

Wellington therefore, as is clear, foresaw the whole course of subsequent operations, and even fixed the exact battle-spot on which the fate of Soult’s attempt to relieve Badajoz would be decided. The only point left to Beresford’s decision was whether the strength of the French army was such as to render a successful resistance possible. And when we come to consider the respective forces at the disposition of the two parties, it can hardly be urged that Beresford was wrong to accept battle. That his victory was a hard-fought and costly one came from minor tactical circumstances, which will be explained in their due place.

As to the details of the projected siege of Badajoz, Wellington laid down an equally clear policy. All guns and material were to be collected in Elvas, Campo Mayor, and Olivenza, and not to move till everything was ready. The main communications of the army were to be across a floating bridge to be constructed at the junction of the Caya and the Guadiana, five miles below Badajoz and six from Elvas. The permanent bridge of Merida and the temporary bridges at Jerumenha would be subsidiary resources. Lastly, and here was the most important point, the general scheme to be pursued was that the besiegers should first reduce the outlying defences of Badajoz, Fort San Cristobal on the north bank of the Guadiana, the Pardaleras and the Picurina on the south bank. Only when all these were taken would operations against the city itself be begun. To quote the concluding paragraph of Wellington’s memorandum: ‘When the British army shall be in possession of San Cristobal, Picurina and Pardaleras, Marshal Beresford will determine upon the point at which he will attack the body of the place. It is believed that, upon the whole, one of the south faces will be the most advantageous.’

There can be no doubt that all the mishaps of the two first British sieges of Badajoz had their origin in these original orders of Wellington, which were drawn up on the advice of his chief engineer, Colonel Fletcher. The great mistake was the choosing of the almost impregnable fort of San Cristobal as one of the three first points of attack, and the making all subsequent operations depend upon its capture. No doubt the possession of this lofty and commanding work would render the fall of Badajoz certain, since it overlooked the castle and all the northern end of the city. But it was the strongest part of the whole defences, and when the miserable and antiquated train of artillery at Beresford’s disposition is taken into consideration, and it is remembered that the siege was to be conducted ‘against time’ as it were, i. e. with the hope that it might be concluded before Soult could collect a relieving force, it is clear that San Cristobal ought to have been left alone. The other points designated by Wellington, the Pardaleras and Picurina, were much more accessible, and the capture of one or other of them would have brought the besiegers close to the walls, though neither of them commanded the whole city in the same fashion as San Cristobal. The best commentary on the sieges of 1811 is that a year later, at the third and successful leaguer, Wellington left the high-lying fort on the other side of the Guadiana entirely alone. The original orders of 1811 gave three separate and distinct objectives, and none of these were to be mere ‘false attacks,’ since it is distinctly said that operations against the enceinte were only to begin when all three of the forts were in British hands. Wellington was not a trained engineer; he was dependent on the advice of the officers of that arm, and it seems that they gave him bad counsel, as they certainly did to Beresford during the subsequent weeks. The most puzzling thing is to make out why Colonel Fletcher and his colleagues ignored Soult’s precedent; the French engineers had concentrated their attack on the Pardaleras front as their sole objective, for their other operations were false attacks. The English engineers, instead of concentrating their efforts in the same way, wasted their work on three separate points, which was all the more inexcusable because they knew that the resources which their comrades of the artillery arm had at their disposition were most inadequate for a great siege. Hence came a very costly and deplorable series of failures.

On the day of Wellington’s departure from Elvas heavy rain fell, and the Guadiana rose high, not merely washing away the cask-bridge at Jerumenha, but rendering the working of the flying bridges impossible. This was a serious matter; not only did it put a temporary stop to the communications between Elvas and the army, but it raised the question as to what might happen if a similar mischance were to occur when Soult was invading Estremadura. For if the Jerumenha bridges should break when the Allies were concentrated at Albuera, they would have no line of retreat and an impassable river behind. Beresford, with this possibility in his eye, ordered an alternative line of communication to be established via Merida, and sent a brigade of the 2nd Division thither to reinforce the Spaniards of Morillo, and a day later the whole 4th Division. His anxiety on this point did not cease till, on the 5th of May, a strong pontoon bridge had been built at the point selected by Wellington, the place where the Caya falls into the Guadiana. By this time the Jerumenha bridges were again in working order, but it was clear that it would be a mistake to trust the whole safety of the army to them.

It was only on this same day (May 5) that Colonel Fletcher and Major Alexander Dickson reported to Beresford that they were ready to produce the means for the attack on Badajoz: the former had his stock of platforms, fascines, and gabions prepared; the latter had organized the first convoy of artillery and ammunition from Elvas. On the 6th, therefore, the investment of Badajoz on the south side of the Guadiana was completed by the British brigades of Lumley and Alten and the Portuguese brigade of Fonseca, while on the following day the brigade of Kemmis and the 17th Portuguese (part of the garrison of Elvas) appeared opposite San Cristobal, and shut in the place on the northern bank of the river. The rest of the infantry[352] encamped as a support of the besieging force in the woods between Badajoz and Albuera, but the cavalry still remained in southern Estremadura, and Colborne’s brigade had been for some days (April 30th-May 11th) executing a demonstration in the Sierra Morena, with the object of keeping Latour-Maubourg employed. This last operation, owing to Colborne’s skilful management of his column of 2,000 infantry and two squadrons each of Spanish and Portuguese horse, had been very successful. On hearing of British infantry in his front, the French general evacuated his posts on the crest of the mountains, Guadalcanal, Fuente Ovejuna, Azuaga, and Monasterio, and fell back south-eastward towards Constantina on the Cordova road, abandoning the direct line of retreat on Seville, which he had hitherto covered. Colborne, after clearing all these places, extended his march eastward into a very wild and unexplored country, and summoned the isolated castle of Benalcazar, the only French garrison left north of the Sierra Morena. When it refused to listen to a summons, he had to leave it, having neither guns to batter it nor time to waste. From thence he returned by a circular sweep through Campanario to Almendralejo, where he once more was in touch with the British army (May 11th).

The first episodes of the siege of Badajoz were not very encouraging to the besiegers. On the 8th trenches were opened opposite all the three points of attack designated by Wellington, the Picurina and Pardaleras forts on the south side, and San Cristobal on the north. In each case the first parallel was started at about 400 yards from the walls; Dickson had told off fourteen 24-pounders and two 8-inch howitzers for the work on the left bank, five 24-pounders and two other howitzers for the attack on San Cristobal. More energy was displayed in this last quarter than in the others, apparently from the notion that if this commanding work could be subdued the rest of the siege would be an easy matter. But the results were disappointing: on the stony slopes of San Cristobal there was little earth to throw up, and the spade gritted against rock at three inches from the surface. At the end of the first night’s digging the trench was but a seam, and there was only one section at which about ten men could work under cover. The rest had to be abandoned during daylight, for the enemy kept up a furious fire, not merely from the fort, but from the citadel on the other side of the river, whose flank battery enfiladed the projected trench. Three out of nine engineer officers present on this front were killed or wounded in the first twenty-four hours, and many of the workers from Kemmis’s brigade and the 17th Portuguese. It soon became obvious that the trenches would have to be built with earth from a distance and gabions, rather than excavated. Nevertheless, a battery for Dickson’s five 24-pounders was sketched out, and began to be visible to the enemy. On the night of the 10th Phillipon sent up a reserve battalion into San Cristobal, and executed a sortie upon the British works. It penetrated into the trenches, but was driven out after a sharp struggle by the covering party. But, pursuing too far, the British came under the guns of San Cristobal, and had to retire to their trenches with lamentably heavy loss[353]. Next day the battery was completed, despite of a deadly fire both from the fort and the castle, and opened upon its objective. But it was completely overmastered, and before night four of its five guns had been damaged or dismounted; three more engineer officers were hurt, leaving only three surviving of the original nine. It is said that the battery opened before it had been intended—a fault of over-zeal on the part of the Portuguese major in command. Beresford’s purpose had been to wait till the other attacks, on the Picurina and Pardaleras, were ready, before beginning to batter San Cristobal. These attacks had met with less difficulty, the ground being easier to dig, and on the evening of the 11th the trenches in front of the Picurina were well advanced, and the battery of ten guns there opened upon the fort with some, but not great, effect.

Seeing the San Cristobal attack faring so badly, the engineers got leave to erect a second battery on that side, further down the hill, which was intended to check the enfilading fire from the other side of the Guadiana. More guns were brought up to the original battery, to replace those that had been damaged. But both batteries were overpowered and badly maltreated on the morning of the 12th. A few hours later news arrived from the south, sent by Ballasteros, to the effect that the French were in motion from Seville with a large relieving army, and were marching hard across the Sierra Morena, just as Wellington had expected; they had reached Santa Olalla on the 11th, and were already in touch with Latour-Maubourg. Since their force was estimated at only 23,000 men—not far from the real amount—Beresford resolved to fight, and sent requests to Castaños, Ballasteros, and Blake to concentrate on Albuera, the battle-ground selected by Wellington. It was fortunate that Blake was now in close touch and available—he had reached Fregenal on the 9th and Barcarrota on the 12th, so that his arrival was certain, unless some unforeseen accident should occur. Without his aid it would have been doubtful policy to wait for Soult and risk a general action: but with his 10,000 men in line the Allies would have wellnigh 35,000 men available, if every unit came complete to the field.

Meanwhile, pending the confirmation of the news of the French advance, Beresford’s engineers asked leave to open another parallel, and 1,400 men had been paraded for the purpose of starting it, when complete details as to Soult’s progress came to hand. It had been so rapid that the Marshal at once ordered all the siege operations to be discontinued, though the engineers tried to persuade him to risk two days’ more work, by the vain promise that they would undertake to produce two practicable breaches in that space of time. Beresford wisely refused to listen to them, and ordered that all the guns and ammunition should be returned at once to Elvas, with such of the siege stores as could be readily moved. But the mass of gabions and fascines had to be burnt, as these would be profitable to the garrison, and would certainly be carried into the town if they were left intact. The troops, English, Portuguese, and Spanish (three battalions of Castaños’s infantry had come up from Merida), were ordered to prepare to march for Albuera in successive detachments, the 4th Division and the Spaniards being left to the last in the trenches, to cover the removal of the guns and stores.

Beresford’s total casualties in this mismanaged fragment of a siege, from May 6th to May 12th, had been 533 British and 200 Portuguese, or 733 in all, lost in the trenches and in the sortie. It will be seen that the sortie cost far more lives than the actual beleaguering work. All the British loss save seven casualties was in Kemmis’s brigade of the 4th Division[354], which lay on the Cristobal side, and suffered both in the trench-building and in the sortie. The Portuguese loss was partly in the 17th Line, which acted with Kemmis, partly among the artillery.