From the first part of Napoleon’s calculations it is clear that he thought Soult would require about 25,000 men—the 15,000 who were to be left about Badajoz and the 10,000 who were to be brought up from Andalusia: they are increased to 30,000 by erroneous addition only. As a matter of fact Soult, in order to cover Seville and to rescue Victor, had left only 11,000 men in Estremadura on March 14th, and 3,000 of these were now shut up in Badajoz. But on the other hand he collected from Andalusia (including Maransin’s column)[460] 16,000 men, so that his fighting force was within a few hundreds of the 25,000 named by the Emperor. The 35,000 spoken of in a later sentence would only be required, so Napoleon thought, if Wellington came down to invade Andalusia with all his British troops. We may point out, by the way, that the Imperial calculations were all wrong in detail, as was bound to be the case when they were made at Paris on data many weeks old. Wellington, owing to the reinforcements which had landed at Lisbon in the first days of March, had now nearly 40,000 British troops. He had detached 12,000 of them under Beresford[461], and these were accompanied not by 6,000 Portuguese at the most, as the Emperor guessed, but by a full 10,000. There was therefore a serious miscalculation. We may add that if Wellington had taken the unlikely step of concentrating his whole army against Andalusia, he would have had not only 38,000 British troops with him, but nearly 25,000 Portuguese troops of good quality[462]. The united force could have smashed up in one morning’s work the 35,000 French under Soult, whom the Emperor thought enough to restrain them. But, as Napoleon truly observed, it was practically impossible for Wellington to make this move, so long as Masséna’s force was still opposed to him in the north. It was only when the Army of Portugal moved down to the Guadiana in June that the British general concentrated practically his whole force in one line, behind the Caya, in the southern sphere of operations. And when he had done so the Armies of Portugal and Andalusia united, though about 60,000 strong, did not dare to attack him. But of this more anon.

In early May Soult, under-valuing Beresford’s fighting force, thought that 25,000 men would suffice to sweep him behind the Guadiana, even when he had the help of Castaños and Ballasteros. The force was collected in the following fashion: Latour-Maubourg at Constantina had 8,000 men, who had just been rejoined by Maransin’s column, thus the 5th Corps was once more concentrated and complete (with the exception of five battalions in Badajoz, and one or two more which Soult was bringing up from Seville). When all came in, the corps amounted to 10,000 men of all arms. The remaining part of the expedition was made up by requisitioning from Victor’s 1st Corps and the lines before Cadiz four battalions and two regiments of cavalry[463], and from Sebastiani’s 4th Corps four battalions and three regiments of cavalry[464]. Of the independent division under Godinot, which garrisoned the kingdom of Cordova, Soult took nearly the whole—nine battalions and two regiments of cavalry[465]. The borrowed troops were divided into two large brigades (one might better have been called a small division) under Generals Godinot and Werlé[466], and three provisional brigades of cavalry. They took five batteries (thirty guns) with them, to add to the eighteen guns of Latour-Maubourg, and some companies of sappers and train. The total force available came to just under 25,000 men, unequally divided in numbers between the 5th Corps and the Andalusian reserves. The cavalry was very strong in proportion, about 4,400 sabres[467].

It will be noted that Soult had (as in January) refrained from adopting the general plan which Napoleon favoured, that of abandoning all Andalusia save Seville and the Cadiz Lines, and leaving only a corps of observation at Granada. It is true that the Duke of Dalmatia took very little from Victor, and left the 1st Corps almost intact in the lines, but Sebastiani’s 4th Corps was also left only slightly diminished, and was expected to hold down all eastern Andalusia, instead of being requisitioned for 10,000 men and reduced to a flying column as the Emperor would have wished. The unit that had been most heavily drawn upon was the garrison of the kingdom of Cordova, and the result of this was that (as in January) very few troops could be detailed for the defence of Seville, since nearly all that had been in its neighbourhood were summoned off to Estremadura. The great city which formed Soult’s base was once more left inadequately defended by dépôts and drafts, and Juramentados of doubtful fidelity. The Marshal had lately raised some companies of so-called Swiss[468], deserters of all nations, and these were also utilized. But the total left under General Daricau was dangerously small. So keenly was this realized that the governor was directed to retire within the great fortified enclosure of the Carthusian convent (La Cartuja) if pressed, and all the military stores had been placed in this immense building, which had been surrounded with a bastioned enceinte, and armed with cannon, so as to form a sort of citadel. The Castle of the Inquisition at Cordova in a similar fashion had been fortified, and converted into a work that could be held against any irregular force, and similar precautions had been taken at Jaen, Andujar, Ronda, Alcala Real, and Niebla, to provide centres of resistance against possible assaults by guerrilleros. Probably, however, Napoleon was right, and if the minor places and eastern Andalusia had been evacuated, Soult might have brought 10,000 more infantry against Beresford, in which case the latter would never have dared to fight him, and must have retired behind the Guadiana. There would have been no battle of Albuera—but on the other hand all the evacuated districts would have flared up into insurrection, and it is difficult to see how Soult could have reconquered them, since he was to be for several months tied up in operations against the British, from which he could not have withdrawn a man.

But having taken another decision, and resolved to surrender nothing, the Marshal had only gathered 12,000 men to reinforce the 5th Corps. They required many days to concentrate, and it was only on May 8th that he reviewed them in their new provisional brigading at Seville, and delivered an allocution in which he announced to them that they were destined to save Badajoz and drive the British from Estremadura, and that the force would march at midnight on the 10th[469]. This threat did not escape the Spanish patriots in the city, who passed the news on so swiftly that Ballasteros was able to forward it to Beresford by the afternoon of the 12th.

Having once started, Soult hoped to surprise his enemy by the swiftness of his movements. The head of the column which marched at 12 p.m. on the night of the 9th-10th was at Santa Olalla, more than thirty miles away, on the 11th. The pace had to slacken in crossing the Sierra Morena, but on the 12th head quarters were at Monasterio (fifteen miles further on) from which Ballasteros’s scouts withdrew. Latour-Maubourg and the 5th Corps, far away to the right, had advanced at the same time from Cazalla and Constantina, and driven Castaños’s advanced posts from Guadalcanal and then from Llerena. On the 13th the two French forces joined at Fuente Cantos, and their leading cavalry squadrons reached Los Santos, from which the 13th Light Dragoons retired. As Wellington had directed, nearly a month before, in his Elvas memorandum, the Spaniards made no attempt to check the advance: their cavalry withdrew as the French pushed forward; their infantry were prepared to fall back on the rendezvous at Albuera.

From the 13th the British cavalry as well as the Spanish were in touch with Soult; General Long had been lying about Villafranca and Los Santos till that day, with three British and four Portuguese regiments[470]. He retired to Fuente del Maestre, and then to Santa Marta, contenting himself with reporting the successive advances of the French to Beresford, who was apparently not over well contented with his operations on this and the two following days, and thought that he might have gone back more slowly, and have compelled the leading squadrons of the enemy to deploy and lose time. At Fuente del Maestre the allied cavalry split itself up, Madden with two Portuguese regiments covering the roads to Almendralejo and Solana, while Long and the main body stayed on the high-road to Badajoz via Santa Marta and Albuera[471].

Having such long warning of his adversary’s movements, Beresford was able to carry out the concentration of his fighting force at leisure. There was still some uncertainty as to which road the enemy might choose, three[472] being open to him when his advance had reached Los Santos, viz. (1) the obvious central and shortest route by Albuera-Badajoz, (2) the eastern route Solana-Talavera Real-Badajoz, (3) the western route Almendral-Valverde-Badajoz. The former was rather circuitous, its main advantage to the French being that it was all across open flat country, where their superior cavalry would have had excellent ground; but the Albuera route was not perceptibly inferior in this respect, as subsequent operations showed. To take the third or western road, that by Almendral-Valverde (though this is not so long as that by Solana) would have forced Soult to execute a flank march across Beresford’s front, and (what probably weighed more with the French Marshal) would have fixed the decisive spot, where the fate of the campaign would be settled, nearer to the point towards which Blake’s army was known to be marching: and Soult still hoped to fight his battle in that general’s absence.

On the 13th of May Beresford marched out from his lines in front of Badajoz to Valverde, a point convenient for occupying a position across two of the possible roads, and not very remote from the third and least likely one. He took with him the 2nd Division and Hamilton’s Portuguese, with three batteries. The rest of the army remained before Badajoz, covering the removal of artillery and stores, but ready to come up at any moment. On the same afternoon he had a conference with General Blake, who rode over from Barcarrota. On the following day Soult’s movement seemed to be growing much slower—the heads of his columns only reached Fuente del Maestre and Villafranca. The fact, duly reported to Beresford, that part of the French army had reached the latter place, which is off the main chaussée to the right, seemed to make it possible that Soult was, after all, going to move by Talavera Real. Beresford waited for a more precise indication of his adversary’s final route, and sent pressing orders to Madden to cover with his scouts all the open country between Talavera Real and Almendralejo. Blake on this day, finding that his cavalry could discover no signs of the French in his front, to the west of the great chaussée, drew in from Barcarrota to Almendral, as he promised to do when he met Beresford at Valverde on the 12th.

It being perfectly clear by this time that the French were not about to take the western route, Beresford on the 15th marched the 2nd Division and Hamilton’s Portuguese to Albuera, where they were joined by more troops from in front of Badajoz, Alten’s German brigade, and a provisional brigade of Portuguese under Colonel Collins[473]. Only the 4th Division and the Spanish brigade belonging to Castaños, lately arrived from Merida, now remained in front of the fortress—all on the south bank save Kemmis’s brigade of the first-named unit. The last of the stores were moved on this day from the trenches to Elvas, and the flying bridge opposite the mouth of the Caya was taken up. This last proved a mistake—it was intended that Kemmis should join the army by using a ford below Badajoz, which had been practicable for the last ten days; but on the night of the 15th-16th the water rose, and the brigade was forced to march round by the next passage, that at Jerumenha, which involved a circuit of thirty miles, and made it late for the battle. Only three companies, which chanced to be on the south bank of the Guadiana when the freshet came down, were able to march off with Cole and the rest of the division, when the order came.

About 15,000 men were already in line at Albuera when Soult’s intention at last became perfectly clear: his chasseurs and hussars vigorously attacked Long’s horsemen at Santa Marta, and began to drive them along the chaussée. Long made no stand, though, having three British and two Portuguese regiments (Otway) besides some 600 of Castaños’s cavalry, he was in considerable strength. ‘He was driven rather faster than one could have wished, and retiring precipitately crossed the Albuera stream, and gave up the whole right bank to the enemy. This haste is a bad thing, because the woods there mask all the enemy’s movements,’ wrote D’Urban in his diary. Beresford was so vexed with him that he that night assigned to the command of the cavalry of the whole allied army General Lumley, who was Long’s senior, leaving the latter in nominal command of the British horse alone. Lumley, though then in charge of an infantry brigade in the 2nd Division, was an old light dragoon, and showed himself next day well able to manage a mass of mounted men[474].