On the next day but one Mendizabal withdrew from Badajoz the divisions of Carlos de España and Virues, and part of that of Garcia[70], leaving the original garrison strengthened by the remainder of the last-named unit up to a force of 7,000 men. The field army retired across the river, and encamped on the strong position of the heights of San Cristobal, its right wing resting on the fort, while the remainder of its camps lay along the reverse slope of the range for a distance of a mile and a half. There were some 9,000 infantry on the position, and the 3,000 horse of Butron and Madden were encamped behind it in the plain of the Caya. By some inconceivable folly Mendizabal made no attempt to use this large force of cavalry, which he should have sent forward to seize and hold the valley of the Gebora, in front of his position. All beyond that stream, which flows at the very foot of the San Cristobal heights, was abandoned to Latour-Maubourg. It seems certain that the French cavalry general could have been driven to a respectful distance if a force of all arms had been sent against him, for he had on the north of the Guadiana only five regiments of horse and not a single battalion of infantry. But the Spaniard allowed himself to be cooped up on the hill, and kept no guard of cavalry far out in the plain to shield his front and report the motions of the enemy. What was worse, he made no attempt to entrench the long hillside, though this was a point on which Wellington and La Romana had given very clear and definite instructions. The position was strong, but as no care was taken to keep the enemy at a distance, it was always possible that he might make a sudden dash at it, and the Spanish army—scattered in its camps—would require time to take up its ground and form its fighting-line.

For some days after the sortie, however, Soult paid little attention to Mendizabal, and concentrated all his efforts against the fortress. Having completed the first parallel, and established several new batteries in it, he proceeded with his operations against the Pardaleras fort. His plan was very daring—not to say hazardous—for on the afternoon of the 11th of February, when the work was much battered but still quite defensible, he determined to try to capture it by escalade. At dusk two columns, making about 500 men, issued from the trenches and dashed at the Pardaleras: the left-hand column coasting round its flank made for the gorge, which was only defended by a row of palisades. These were so weak that they were broken down or hewn to pieces by the assailants without much difficulty. At the same time the right column, which had entered the ditch, found an open postern into which it made its way. Attacked on two sides, the garrison evacuated the work, and fled into the city, leaving 60 men killed or prisoners behind them. The French, who had lost only 4 killed and 32 wounded in this reckless venture, established themselves in the Pardaleras. But the governor turned against the fort all the guns of the next two bastions, and the captors had to burrow and lie low, till on the night of the 12th-13th a trench was run out from the first parallel, which gave safe ingress and egress. During the intervening day the besiegers lost more men in holding the work than they had in storming it[71], and the Pardaleras, close though it was to the walls, proved to be ground from which it was most difficult to push forward while the artillery fire of the town was unsubdued. To transform the open gorge in its rear into a base for new approaches was a slow and expensive business, and the siege made a much less rapid progress than had been hoped.

Meanwhile Soult resolved to make a blow at Mendizabal and his field army, which was visible day after day encamped on the San Cristobal heights, in a position imposing but unfortified and ill-watched. The Marshal had intended to cross the Guadiana and deliver his attack even before the Pardaleras was taken, but much rain was falling, and the river had overflowed its banks, so that access to the point where the French flying-bridge had been established, a mile above Badajoz, was difficult. Moreover the Gebora was also in flood, and reported to be unfordable, though usually a slender stream. The only thing which the Marshal was able to do between the 11th and the 18th of February was to shell the nearer end of the heights of San Cristobal from the batteries in his right attack, with the object of inducing the Spanish battalions there encamped to move further from the protection of the fort, which effectually covered their right flank. On February 13th this plan was seen to have been effective: the Spaniards had withdrawn from the neighbourhood, and had left half a mile unoccupied between San Cristobal and their new camp.

On the afternoon of the 18th it was reported to Soult that both the rivers had fallen, and that the Gebora had again become fordable. He made no delay, and at dusk his striking-force began to cross the Guadiana—the operation was slow, since only two flying-bridges and a few river-boats were available. But at dawn nine battalions[72], three squadrons, and two batteries were on the north bank, while Latour-Maubourg had come up from his usual post at Montijo with six cavalry regiments more. The whole force assembled in the angle between the Guadiana and the Gebora amounted to no more than 4,500 infantry[73] and 2,500 horse, with twelve guns, a total so much below Mendizabal’s 9,000 infantry and 3,000 cavalry that the adventure seemed most hazardous. But fortune often favours the audacious, and on this day Soult chanced on the unexpected luck of a very foggy morning. He was practically able to surprise the enemy, for the first warning that a battle was at hand only came to Mendizabal when, shortly after dawn, his picket at the broken bridge of the Gebora—a short mile or so in front of the heights—was driven in by masses of French infantry. At the same moment a tumult broke out on his left rear: the 2nd Hussars, sent on by Latour-Maubourg to discover and turn the Spanish northern flank, had been able to mount the heights unseen and unopposed, by making a long détour, and rode unexpectedly into the camp of one of Carlos de España’s regiments. Mendizabal’s troops flew to arms, and began hastily to form their line upon the heights, but they had no time to get into good order, for the enemy was upon them within a few minutes.

Mortier, to whom Soult had committed the conduct of the battle, showed great tactical skill. On reaching the line of the Gebora, infantry and cavalry poured across it without a moment’s delay; all the three fords of which the French knew proved to be practicable, though on the southernmost one, near the bridge, the infantry had to cross with the chilly water up to their waists. The order of battle was very simple: the right wing, composed of the whole of the cavalry, was to pass by the most northern ford and ascend the heights beyond the Spanish left. Arrived at the crest, one brigade was to push along it, and fall on the flank of the hostile line, while the other descended into the valley of the Caya, and charged into the Spanish camps, so placing itself directly in Mendizabal’s rear. Of the infantry three battalions (the 100th regiment) were to ascend the hillside in the gap between the fort of San Cristobal and the nearest Spanish camp, a gap which had been caused (as it will be remembered) by the withdrawal of the Spaniards from the southernmost heights under the stress of bombardment six days before. This column was to risk the fire of the fort, which it had to disregard, and fall on the hostile flank. Meanwhile the centre—very weak and composed only of six battalions of infantry (34th and 88th regiments)—was to attack the Spanish front, when the two turning movements were well developed.

The San Cristobal heights are a most formidable position, two miles of smooth steep slopes with an altitude of 250-300 feet, overlooking the whole plain of the Gebora and with hardly any ‘dead ground’ in their sides. They form an excellent glacis for an army in position ready to defend itself by its fire, for the assailant must come up the hill at a slow pace and utterly exposed. Cavalry could only climb at a walk, and with difficulty; but Mortier had sent all his horse far to the north, where they ascended, and partly crossed, the range at its lowest point, beyond the extreme flank of the enemy. Just at this moment the fog rose, and everything became visible. On gaining the heights unopposed, Briche’s light cavalry formed up across them, and commenced to move along the summit towards the Spanish left wing, while Latour-Maubourg, with three dragoon regiments, descended the reverse slope and moved towards the hostile camp, in front of which Butron’s Spanish horse and Madden’s Portuguese could be seen hastily arraying their squadrons.

It may be said that the battle of the Gebora was lost almost before a shot had been fired, for on seeing themselves threatened in flank and about to be charged by Latour-Maubourg, the Spanish and Portuguese horse broke in the most disgraceful style, disregarding the orders of their commanders, and went off in a disorderly mass across the plain of the Caya, towards Elvas and Campo Mayor. They outnumbered the enemy, and could have saved the day if they had fought even a bad and unsuccessful action, so as to detain the French dragoons for a single hour. But the cavalry of the Army of Estremadura had a bad reputation—they were the old squadrons of Medellin and Arzobispo, of which Wellington preserved such an evil memory, and Madden’s Portuguese this day behaved no better[74]. They escaped almost without loss, for Latour-Maubourg let them fly, and turned at once against the flank and rear of Mendizabal’s infantry.

The combat on the southern part of the heights had not yet assumed a desperate aspect. Though the column which was formed by the 100th regiment had got up the hillside under the fort of San Cristobal, and had penetrated into the gap between that work and its extreme left, the Spanish infantry was still holding its own. The fog having cleared, they were able to estimate the smallness of the number of the hostile infantry, and stood to fight without showing any signs of failing. But the fusilade was only just beginning all along the hillside when the victorious French cavalry came into action. Briche’s light horse came galloping along the crest of the heights, while Latour-Maubourg’s dragoons were visible in the plain behind, well to the rear of the Spanish line. Mendizabal, horrified at the sight, ordered his men to form squares, not as usual by battalions, but vast divisional squares, each formed of many regiments, and with artillery in their angles. If the French cavalry alone had been present, it is possible that in this formation the Spaniards might have saved themselves. But Mortier’s infantry was also up, and well engaged in bickering with Mendizabal’s men. The squares, when formed with some difficulty, found themselves exposed to a heavy fire of musketry from the front at the same moment that the cavalry blow was delivered on their flank. Briche’s hussars penetrated without much difficulty through battalions already shaken by the volleys of the French infantry. First the northern and soon afterwards the southern square was ridden through from the flank and broken. The disaster that followed was complete: some of the Spanish regiments dispersed, many laid down their arms in despair, a limited number clubbed together in heavy masses and fought their way out of the press towards the plain of the Caya and the frontier of Portugal. General Virues and three brigadiers were taken prisoners, with at least half of the army. Mendizabal and two other generals, La Carrera and Carlos de España, got away, under cover of the battalions which forced their passage through toward the west. In all about 2,500 infantry escaped into Badajoz[75], and a somewhat smaller number towards Portugal[76]. The rest were destroyed—only 800 or 900 had been killed or wounded, but full 4,000 were taken prisoners, along with seventeen guns—the entire artillery of the army—and six standards. The French loss, though under-estimated by Soult in his dispatch at the ridiculous figure of 30 killed and 140 wounded, was in truth very small—not exceeding 403 in all. It fell almost entirely on the cavalry—who had done practically the whole of the work. The regimental returns show that only four officers in the infantry were killed or wounded, to thirteen in the mounted arm. In proportion this battle was more disastrous to the vanquished and less costly to the victors than even Medellin or Ocaña. It is difficult to write with patience of the culpable negligence of Mendizabal, in allowing himself to be surprised in such a position, when he was amply provided with cavalry, or of the conduct of the Spanish and Portuguese horse in abandoning their infantry without striking a single blow, when even a show of resistance might at least have given their comrades time to save themselves. For the battalions on the heights could have escaped into Badajoz, or even have retreated along the Guadiana without desperate loss, if they had been granted an hour’s respite: while if the French cavalry could have been detained till the infantry battle on the heights was decided, it is quite clear that Mendizabal in his splendid fighting position, and with double numbers, could have held his own and driven off the attack of the nine battalions of Mortier.