Two months elapsed between Wellington’s passage of the Agueda on his offensive march into the kingdom of Leon, and his triumphal entry into Madrid. During this critical time there had been constant alarms and excursions in Andalusia and Estremadura, but nothing decisive had occurred. This was all that Wellington wanted: if employment were found for the French Army of the South, so that it got no chance of interfering with the campaign on the Douro, he was perfectly satisfied, and asked for nothing more.

It will be remembered that his instructions to Hill, before he started on the march to Salamanca, were that Soult must be diverted as far as possible from sending troops northward. The main scheme was that Ballasteros and Hill should, if possible, combine their operations so as to bring pressure upon the enemy alternately[669]. The Cadiz Regency had readily agreed to stir up the Spanish general to activity: if he would demonstrate once more (as in April) against Seville, so as to attract Soult’s attention, and cause him to concentrate, Hill should press in upon Drouet and the French troops in Estremadura, so as to force the Marshal to draw off from the Spaniard. Similarly, if Soult should concentrate against Hill, Ballasteros was to strike again at Seville, or the rear of the Cadiz Lines, which would infallibly bring the Marshal southward again in haste[670].

When Wellington crossed the Agueda [June 13] Hill had his corps collected in central Estremadura—head-quarters at Almendralejo, the troops cantoned about Ribera, Villafranca, Fuente del Maestre, and Los Santos, with Penne Villemur’s Spanish horse in front at Zafra. Hill had in hand his old force—the 2nd Division and Hamilton’s Portuguese, with two (instead of the usual one) British and one Portuguese cavalry brigades. He could also call up, if needed, the three strong Portuguese infantry regiments (5th, 17th, 22nd) which were holding Badajoz till a sufficient native garrison should be provided for it. At present only a few hundred Spaniards [Tiradores de Doyle] had appeared. Far away, to the north of the Guadiana, observing the French posts on the Tagus, there was a detached Portuguese cavalry regiment at Plasencia. This outlying unit was also put under Hill’s charge: its object was to give early notice of any possible stir by the French, in the direction of Almaraz or the recently restored bridge of Alcantara. Morillo’s infantry division of Castaños’s army was lying on the right of Hill, in south-western Estremadura: Wellington suggested that the Spanish general might be willing to throw it into Badajoz, and so liberate the Portuguese regiments lying there, if Soult should advance before the regular garrison intended for the great fortress should arrive from Cadiz. The whole force watching Soult amounted to nearly 19,000 men, not including the Spaniards. Of this total about 7,500 sabres and bayonets were British—something over 11,000 were Portuguese. In addition, Morillo and Penne Villemur had not quite 4,000 Spanish horse and foot. Supposing that a minimum garrison were thrown into Badajoz—Morillo’s infantry for choice—Hill could dispose of 18,000 Anglo-Portuguese for field-operations, not including the Portuguese cavalry by the Tagus, who had the separate duty of watching the Army of the Centre.

The French in Estremadura still consisted of the old contingent which D’Erlon had been administering since the year began, viz. his own and Daricau’s infantry divisions, with Lallemand’s and Perreymond’s cavalry—altogether not more than 12,000 men, for several of the infantry regiments had lost a battalion apiece when Badajoz fell. Since his excursion to Don Benito and Medellin at the time of Hill’s raid on Almaraz, D’Erlon had drawn back, abandoning all southern and most of eastern Estremadura to the allies. He himself was lying at Azuaga and Fuente Ovejuna, on the slopes of the Sierra Morena, while Daricau was more to the north, about Zalamea, rather too far off to give his chief prompt support. Daricau’s detachment in this direction seems to have been caused by a desire to make communication with the Army of Portugal easy, if the latter should ever come southward again from the Tagus, and push to Truxillo as in 1811. It was clear that unless Soult should reinforce his troops to the north of the mountains, Hill need fear nothing: indeed he had a distinct superiority over D’Erlon.

Early in June, however, there was no danger that any troops from Seville would come northward, for Ballasteros’s diversion had taken place somewhat earlier than Wellington had wished, and the disposable reserve of the Army of Andalusia was far away in the extreme southern point of the province. After his success at Alhaurin in April, and his subsequent pursuit by Soult’s flying columns, Ballasteros had taken refuge—as was his wont when hard pressed—under the guns of Gibraltar. The French retired when they had consumed their provisions, and fell back to their usual stations at Malaga and Ronda, and along the line of the Guadalete. When they were gone, the Spanish general emerged in May, and recommenced his wonted incursions, ranging over the whole of the mountains of the South. Having received the dispatches of the Regency, which directed him to execute a diversion in favour of the allied army in Estremadura, he obeyed with unexpected celerity, and took in hand a very bold enterprise. General Conroux, with the column whose task it was to cover the rear of the Cadiz Lines, was lying at Bornos, behind the Guadalete, in a slightly entrenched camp. He had with him about 4,500 men[671]. Ballasteros resolved to attempt to surprise him, on the morning of June 1. Having got together all his disposable troops, 8,500 infantry and a few squadrons of horse[672], he made a forced march, and, favoured by a heavy mist at dawn, fell upon the enemy’s cantonments and surprised them. He won a considerable success at first: but the French rallied, and after a hard fight broke his line by a general charge, and drove him back across the Guadalete. Conroux was too exhausted to pursue, and Ballasteros remained in position, apparently meditating a second attack, when on seeing some cavalry detachments coming up to join the enemy, he sullenly retired. He had lost 1,500 men and 4 guns, the French over 400[673]. The first note of alarm from Bornos had caused Soult to send what reserves he could collect from the Cadiz Lines and Seville—six battalions and two cavalry regiments, and since Ballasteros had been beaten, but not routed, he thought it necessary to give prompt attention to him. Thereupon the Spaniard retreated first to Ubrique, and when threatened in that position, to his old refuge in the lines of San Roque before Gibraltar.

Soult would have liked to make an end of him, and would also have been glad to direct a new attack upon Tarifa, which served as a second base to the roving Spanish corps; he mentions his wish to capture it in more than one of his dispatches of this summer. But his attention was drawn away from Ballasteros and the South by the prompt advance of Hill, who (as had been settled) pressed in upon Drouet at the right moment. On the 7th June he moved forward his head-quarters from Almendralejo to Fuente del Maestre, and two days later to Zafra. On the 11th, Penne Villemur’s cavalry pushed out from Llerena towards Azuaga, while Slade’s brigade, advancing parallel with the Spanish general, pressed forward from Llera on Maguilla, a village some fifteen miles in front of Drouet’s head-quarters at Fuente Ovejuna. This reconnaissance in force brought on the most unlucky combat that was ever fought by the British cavalry during the Peninsular War, the skirmish of Maguilla.

Slade, an officer whose want of capacity we have before had occasion to notice[674], after some hours of march began to get in touch with French dragoon vedettes, and presently, after driving them in, found himself facing Lallemand’s brigade. Their forces were nearly equal—each having two regiments, Slade the 1st Royals and 3rd Dragoon Guards, Lallemand the 17th and 27th Dragoons—they had about 700 sabres a side: if anything Slade was a little the stronger. The French general showed considerable caution and retired for some distance, till he had nearly reached Maguilla, where he turned to fight. Slade at once charged him, with the Royals in front line and the 3rd Dragoon Guards supporting. The first shock was completely successful, the French line being broken, and more than 100 men being taken. But Slade then followed the routed squadrons with headlong recklessness, ‘each regiment,’ as he wrote in his very foolish report of the proceedings, ‘vying with the other which should most distinguish itself.’ The pursuit was as reckless as that of the 13th Light Dragoons at Campo Mayor in the preceding year, and resolved itself into a disorderly gallop of several miles. After the French had passed a defile beyond Maguilla a sudden cry was heard, ‘Look to your right’—a fresh squadron which Lallemand had left in reserve was seen bearing down on the flank of the disordered mass. Charged diagonally by a small force, but one in good order, the British dragoons gave way. Lallemand’s main body turned upon them, and ‘the whole brigade in the greatest disorder, and regardless of all the exertions and appeals of their general and their regimental officers, continued their disgraceful flight till victors and fugitives, equally overcome and exhausted by the overpowering heat and the clouds of thick dust, came to a standstill near Valencia de las Torres, some four miles from Maguilla, where at last Slade was able to collect his regiments, and to retire to the woods beyond Llera[675].’

In this discreditable affair Slade lost 22 killed, 26 wounded, and no less than 2 officers and 116 men taken prisoners—most of the latter wounded—a total casualty list of 166. Lallemand acknowledges in his report a loss of 51 officers and men[676]. The defeated general irritated Wellington by a very disingenuous report, in which he merely wrote that ‘I am sorry to say our loss was severe, as the enemy brought up a support, and my troops being too eager in pursuit, we were obliged to relinquish a good number of prisoners that we had taken, and to fall back on Llera.’ He then added, in the most inappropriate phrases, ‘nothing could exceed the gallantry displayed by both officers and men on this occasion, in which Colonels Calcraft and Clinton, commanding the two regiments, distinguished themselves, as well as all the other officers present[677].’

Wellington’s scathing comment, in a letter to Hill, was: ‘I have never been more annoyed than by Slade’s affair, and I entirely concur with you in the necessity of inquiring into it. It is occasioned entirely by the trick our officers of cavalry have acquired, of galloping at everything—and then galloping back as fast as they galloped on the enemy. They never consider their situation, never think of manœuvring before an enemy—so little that one would think they cannot manœuvre except on Wimbledon Common: and when they use their arm as it ought to be used, viz. offensively, they never keep nor provide for a reserve.... The Royals and 3rd Dragoon Guards were the best cavalry regiments in this country, and it annoys me particularly that the misfortune has happened to them. I do not wonder at the French boasting of it: it is the greatest blow they have struck[678].’ It is curious to find that Slade retained command of his brigade till May 1813. One would have expected to find him relegated to Great Britain at a much earlier date. But Wellington was not even yet in full control of the removal or promotion of his senior officers. Other generals with whom he was equally discontented, such as Erskine and Long, were also left upon his hands after he had set a black mark against their names.

The combat of Maguilla, however unsatisfactory in itself, made no difference to the general strategy of the campaign. Drouet, having drawn back on Hill’s advance, sent messages to Soult, to the effect that unless he were strongly reinforced he must retire from the Sierra Morena, and cover the roads to Cordova on the Andalusian side of the mountains. He reported that he had only 6,000 men in hand, and that Hill was coming against him with 30,000, including the Spaniards. Both these figures were fantastic—for reasons best known to himself D’Erlon did not include Daricau’s division in his own total, while he credited Hill with 15,000 men in the 2nd British division alone [which was really 8,000 strong, including its Portuguese brigade], and reported with circumstantial detail that the 7th Division had come down from Portalegre and joined the 2nd[679].

Soult sent on D’Erlon’s dispatch to Madrid, with the comment that Hill’s advance showed that the main intention of Wellington was certainly to attack Andalusia, and not to fall upon Marmont. But that he did not consider such an attack very imminent is sufficiently shown by the fact that he detached to Drouet’s aid only one division of infantry, that of Barrois—which composed his central reserve—and one of cavalry, that of Pierre Soult, or a total of 6,000 infantry and 2,200 cavalry: such a reinforcement would have been futile if he had really believed that Wellington was marching against Seville. His real view may be gathered from his estimate of Hill’s force at 15,000 Anglo-Portuguese and 5,000 Spaniards—a total very remote from the alarmist reports of Drouet, and not far from the truth. The reinforcement sent under Barrois would give the Estremaduran detachment a practical equality in numbers with Hill, and a great superiority in quality. The orders sent to Drouet were that he was to advance against Hill, to strive to get him to an engagement, at any rate to ‘contain’ him, so that he should not detach troops north of the Guadiana to join Wellington or to demonstrate against Madrid. If things went well, Drouet was to invest Badajoz, and to occupy Merida, from whence he would try to get into communication via Truxillo with the troops of the Army of the Centre. The final paragraph of his directions stated that Drouet’s main object must be to make such a formidable diversion that Wellington would have to reinforce Hill. ‘When the Army of Portugal finds that it has less of the English army in front of it, we may perhaps persuade it [i. e. Marmont] that the enemy’s plan is certainly to invade the provinces of the south of Spain before he acts directly against the North: then, no doubt, changed dispositions will be made.’ Unfortunately for the strategical reputation of Soult, Wellington crossed the Agueda with seven of his eight divisions to attack Marmont, on the very day after this interesting dispatch was written.

D’Erlon had been promised that Barrois should march to his aid on the 14th, but it was not till the 16th that the column from Seville started to join him, and then it marched not by the route of Constantina and Guadalcanal, as D’Erlon had requested, but by the high-road from Andalusia to Badajoz, via Monasterio. If Hill had been pressing the troops in front of him with vigour, the French would have been in an awkward position, since they were on separate roads, and might have been driven apart, and kept from junction by a decisive movement from Llerena, where Hill’s cavalry and advanced guard lay. But the British general had orders to attract the attention of Soult and to ‘contain’ as many of the enemy as possible, rather than to risk anything. He resolved, when he heard of the approach of Barrois, to retire to the heights of Albuera, which Wellington had pointed out to him as the most suitable position for standing at bay, if he were pressed hard. Accordingly he drew back by slow stages from Zafra towards Badajoz, covering his rear by his cavalry, which suffered little molestation. Barrois joined Drouet at Bienvenida near Zafra on the 19th, and their united force, since Daricau had come in to join them from the direction of Zalamea, with the greater part of his division, must have amounted to over 18,000 men, though Drouet in a report to King Joseph states it at a decidedly lower figure[680]. They advanced cautiously as far as Villafranca and Fuente del Maestre, which their infantry occupied on June 21, while their numerous cavalry lay a little way in front, at Villalba, Azeuchal, and Almendralejo. On the same day Hill had taken up the Albuera position, on which several points had been entrenched.

As Hill had just called out the three garrison regiments of Portuguese from Badajoz, he had now between 18,000 and 19,000 of his own army in position, besides Villemur’s Spanish cavalry. This last, together with Long’s and Slade’s squadrons, were thrown out in front of the Albuera river, with their vedettes in Santa Marta, Almendral, and Corte de Peleas, only a mile or two from the French advanced posts. They were directed not to give way till they were severely pressed, as Hill wished to avoid at all costs the kind of surprise that had befallen Beresford in 1811, when Long had retired so precipitately before the French horse that he could give no account of their strength, nor of the position of Soult’s infantry. But the expected advance of the enemy hung fire—from the 21st onwards Hill was waiting to be attacked, and sending almost daily accounts of the situation to Wellington: but the main body of the French moved no farther forward. This was all the more surprising to the English general because he had intercepted a letter written on May 31 from King Joseph to Drouet, in which the latter was directed to ‘passer sur le corps à Hill[681],’ and then to come up to the Tagus to join the Army of the Centre. Not knowing how entirely Soult and D’Erlon were ignoring all orders from Madrid, both Wellington and his trusty lieutenant thought that such instructions must almost certainly bring about an action. The former wrote to the latter on June 28th, after receiving several statements of the situation: ‘if you should find that Drouet separates his troops, or if he pretends to hold you in check with a smaller body of men than you think you can get the better of, fall upon him, but take care to keep a very large proportion of your troops in reserve.... I should prefer a partial affair to a general one, but risk a general affair—keeping always a large body of reserve, particularly of cavalry—rather than allow Drouet to remain in Estremadura and keep you in check.’ But the enemy neither came on for a general action, nor scattered his troops so widely as to induce Hill to risk an attack on any point of his line. He remained with his infantry massed about Villafranca and Fuente del Maestre, and only demonstrated with his cavalry.

The cause of this inactivity on Drouet’s part was partly, perhaps, his over-estimate of Hill’s strength, but much more Soult’s unwillingness to obey the orders sent him from Madrid. He was determined not to detach a third part of his army to the Tagus, to join the Army of the Centre. He was by this time fully embarked on his long course of insubordinate action, with which we have already dealt when writing of the King’s desires and their frustration[682]. On the 26th May Joseph had sent him the dispatch which directed that D’Erlon must come up northward, if Wellington’s main attack turned out to be directed against Marmont and the Army of Portugal: ‘his corps is the pivot on which everything turns: he is the counterpoise which can be thrown into the balance in one scale or the other, according as our forces have to act on the one side or the other[683].’ Drouet himself had at the same time received that order to the same effect, sent to him directly and not through his immediate superior, which so much scandalized Soult’s sense of hierarchical subordination[684]. On getting the Madrid dispatch of May 26 upon June 8th, Soult had written to say that Wellington’s real objective was Andalusia and not the North, that Marmont was utterly misled if he supposed that he was to be attacked by the main body of the allies, that Graham, with two British divisions, was still at Portalegre in support of Hill, and that Drouet had therefore been forbidden to lose touch with the Army of the South by passing towards the Tagus. If he departed, the whole fabric of French power in the South would go to pieces, ‘I should have to pack up and evacuate Andalusia after the smallest check.’ Drouet should ‘contain’ Hill, but could do no more. In a supplementary dispatch of June 12, provoked by the receipt of Joseph’s direct orders to Drouet, Soult went further, definitely stating that the troops in Estremadura should not go to the Tagus, ‘where they would be lost to the Army of the South, but would never arrive in time to help the Army of Portugal.’ If Drouet passed the Tagus, Hill would march on Seville, and on the sixth day would capture that insufficiently garrisoned capital, put himself in communication with Ballasteros, and raise the siege of Cadiz. ‘I repeat that the Army of the South cannot carry out its orders, and send Count D’Erlon and 15,000 men to the valley of the Tagus, without being compelled to evacuate Andalusia within the fortnight.... If your Majesty insists, remove me from command, I do not wish to be responsible for the inevitable disaster that must follow[685].’

At the same time Drouet, much vexed at having personal responsibility thrown upon his shoulders, by the King’s direct orders to him to march without consulting Soult, wrote to Madrid that he was very weak, that Hill was in front of him with a superior force, and that Barrois and Pierre Soult, who had just joined him, were under strict orders not to go beyond the Guadiana, so that if he himself marched towards the Tagus it would be with a very small force. But he dare not make that move: ‘I am absolutely obliged to stop where I am [Villafranca] in presence of Hill, who still remains concentrated on the Albuera position, which he has entrenched, with at least 25,000 men.’ Indeed an attack by Hill was expected day by day: ‘at the moment of writing there is lively skirmishing going on at the outposts, and news has come in that the whole allied army is advancing[686].’ Drouet, in short, was determined to evade responsibility, and summed up the situation by the conclusion that he was acting for the best in ‘containing’ Hill and his very large detachment, who could be of no use to Wellington in the campaign which the latter was now reported to have begun against Marmont in the North. He could do no more.

The deadlock in front of the Albuera position lasted for many days—from June 21st till July 2nd. This was a very trying time for Hill’s corps—the weather was excessively hot, the ground was hideous with the insufficiently buried corpses of the battle of last year, and sickness was very prevalent in some regiments. For the first day or two after the arrival of the French at Villafranca and Almendralejo, an attack was expected each morning, but nothing in particular happened. Drouet kept quiet behind his cavalry screen, and did no more than send foraging parties out on his flanks, which ravaged the countryside as far as Merida and Feria. Over-valuing Hill’s strength, he dreaded to commit himself to an attack on a superior force, covered by field-works and in a fine position. Nothing was seen of him for ten days, save that on the 26th he felt the posts of the allies at Corte de Peleas and Santa Marta, and retired after a little cavalry skirmishing. On July 1, however, he executed a more searching reconnaissance, with three brigades of cavalry under the direction of Pierre Soult, Vinot’s in the centre, Sparre’s on the right, Lallemand’s on the left. Barrois’s infantry division came up in support. Vinot drove in a Portuguese cavalry regiment of J. Campbell’s brigade from Corte de Peleas[687], but retired when he found it supported by Long’s light dragoons in front of the Albuera position. Lallemand found Santa Marta held by Penne Villemur’s cavalry, and turned them out of it with considerable loss, for the Spanish general unwisely offered battle, and was routed after a very short contest. He retired into the wood of Albuera, whose edge was occupied by Slade’s heavy dragoons, supported by the pickets of Byng’s infantry brigade. A troop of the 3rd Dragoon Guards made a gallant charge to cover the retreat of the Spaniards, and suffered some loss in bringing them off. Lallemand at dusk pressed forward, and cut off a small party of the Buffs, who would have been taken prisoners if a troop of the 2nd Hussars K.G.L. had not rescued them by a sudden counter-attack. Sparre’s brigade on the right did no more than skirmish with the allied outposts along the lower course of the river Albuera. At night all the French cavalry retired, and D’Erlon wrote to Soult that his reconnaissance had ‘completely fulfilled its object,’ by making him certain that Hill had 25,000 foot, 3,000 horse, and a very strong force of artillery in position, so that it would be insane to attack him[688].

On the next morning, July 2nd, Hill determined to make use of Wellington’s permission to bring on an action, if he should judge that Drouet was not strong enough to face him. The weakness of the French demonstration had convinced him that the enemy was not ready to fight. Collecting the whole of his army, he advanced from the Albuera position towards Santa Marta, thus challenging Drouet to a fight. The enemy’s vedettes made no stand and retired when pushed. On reaching Santa Marta Hill halted for the night in battle order, and on the morning of the 3rd resumed his movement, which was directed to cutting off Drouet from the great road to Seville. While Erskine with the light cavalry (Long, and J. Campbell’s Portuguese) advanced down the high-road to Villalba, supported by one British and one Portuguese brigade of infantry, Hill himself, with the rest of his army, executed a flank march to Feria, and, having got behind the French left wing, turned inward and moved toward Los Santos. The enemy’s main body, at Villafranca and Fuente del Maestre, were thus prevented from using the high-road to Seville, and placed in a position which compelled Drouet either to fight, or to retire south-eastward towards Usagre and Llerena.

Next morning (July 4) Hill expected a battle, for Barrois’s division and all Pierre Soult’s cavalry were found in a strong position at Fuente del Maestre, and the rest of the French were close behind at Almendralejo. But when he continued his movement toward the right, outflanking Barrois instead of attacking him, the enemy gave way and retired, protected by his cavalry, retreating on Ribera, Hinojosa, and Usagre[689]. There was lively skirmishing between the squadrons of the British advanced guard, and those of the French rearguard, but no serious engagement.

The same general plan of action continued on the 5th. Hill, keeping his army well concentrated, moved in two columns on Usagre and Bienvenida, the bulk of his cavalry riding at the head of his left-hand column and pressing in the French horse. Drouet took up a position at Valencia de las Torres, where he had found strong ground, and thought on the 6th that he would risk a defensive action. But Hill, instead of marching in upon him, continued his flanking movement towards Llerena. Thereupon Drouet, finding that he would be cut off from Andalusia if he remained in his chosen position, evacuated it and fell back by Maguilla on Berlanga and Azuaga [July 7]. The two armies had thus got back into exactly the same positions in which they had lain on June 19th, before Hill’s retreat to Albuera. The tale of their manœuvres bears a curious resemblance to the contemporary movements of Wellington and Marmont between Salamanca and Tordesillas. In each case one combatant, when pressed, retired, and took up a strong position (Marmont at Tordesillas-Pollos-Toro, Hill at Albuera). He then issued from it after some days, and by persistent flank movements dislodged his opponent, and drove him back to the same position from which he had started, so that the situation came back to that which it had been three weeks before. But here the parallel ended—Marmont pressed his advantage too far, and got entangled in the disastrous manœuvre of July 22, which brought on the battle of Salamanca and his own ruin. Hill, contented with what he had achieved, halted at Llerena, and did not push matters to a decisive action. He had done all that Wellington desired in keeping Soult’s attention diverted from Marmont’s peril, and in ‘containing’ a hostile force as great as his own. Moreover he had driven it off the road to Seville, and if it retreated on Andalusia it would have to be on Cordova, by the road of Constantina, since no other remained available.

But a new development of this complicated and indecisive campaign began on July 10th. Drouet, thinking apparently that Hill’s farther advance might be stopped as effectively by assuming a position on his flank as by direct opposition in front, shifted his right wing (Daricau’s division and Sparre’s and Vinot’s cavalry) back to Zalamea and its neighbourhood, where Daricau had lain in May and June. He himself resumed his old head-quarters at Fuente Ovejuna. Now just at this time Hill received an intercepted letter of King Joseph to Drouet, dated June 21st, which repeated in angry terms the long-ignored orders that the Estremaduran detachment of the Army of the South was to march on Toledo without delay. ‘Vous aurez sans doute reçu les renforts que j’ai donné l’ordre au duc de Dalmatie de vous envoyer. Vous devez avoir quinze mille hommes. Agissez avec ce corps, et tout ce qui est sous le commandement du général Daricau. Rapprochez-vous de moi: passez le Tage, et mettez-vous en état d’agir suivant les événements; n’attendez aucun ordre[690].’

The capture of this dispatch coincided with the news that Drouet had pushed Daricau and a large body of cavalry towards Zalamea. Hill drew the natural deduction that the French opposite him were at last about to obey the King’s orders, and to march to the Tagus, via Zalamea, Medellin, and Truxillo. ‘The intelligence that I have of the enemy’s movements’ (he wrote to Wellington) ‘indicates his intention of carrying Joseph’s instructions into execution.... I have received information [false as it chanced] that Drouet was yesterday at Zalamea, with his main body, having sent troops by Berlanga and Azuaga. I shall move immediately in the direction of Zalamea.’ That is to say that if Drouet was going off northward towards the King, Hill was prepared to carry out the original instructions which Wellington had left him, and if he could not stop the enemy, would move parallel to him, so as to join his chief before Drouet could transfer himself to the northern sphere of operations. His route would be by Badajoz or Merida and the newly-restored bridge of Alcantara on Ciudad Rodrigo, a much shorter one than that of his opponent. He had just begun to move his left wing in the direction of Merida, when he received a letter from Wellington exactly conforming to his own ideas. If Drouet is making for the Tagus in full force, wrote Wellington, you must take all the cavalry except one English regiment and Campbell’s Portuguese, along with Byng’s and Howard’s brigades of the 2nd Division, and Hamilton’s division, and send orders to have all preparations made at Alcantara to lay down the bridge: your route across the mountains will be by the pass of Perales: you will find elaborate instructions for the further movement at Ciudad Rodrigo. If Drouet only takes a small force, more allied troops may be left in Estremadura; Zafra had better be their head-quarters. Hill would conduct the marching column as far as Perales, and then return to take charge of whatever is left in the South to watch Soult[691].

A few days later it became evident that no general movement of the French towards the Tagus was in progress. Daricau’s infantry and the attached cavalry settled down at and about Zalamea, and pushed nothing but reconnaissances in the direction of the Guadiana—parties of horse appeared about Don Benito and Medellin, but no solid columns in support[692]. Hill therefore halted, with his head-quarters at Zafra and his rearguard (which had but a moment before been his advanced guard) at Llerena: only a few of J. Campbell’s Portuguese squadrons moved to Merida, though some Spanish infantry came up to the same direction[693]. Things then remained very quiet till July 24th, when Drouet at last appeared to be on the move with some definite purpose. On that day Lallemand’s dragoons appeared at Hinojosa, pressed in a Portuguese cavalry regiment, and seemed inclined to push towards Ribera, but retired when Long’s brigade came up against them: the losses on both sides were trifling. Three days later (July 27) a brigade of Daricau’s infantry advanced to Medellin and drove off the observing force of the Spanish infantry, while Vinot’s cavalry executed a raid on Merida, expelled the Portuguese detachment there, and exacted a requisition of food from the town. They then retired in haste; but Hill thought it well for the future to strengthen his left, and moved up Byng’s British and A. Campbell’s Portuguese infantry brigades to Merida. But Drouet was only feinting, and had no serious intentions of drawing up to the Guadiana, or crossing that river northward. His main purpose was simply the raising of requisitions; for his detachments in the mountains of the Serena were living on the edge of famine, and could only feed themselves by keeping constantly on the move. It is curious to find from the dispatches of the two opposing generals at this time that both were fairly satisfied with themselves: each thought that he was ‘containing’ a somewhat superior force of the enemy, and was doing his duty by keeping it from interfering in the more important theatre of war. Hill knew that he was detaining Drouet, when he was much wanted at Madrid: Drouet knew that he was preventing Hill from joining Wellington on the Douro. But the real balance of advantage was on the side of the allies: Hill, with only 8,000 British and 11,000 Portuguese was claiming the attention of three veteran divisions of the infantry of the Army of the South, and of the major part of Soult’s cavalry. The French in Andalusia were left so weak by the absence of 18,000 men beyond the Sierra Morena, that they could neither molest Cadiz nor the Army of Murcia. Indeed, Ballasteros, though his forces were less than they had been at the time of his defeat at Bornos, was able to provide employment for all the troops that Soult could spare for operations in the open field.

Six weeks after his disaster of June 1st, that enterprising, if irresponsible, general started out again from the lines of San Roque with between 5,000 and 6,000 men. Keeping to mountain roads and concealing his march, he surprised, on July 14, the great harbour-city of Malaga, though he failed to capture its citadel, Gibalfaro, into which the wrecks of the garrison escaped. Ballasteros got money, stores, and recruits from the captured town, but knew that he dare not tarry there for long. For Soult, naturally enraged at such a bold and successful raid, turned troops toward him from all sides. Leval, the governor of Granada, marched against him with every spare battalion that could be got together from the eastern side of Andalusia, some 5,000 bayonets. Villatte, in command before Cadiz, came from the other quarter with 6,000 men; they had orders to catch Ballasteros between them, to intercept his retreat upon Gibraltar, and annihilate him.

In order to cut off the Spaniard from his usual place of refuge, Villatte took a turn to the south, appeared in sight of Gibraltar on July 20, and then, keeping himself between the British fortress and Ballasteros, advanced northward to wait for him. Leval was to have driven him into Villatte’s arms, advancing from Antequera and pressing the hunt southward. But the raider, instead of retreating in the expected direction, slipped unseen across Villatte’s front by Alora, and made off into the plains of central Andalusia. On the 25th at dawn he appeared, most unexpectedly, at Osuna and surprised the small French garrison there. The governor, Colonel Beauvais, cut his way through the streets to a fortified convent, where he held out. But Ballasteros, satisfied with having captured a quantity of stores, mules, and baggage, and a few prisoners, vanished. Leval was on his track, and he had to evade his pursuer by a flank march, first to Grazalema and then to Ubrique. This was bringing him dangerously near to Villatte’s position. But that general had no accurate knowledge of what was going on to the north, and having waited for ten days in the mountains beyond Gibraltar for a prey that never appeared, found himself starved out. On the 30th he started on his enforced return towards the Cadiz Lines, and had reached Medina Sidonia when Ballasteros, who had quite outmarched Leval, came down in safety to Ximena on August 1, and placed himself in touch with Gibraltar once more. Thereupon Leval, seeing that it was no use to push the Spaniard (for about the tenth time) under the guns of the British fortress, and finding his column utterly worn out, went home to Granada[694].

Thus Ballasteros gave no small help to the allied cause by distracting some 11,000 or 12,000 French troops for a long fortnight, while Hill was detaining Drouet in Estremadura. By the time that the hunt after the evasive Spaniard had come to an end, the battle of Salamanca had been fought, and the aspect of affairs in the Peninsula had been completely changed. Even Soult, who had so long shut his eyes to the obvious, had at last to acknowledge that a new situation had arisen.

The news of Salamanca had reached Hill on July 29th, and caused a general expectation that the French in Estremadura would retreat at once, and that Soult would be retiring from Andalusia also in a few days. No such results followed—the intelligence was late in penetrating to the French camps; and Soult, still hoping to induce King Joseph to join him, lingered for many days in his old posture. On August 4 Hill wrote that the ‘recent glorious event’ appeared to have had very little effect on his immediate opponent, who continued in a strong position in his front. ‘Therefore for the present I shall remain where I am, and watch for a favourable opportunity of acting[695].’ Soult at Seville had, as late as August 8, no official news of Marmont’s defeat, and only knew of it by Spanish rumours, which he—of set purpose—discounted. ‘Les relations qu’ils ont publiées exagèrent sans doute les avantages: mais il paraît que quelque grand événement s’est passé en Castille[696].’ He continued to urge King Joseph to come to Seville, join him, and attack Hill with such superior forces that Wellington would be forced to fly to the aid of his subordinate. It was only on August 12th that certain information regarding the battle of July 22nd reached the head-quarters of the Duke of Dalmatia, in the form of Joseph’s Segovia dispatch of July 29th, containing the orders for the complete evacuation of Andalusia, and the march of the whole Army of the South upon Toledo. Even then Soult did not think it too late to make a final appeal to the King: ‘the loss of a battle by the Army of Portugal was nothing more than a great duel, which can be undone by another similar duel. But the loss of Andalusia and the raising of the siege of Cadiz would be events whose effects would be felt all round Europe and the New World.... What does it matter if the enemy is left in possession of the whole space between Burgos and the Sierra Morena, until the moment when great reinforcements come from France, and the Emperor has been able to make his arrangements? But this sacrifice of Andalusia once made, there is no way of remedying it. The imperial armies in Spain will have to repass the Ebro—famine perhaps will drive them still farther[697],’ &c.

On reflection, however, Soult did not venture to disobey, and, before his last appeal could possibly have reached the King’s hands, began to issue orders for evacuation. But so great was his rage that he wrote an extraordinary letter to Clarke, the Minister of War at Paris, in which he made the preposterous insinuation that Joseph was about to betray his brother the Emperor, and to come to an agreement with the Cadiz Cortes. The evidence which he cited for this strange charge was flimsy in the extreme. ‘I have read in the Cadiz newspapers the statement that His Majesty’s Ambassador in Russia has joined the Russian army: that the King has opened intrigues with the Insurrectional Government[698]. Sweden has made peace with England, and the Hereditary Prince (Bernadotte) has begun to treat with the Regency at Cadiz[699].... I draw no deduction from all these facts, but I am all the more attentive to them. I have thought it necessary to lay my fears before six generals of my army, after having made them take an oath not to reveal what I told them save to the Emperor himself, or to some one specially commissioned by him. But it is my duty to inform your Excellency that I have a fear that all the bad arrangements made [by the King] and all the intrigues that have been going on, have the object of forcing the imperial armies to retreat to the Ebro, or farther, and then of representing this event as the “last possible resource” (an expression used by the King himself in a letter of July 20), in the hope of profiting by it to come to some compromise[700].

This letter, as obscurely worded as it was malicious, was not sent to France by the usual channels, lest the King should get wind of it, but consigned to the captain of a French privateer, who was about to sail from Malaga to Marseilles. By an ill chance for Soult, the vessel was chased by a British ship, and compelled to run for shelter into the harbour of Valencia. There the King had recently arrived, on his retreat from Madrid. The privateer-captain, who did not know what he was carrying, sent the letter in to the royal head-quarters. Hence came an explosion of wrath, and a series of recriminations with which we shall have to deal in their proper place.

The evacuation of Andalusia commenced from the western end, because the retreat of the army was to be directed eastwards. The evacuation of the Castle of Niebla on August 12th was its first sign—the troops in the Condado had retired to San Lucar near Seville by the 15th. A little later the garrisons in the extreme south, at Ronda and Medina Sidonia, blew up their fortifications and retired. These were small movements, but the dismantling of the Cadiz Lines was a formidable business, and took several days. Soult covered it by ordering a furious bombardment of the city and the Puntales fort from his batteries across the bay; during each salvo of the heavy guns one or two of them were disabled, others being fired at an angle against their muzzles, so as to split them. More were burst by intentional over-loading, others had their trunnions knocked off, but a good many were only spiked or thrown into the water. The ammunition remaining after two days of reckless bombardment was blown up; the stores set on fire; the flotilla of gunboats was sunk, but so carelessly that thirty of them were afterwards raised with no difficulty and found still seaworthy. This orgy of destruction continued for the whole of the 24th: at night the sky was red all round the bay, from Rota to Chiclana, with burning huts and magazines, and the explosions were frequent.

This was the moment when the large allied force in Cadiz might well have made a general sortie, for the purpose of cutting up the enemy while he was engrossed in the work of destruction. Wellington had written a week before, to General Cooke, then in command of the British contingent in the Isla de Leon, to bid him fall upon the enemy when opportunity should offer, considering that the French troops in the Lines were reduced to a minimum by the detachment of the division that had gone out to hunt Ballasteros. He suggested that the allies should cross the Santi Petri river and attack Chiclana, taking care, however, not to be cut off from their retreat. Unfortunately this letter of August 16th came too late, for Cooke (after conferring with the Spanish authorities) had committed himself to another and a more circuitous expedition to molest the French. General Cruz Murgeon, with a Spanish division of 4,000 men (which had originally been intended for the reinforcing of Ballasteros) had landed at the port of Huelva, in the Condado de Niebla, on August 11th. Cooke reinforced him with the pick of the British contingent—six companies of Guards, half of the 2/87th[701], two companies of Rifles[702], part of the 20th Portuguese, and the squadron of the 2nd Hussars K.G.L., which was the only cavalry at his disposition. These, placed under the charge of Colonel Skerrett, made up 1,600 men in all[703]; they landed at Huelva, joined Cruz Murgeon, and advanced with him against Seville. On the 24th they discovered the French outposts at San Lucar la Mayor, and drove them out of that town. But they hesitated over the idea of attacking Seville, where French troops were collecting from all quarters, though the divisions of Conroux and Villatte from the Cadiz Lines had not yet come up.

On the night of August 26th-27th, however, Soult, apprised of the near approach of his column from the Lines, evacuated Seville with the main part of his force, escorting a vast horde of Spanish refugees, who feared to remain behind to face their countrymen, and a long train of wagons and carriages loaded with the accumulated spoils of three years of tyrannous misrule in Andalusia. He left a rearguard to occupy the outworks of the city, which was to be picked up and taken on by Villatte when he should appear on the next day.

On hearing of the departure of the Marshal, Cruz Murgeon and Skerrett resolved to attack Seville, knowing that the troops left behind to guard it were insufficient to man effectively all its long line of defences. Being on the western side of the Guadalquivir, they had first to win the large transpontine suburb of Triana, the home of potters and gipsies, through which alone access could be got to the city. It was attacked at several points and stormed, but the enemy then held to the great bridge over the river linking Triana and Seville, and made a long resistance there. The bridge had been barricaded, part of its planks had been pulled up, and artillery had been trained on it from the farther side. Notwithstanding these obstacles the Spaniards attacked it; the well-known Irish adventurer Colonel Downie charged three times at the head of his Estremaduran Legion. Repelled twice by the heavy fire, he reached the barricade at the third assault, and leaped his horse over the cut which the French had made in front of it, but found himself alone within the work, and was bayoneted and made prisoner[704]. But soon after the allies brought up guns through the streets of Triana, and so battered the barricade that the French were compelled to evacuate it. Skerrett sent the Guards across: they passed by the beams which had been left unbroken, and many Spanish troops followed. After a running fight in the streets of the city, in which some of the inhabitants took part, the garrison was completely driven out, and fled by the Carmona Gate towards Alcala. The victors captured two field-pieces, about 200 prisoners[705], and a rich convoy of plunder, which was to have been escorted by the French rearguard[706]. Villatte’s column, approaching the city in its march from the Cadiz Lines and Xeres, found it in the hands of the allies, so swerved off eastward and followed Soult, picking up the expelled garrison by the way.

Cruz Murgeon and Skerrett did not pursue, not thinking themselves strong enough to meddle with the French, but only sent their cavalry forward to watch their retreat. They stayed in Seville, where the Cadiz Constitution was proclaimed with great enthusiasm on August 29th. On the other flank of the French Ballasteros was trying at this moment to molest the column formed by the garrisons retiring from Ronda, Malaga, and Antequera on Granada. He followed them for ten days, and fought their rearguard at Antequera on September 3rd, and at Loja on September 5th; but though he captured many stragglers and some baggage, as also three guns, he was unable to do any material harm to the main body, which General Sémélé brought in to join Leval at Granada on September 6th.

Soult, meanwhile, with the troops from Cadiz and Seville, had to halt at Cordova for some days, to allow of the junction of Drouet from Estremadura; for that general had to collect his troops and to bring down detachments from places so far away as Don Benito and Zalamea, before he could concentrate and march across the Sierra Morena to join his chief. Drouet had kept up a bold countenance in front of Hill to the last moment, even after he had received orders from Soult to prepare for a sudden retreat. Indeed one of the most lively of the many cavalry affairs fought in Estremadura during the summer of 1812 took place in August. On the 1st of that month, when Hill was already expecting that the news of Salamanca would have driven his opponent away, Pierre Soult tried a raid upon Ribera, with two regiments of cavalry and two battalions, and drove in the 2nd Hussars of the Legion, who maintained a long and gallant skirmishing fight, till General Erskine came up with Long’s brigade, when the French retreated. Erskine was thought to have missed a fine opportunity of cutting up the raiding detachment by his slow and tentative pursuit[707]. On the 18th Soult made another reconnaissance in force, with four regiments, in the same direction, on a false report that Hill had moved from Ribera and Almendralejo. This brought on another long day of bickering, with no definite result: it was mainly remembered afterwards for the courteous behaviour of Drouet in sending back unharmed Erskine’s aide-de-camp Strenowitz, the most daring officer for raids and reconnaissance work in the German Legion. He had been captured while scouting, and a general fear prevailed that he would be shot, for he had served for a short time in the French army, and might have been treated as a deserter. Drouet most handsomely dispatched him to the British camp on parole, with a request that he might be exchanged for an officer of his own, who had been taken a few days before. ‘A most courteous and liberal enemy!’ wrote a diarist in Hill’s camp, ‘Strenowitz’s exploits are well known: certainly in strict law he might have been hung[708].’

It was not till August 26th that all the French troops in front of Hill suddenly vanished, Drouet having had orders to keep his position till Seville was ready to be evacuated; for Soult feared that if he withdrew his forces in Estremadura too early, in the direction of Cordova, the allied troops might make a forced march on Seville, and arrive there before the divisions from the Cadiz Lines had gone by. Wherefore Drouet was in evidence before Hill till the precise day when Soult left Seville. He then retired through the Sierra Morena, going by the remote mountain road by Belalcazar with such speed that he reached Cordova on the fourth day (August 30). He was not pursued by Hill, whose orders from Wellington were to come up to the Tagus and join the main army, and not to involve himself in operations in Andalusia. Only some of Penne Villemur’s Spanish horse, under the German colonel Schepeler—one of the best historians of the war—followed on Drouet’s track, and saw him join Soult at Cordova[709]. The united French force then marched on Granada, where the garrisons of eastern Andalusia, under Leval, had concentrated to meet the Marshal. Up to this moment Soult had been uncertain whether he should retreat by way of La Mancha, or across the kingdom of Murcia. His decision was settled for him by news brought by Drouet, who had heard in Estremadura of King Joseph’s evacuation of Madrid and Toledo. Since the Army of the Centre was now known to be on the road for Valencia, to join Suchet, it would be too dangerous to cross La Mancha in search of it. Wellington might descend from Madrid in force, upon an enemy who dared to march across his front. Wherefore Soult resolved that his retreat must be made across the kingdom of Murcia. It was true that O’Donnell’s army was in occupation of the inland in that direction, but it was weak and disorganized. Moreover, Suchet had lately inflicted a severe defeat upon it at Castalla (on July 21st), and O’Donnell was practically a negligible quantity in the problem. A far more important factor in determining Soult’s exact route was the news that the yellow fever had broken out at Cartagena and was spreading inland: it had reached the city of Murcia. Wherefore the French army avoided the coast, and took the inferior roads across the northern part of the province.

Soult, when once he had concentrated 45,000 men at Granada, had nothing to fear from any enemy. The gloomy picture of ‘a retreat harrassed by 60,000 foes,’ with which he had tried to scare King Joseph a month before, turned out to be a work of pure imagination. Hill had turned off towards the Tagus: Cruz Murgeon and Skerrett remained at Seville, awaiting the appearance of the 10,000 men left in Cadiz. But these were slow to move, because they had been on garrison duty for long years, and had to provide themselves with transport. Only Ballasteros hung about Granada, bickering with the outposts of the French army, and as he had no more than 5,000 or 6,000 men he was not dangerous, but only tiresome.

Soult therefore was able to spend many days at Granada, making deliberate preparations for the toilsome march that was before him. He only started out, after destroying the fortifications of the Alhambra and other posts, on September 16th. His route was by Baza, Huescar, Caravaca, and Hellin, through a mountainous and thinly-peopled country, where his troops suffered considerable privations. But these were nothing compared to the misery of the immense convoy of Afrancesados of all ages and both sexes, who had joined themselves to his train, and had to be brought through to a place of safety. Nor did the 6,000 sick and wounded whom he was dragging with him enjoy a pleasant journey. Yet it was only the September heat and the mountain roads that harassed the army and its train: Ballasteros did not pursue farther than the borders of Andalusia: the Murcians were cowed by the approach of a force which could have destroyed them with ease if it had lingered within their borders. Some of them shifted north toward Madrid, others south toward Alicante: none did anything to attract the notice of such a formidable enemy. Touch with Suchet’s outposts was secured before September was quite ended, and by the appearance of the whole Army of Andalusia near Valencia, a new military situation was produced by October 1st. With this we shall have to deal in its proper place—the fortunes of Wellington and the main army of the allies have not been followed beyond the middle of August.

Summing up the events of June-July-August 1812 in southern Spain, it is impossible to avoid the conclusion that Soult’s personal interests wrecked any chance that the French might have had of retaining their dominant position in the Peninsula, when once Wellington had committed himself to his offensive campaign upon the Tormes and the Douro. If the Duke of Dalmatia had obeyed in June King Joseph’s peremptory orders to send Drouet to Toledo, he would have had, no doubt, to evacuate certain parts of Andalusia. But Joseph and Jourdan could have marched many weeks earlier, and with a doubled force, to interfere with Wellington’s campaign against Marmont. It is true that Hill would have made a corresponding movement by Alcantara, and would have joined the main allied army under his chief many days before the King and Drouet would have been able to link up with Marmont. But Hill, on leaving Estremadura, would have removed the larger and more efficient part of his corps from Soult’s vicinity, and the Marshal might easily have held Seville and the Cadiz Lines, when faced by no stronger enemies than Ballasteros and the garrison in the Isla. If Soult had made up his mind to sacrifice Andalusia, and had marched with his whole army on Toledo, in June or even early in July, Wellington’s whole game would have been wrecked. It was, perhaps, too much to expect that the Marshal would consent to such a disinterested policy. But if, without making this sacrifice, he had merely obeyed King Joseph, and reinforced the Army of the Centre at an early date, he would have made the Salamanca campaign impossible. Wellington would probably have retired behind the Agueda and abandoned his conquests in Leon, without risking a battle, if the French forces in contact with him had been 25,000 men stronger than they actually were. The junction of Hill and some 12,000 men of the best of his Estremaduran detachment would have given him the power to fight out a defensive campaign on the Portuguese frontier, but hardly to deliver an offensive battle like Salamanca. The net results of all his manœuvres in June would then have been no more than an indirect success—the delivery of eastern Andalusia from Soult. Seville and the Cadiz Lines might still have remained occupied by the French.

It is scarcely necessary to repeat once more that Soult’s counter-plan of inviting the King and the Army of the Centre to retire to Andalusia, throwing up all communication with France and the imperial armies beyond the Douro, was wrongheaded in the extreme, though Napier calls it ‘grand and vigorous[710].’ Joseph could have brought no more than the 15,000 men that he owned, and they, when added to the 50,000 men of the Army of the South, would not have provided a force large enough to make a decisive move. For, as we have already seen, half the French in Andalusia were necessarily pinned down to garrison duties, and the ‘containing’ of Ballasteros and other partisans. Soult could never bring more than 25,000 men of his own into Estremadura: if 15,000 more are added for King Joseph’s troops[711], only 40,000 in all would have been available for a demonstration (or a serious invasion) in the direction of Portugal. Such a force would have given Wellington no very great alarm. It would have had to begin by besieging Badajoz and Elvas, in face of the existing ‘containing’ army under Hill, a delicate business, and one that would have taken time. Meanwhile Wellington could have come down, with reinforcements strong enough to make up a total sufficient to fight and beat 40,000 men, since he had the advantage of a central position and the shorter roads. At the worst he would have blocked the French advance by taking up an unassailable position, as he had before on the Caya in June 1811. But now he would have had a far superior game in his hands, since Badajoz was his and not his enemy’s, and his total disposable force was considerably larger than it had been in 1811.

Thus, if Soult’s plan had been carried out, all central Spain, including the capital, would have been lost just as much as it was by the actual campaign of July-August 1812, and the disorganized Army of Portugal could have done nothing. For Wellington could have left not Clinton’s one division (as he actually did) but three at least to look after it—not to speak of the Galicians and the partidas. Isolated and cut off from all communication with other French armies, Soult and the King would have had to evacuate Andalusia in the end, if they did not suffer a worse fate—a crushing defeat in a position from which there would have been no retreat possible. Hypothetical reconstructions of campaigns which might have happened are proverbially futile—but it is hard to see how any final profit to the French could have come from Soult’s extraordinary plan.