CHAPTER XI

THE TWO DIVERSIONS: (1) OPERATIONS IN THE NORTH: SIR HOME POPHAM AND CAFFARELLI. (2) OPERATIONS IN THE EAST: SUCHET, O’DONNELL, AND MAITLAND. JUNE-AUGUST 1812

It has already been made clear that the whole of Wellington’s victorious advance, from Ciudad Rodrigo to Madrid, was rendered possible by the fact that he had only to deal with the Army of Portugal, succoured when it was too late by the Army of the Centre. If Caffarelli and his 35,000 men of the Army of the North had been able to spare any help for Marmont, beyond the single cavalry brigade of Chauvel, matters must have taken a very different turn from the first, and the Douro (if not the Tormes) must have been the limit of the activity of the Anglo-Portuguese army. How Caffarelli was to be detained, according to Wellington’s plan, has been explained in an earlier chapter[712]. The working out of the scheme must now be described.

[712] See above, pp. 340-1.

The essential duty of the French Army of the North was twofold, according to Napoleon’s general conception of the Spanish war. It was Marmont’s reserve, bound to assist him in time of trouble; but it was also the force of occupation for the Biscayan provinces, Navarre, Santander, and Burgos. Of its 35,000 men more than half were at all times immobilized in the innumerable garrisons which protected the high-road from Bayonne to Burgos, and the small harbours of the coast, from San Sebastian to Santoña. The system of posts was complicated and interdependent. Since the great guerrillero Mina started on his busy career in 1810, it had been necessary that there should be fortified places at short intervals, in which convoys moving to or from France (whether by Vittoria and San Sebastian, or by Pampeluna and Roncesvalles) could take refuge when attacked by the bands. And since a convoy, when it had sought shelter in one of the minor garrisons, might be blockaded there indefinitely, unless the high-road were cleared betimes, large movable columns had to be ready in three or four of the larger places. Their duty was to march out on the first alarm, and sweep the guerrilleros away from any post that they might have beset. Such bodies were to be found at Bayonne, where the ‘Reserve of the Army of Spain’ kept a brigade of 3,500 men under General D’Aussenac to watch for any exceptional outbreak of trouble in Guipuzcoa; at Pampeluna, where General Abbé had the head-quarters of his division; and at Vittoria, where the General-in-Chief, Caffarelli himself, normally lay with the brigade of Dumoustier--the last unit of the Imperial Guard still remaining in Spain--ready to keep the line of the Ebro under surveillance, and to communicate when necessary with the three large outlying garrisons of Santoña, Santander, and Burgos. Each of these last consisted of some 1,300 or 1,500 men[713]; even so, they were only strong enough to provide for their own safety in normal times, and might require assistance from head-quarters in face of any specially large and threatening combination of the insurgents. But the larger half of Caffarelli’s army was locked up in small towns, forts, and blockhouses, in bodies ranging down from a battalion to half a company. Every one of the dozen little ports on the Biscay coast had to be held, in order to prevent the bands of the inland from communicating with the English cruisers, which occasionally appeared in the offing to throw weapons and ammunition ashore. And similarly all the little towns along the Ebro had to be garrisoned, in order to keep touch with Reille at Saragossa; wherever there was a gap Mina’s Navarrese slipped in between.

[713] In and about Santander, 2 battalions of the 130th, 2 squadrons of gendarmes, &c. In Santoña, 93 officers and 1,382 men of the 28th, 75th, and 34th. In Burgos, 2 battalions of 34th Line.

Since the autumn of 1810, when Porlier and Renovales had made their vain attempt, with British naval aid, to break up the line of communication along the coast,[714] there had been no general attempt to shake the French occupation of Biscay and Cantabria. The Spanish resources were at a low ebb whenever Bonnet held the Asturias, and (as we have seen) he was generally in possession of that province, or at least of its capital and its chief harbours, from 1810 down to the summer of 1812. The idea of attacking the harbour-fortresses of the northern coast with a considerable naval force, which should get into regular touch with the patriot forces of the inland, and establish posts to be held permanently on suitable points of the northern littoral, had been started by Sir Home Popham, approved by Sir Howard Douglas, the British Commissioner in Galicia, and warmly adopted by Wellington himself, who at once realized the pressure which such a policy would bring upon Caffarelli. He counted upon this diversion as one of his most valuable assets, when he drew up his scheme for the invasion of Leon in May 1812. It more than fulfilled his expectation.

[714] See vol. iii. pp. 486-7.

On June 17th, four days after the Anglo-Portuguese army crossed the Agueda, Sir Home Popham sailed from Corunna with two line of battleships[715], five frigates[716], two sloops[717], and one or two smaller vessels, carrying two battalions of marines, and several thousand stand of small-arms for the insurgents. Popham had credentials from Castaños, as captain-general of Galicia, for Mendizabal, the officer who was supposed to exercise authority over all the bands of Cantabria and Biscay. These scattered forces consisted in the more or less organized brigades of Porlier in the Eastern Asturias, and Longa in Cantabria--both of which were reckoned part of the national army--and in addition of the guerrilleros of Jauregui [’El Pastor’] in Guipuzcoa, Renovales in Biscay, Marquinez, Saornil, the Curé Merino, and others in the mountains between the Douro and the sea. These were bands of varying strength, often scattered by the French, but always reassembled after a space, who roamed from region to region according as the enemy was stronger or weaker at one point or another. Occasionally Mendizabal was in touch with Mina and the Navarrese, but generally the French were in too great force about the high-road from Burgos to Pampeluna to make co-operation practicable.

[715] The _Venerable_ (his flag-ship) and the _Magnificent_. The _Magnificent_ went home with prisoners some weeks later, and was replaced for a time by the _Abercrombie_, from the Brest blockading squadron.

[716] _Medusa_, _Isis_, _Diadem_, _Surveillante_, and _Rhin_. The _Belle Poule_ looked in for a short time later in the season.

[717] _Sparrow_ and _Lyra_.

At the moment of Popham’s start matters were exceptionally favourable along the Biscay coast, because Bonnet was just evacuating the Asturias, in order to join his chief Marmont in the plains of Leon. His departure isolated Santoña and Santander, which had been the links by which he was joined to Caffarelli’s army. It also gave Porlier and Longa an open communication with Galicia, from which they had hitherto been cut off by Bonnet’s presence in Asturias, and a safe retreat thitherward if they should be pressed. In addition Marmont had called up all the small garrisons and detached columns from his rear, for the main struggle with Wellington; so that the Upper Douro valley and the Soria country were much more free from the French than they had been for a long time. The opportunity for molesting Caffarelli and his much-scattered Army of the North was unique.

The idea which lay at the back of Popham’s plan was that a fleet furnished with the heaviest ship guns, and with a landing-force of over 1,000 men, could operate at its choice against any one of the long chain of posts which the French held, calling in to its aid the local bands in each case. The insurgents had never been able to capture any of these places because they lacked a battering-train. The fleet supplied this want, and with few exceptions the French strongholds were not suited for resistance against heavy guns. They were mediaeval castles, fortified convents, or the patched-up walls of little towns, all defensible against irregular bands without cannon, but most vulnerable to 18-or 24-pounders. The number of the French garrisons gave ample liberty of choice between one and another: individually they were generally weak--not over 300, 500, or 1,000 men. There were succouring columns, no doubt, ready to relieve them, at Bayonne and Vittoria and elsewhere. But the squadron had the power of misleading these forces to any extent--of drawing them from one remote port to another by false attacks and demonstrations, and then of attacking some third point when the enemy had been lured as far as possible from it. Here lay the beauty of naval operations--the squadron could appear to threaten any objective that it chose, could attract the enemy thither, and then could vanish, and be fifty miles away next day. The relieving column could only follow slowly over mountain roads, and would invariably be late in learning what new direction the squadron had taken. It was something like the advantage that the Danes had in their attacks on England in the ninth century: a defending army cannot guard all points of a coast-line at once against a movable landing force on shipboard. The weaker points of the scheme were, firstly, the dependence of all operations on fine weather--contrary winds could in those days delay a fleet for whole weeks on end; secondly, the want of a base nearer than Corunna, till some good and defensible haven should have been captured; and third and greatest of all, the difficulty of inducing the local chiefs to combine: they paid a very limited amount of obedience to their nominal chief, Mendizabal: they had private grievances and jealousies against each other: and each of them disliked moving far from the particular region where his men were raised, and where every inch of the mountain roads was known to him.

However, all these dangers were known and were chanced, and the game was well worth the risk. The operations began with the appearance of the squadron before Lequeitio on June 21st. Popham landed a heavy gun and some marines, and the band of El Pastor [’Don Gaspar’ as the English dispatches call him] appeared to co-operate from the inland. The defences consisted of a fort and a fortified convent: the 24-pounder breached the former, which was then stormed by the guerrilleros in a very handsome fashion[718]: its garrison was slain or captured. The gun was then brought up against the fortified convent, whose commander, the _chef de bataillon_ Gillort, surrendered without further fighting; the prisoners amounted to 290 men, a half-battalion of the 119th regiment (June 22). Popham then moved off to Bermeo and Plencia, both of which the French evacuated in haste, leaving guns unspiked and some useful stores of provisions. The British force had set the wildest rumours abroad, and Renovales, the Spanish commander in Biscay, appeared at Orduna with his bands and threatened Bilbao, the capital of the province. It was these reports which made Caffarelli suddenly break off his project for sending reinforcements to Marmont, and prepare rather to march northward when he was most wanted on the Douro[719].

[718] So says Popham in his dispatch at the Record Office: though Napier (iv. p. 246) says that the Spaniards attacked and were repulsed. But Popham must have known best! Sir Howard Douglas corroborates him, _Life_, p. 168.

[719] See above, p. 378.

Popham’s next blow was at Guetaria, a most important post, owing to its nearness to the great _chaussée_ leading to Bayonne, which passes quite close to it between Tolosa and Ernani. If it had fallen, the main road from France to Spain would have been blocked for all practical purposes. But being far to the East and near the French border, it was remote from the haunts of the guerrilleros: few of them turned up: after a few days Popham had to re-embark guns and men, and to take his departure, owing to the arrival in his neighbourhood of a strong French flying column. He then sailed off to Castro Urdiales, where he had much better luck: Longa had left the Upper Ebro with his brigade and joined him there on July 6th, by Mendizabal’s orders. Their united force drove off on the 7th a small French column which came up from Laredo to raise the siege. The governor of Castro then surrendered with some 150 men, and 20 guns on his walls fell into Popham’s hands (July 8). The place seemed so strong that the commodore resolved to keep it as a temporary base, and garrisoned it with some of his marines.

Three days later Popham appeared before Portugalete, the fortified village at the mouth of the Bilbao river, and bombarded it from the side of the sea, while Longa (who had marched parallel with the squadron along the shore), demonstrated against its rear. But a French flying column happened just to have arrived at Bilbao, and the force which marched out against the assailants was so powerful that they made off, each on his own element [July 11th.] Popham now turned his attention for a second time to the important strategical post of Guetaria; he had enlisted the support of the Guipuzcoan bands under Jauregui, and the distant Mina had promised to send a battalion to his aid from Navarre. Popham got heavy guns on shore, and began to batter the place, while Jauregui blockaded it on the land side. This move drew the attention of D’Aussenac, commanding the flying column which belonged to the Bayonne reserve: he marched with 3,000 men towards Guetaria, and drove off Jauregui, whereupon Popham had to re-embark in haste, and lost two guns which could not be got off in time and thirty men [July 19], Mina’s battalion came up a day too late to help the discomfited besiegers.

This petty disaster was in the end more favourable than harmful to Popham’s general plan, for he had succeeded in drawing all the attention of the French to the eastern end of their chain of coast-fortresses, between Santoña and San Sebastian. But now he used his power of rapid movement to attack unexpectedly their most important western stronghold. On July 22nd he appeared in front of the harbour of Santander, while (by previous arrangement) Campillo--one of Porlier’s lieutenants--invested it on the land side. Porlier himself, with his main body, was blockading at the moment the not very distant and still stronger Santoña.

There was very heavy fighting round Santander between the 22nd July and August 2. Popham landed guns on the water-girt rock of Mouro, and bombarded from it the castle at the mouth of the port: when its fire was subdued, he ran his squadron in battle order past it, and entered the harbour, receiving little damage from the other French works (July 24). The enemy then evacuated the castle, which the marines occupied: but an attempt to storm the town with the aid of Campillo’s men failed, with a rather heavy casualty list of two British captains[720] and many marines and seamen disabled (July 27th). However, Popham and Campillo held on in front of Santander, and Mendizabal came up on August 2nd to join them, bringing a captured French dispatch, which proved that the enemy intended to evacuate the place, a strong relieving column under Caffarelli himself being at hand to bring off the garrison. And this indeed happened: the General-in-Chief of the Army of the North had marched with all the disposable troops at Vittoria to save his detachment. The governor Dubreton--the same man who afterwards defended Burgos so well--broke out of the place with his 1,600 men on the night of the 2nd-3rd and joined his chief in safety: he left eighteen guns spiked in his works. Caffarelli then drew off the garrison of the neighbouring small post of Torrelavega, but threw a convoy and some reinforcements into Santoña, which he had determined to hold as long as possible. He then hastened back to Vittoria, being under the impression at the moment that Wellington was in march against him from Valladolid, in pursuit of the routed host of Clausel. But the Anglo-Portuguese main army--as will be remembered--had really followed the retreating French no farther than Valladolid, and no longer than the 30th July. Instead of finding himself involved in the affairs of the Army of Portugal, Caffarelli had soon another problem in hand.

[720] One of them, Sir George Collier, commanding the _Medusa_.

The capture of Santander by the allies was the most important event that had happened on the north coast of Spain since 1809, for it gave the squadron of Popham possession of the sole really good harbor--open to the largest ships, and safe at all times of the year--which lies between Ferrol and the French frontier. At last the Spanish ‘Seventh Army’ had a base behind it, and a free communication with England for the stores and munitions that it so much needed. It might be developed into a formidable force if so strengthened, and it lay in a position most inconvenient for the French, directly in the rear of Clausel and Caffarelli. Popham saw what might be made of Santander, and drew up for Wellington’s benefit a report on the possibilities of the harbour, in which he details, from the information given by Porlier and his staff, the state of the roads between it and Burgos, Valladolid, and other points. Six weeks before the siege of Burgos began, he wrote that by all accounts six or eight heavy guns would be required to take that fortress, and that he could manage that they should be got there--a distance of 115 miles--by ox-draught, if they were wanted[721]. But Wellington, at the moment that this useful information was being compiled, was turning away from Valladolid and Burgos toward Madrid; and when his attention was once more drawn back to Burgos, he made no use of Popham’s offers till it was too late. Of this more in its proper place.

[721] Popham’s prescience is shown by the fact that his papers relating to Burgos began to be drawn up as early as July 26. He cross-questioned not only Porlier but other Spanish officers. Their answers did not always tally with each other. See all Popham’s dispatches of the time, in the Admiralty Section at the Record Office--under the general head ‘Channel Fleet!’ They have this misleading heading because Popham was under Lord Keith, then commanding that fleet.

Having brought all his squadron into Santander, and made himself a fixed base in addition to his floating one, Popham began to concert plans for further operations with Mendizabal, whom he described as a man of ‘vacillating councils,’ and hard to screw up to any fixed resolution. The scheme which the commodore most recommended to the general was one for a general concentration of all his scattered forces against Bilbao, in which the squadron should give its best help. But he suggested as an alternative the sending of Porlier to join Longa, who had already gone south to the Upper Ebro after the failure at Portugalete on July 11th. Porlier and Longa would together be strong enough to cut the road between Burgos and Vittoria, and so divide Clausel from Caffarelli. If the two French generals combined against them, they could always escape north-westward into their usual mountain refuges.

According to Popham’s notes Mendizabal first seemed to incline to the second scheme, and then decided for the first. He even in the end ordered up Longa--then very usefully employed against Clausel’s rear about Pancorbo and Cubo--to join in the attack upon Bilbao. But Longa came late, being busy in operations that he liked better than those which his chief imposed on him. After waiting a few days for him in vain, Mendizabal marched against Bilbao by land with two battalions belonging to Porlier and one recently raised in Alava, while Popham took three Biscayan battalions belonging to Renovales on board his squadron and sailed for Lequeitio, where he put them ashore. He himself then made for Portugalete, at the mouth of the Bilbao river. The triple attack, though made with no very great total force was successful. The officer commanding in Bilbao, went out to meet Mendizabal, and in order to collect as many men as possible, drew off the garrison of Portugalete. The British squadron, arriving in front of the port, found it undefended and threw the marines ashore. Hearing of this descent in his rear the French general, then indecisively engaged with Mendizabal and Renovales, thought that he was in danger of being surrounded, and retired hastily toward Durango, abandoning Bilbao altogether [August 13].

Learning next day that they had overrated the enemy’s force, the French returned and tried to reoccupy the Biscayan capital, but were met outside by all Mendizabal’s troops, arrayed on the position of Ollorgan. An attack entirely failed to move them, and the French fell back to Durango. General Rouget, the commanding officer in the province, then drew in all his minor garrisons, and sent Caffarelli notice that all Biscay was lost, unless something could be done at once to check Mendizabal’s progress [August 14]. Indeed the situation looked most threatening, for Longa had at last come up and joined his chief with 3,000 men, and the Biscayans were taking arms on every side. A general junta of the Basque provinces was summoned by Mendizabal to meet at Bilbao, and the French had for the moment no foothold left save in San Sebastian and Guetaria. Thereupon Caffarelli, collecting every man that he could at Vittoria, marched to join Rouget. Their united forces, making some 7,000 men, attacked Bilbao on August 27th-29th, and after much confused fighting drove Mendizabal and Longa out of the place, only a fortnight after it had come into Spanish hands. The defeated troops dispersed in all directions, each section seeking the region that it had come from--Porlier’s men retired towards Cantabria, Longa’s toward the Upper Ebro. Renovales and his Biscayan battalions were caught in their retreat, and badly cut up at Dima.

While this fighting was going on around Bilbao, Popham was trying a last attack on Guetaria, with his own resources only, as nearly all the Spaniards were engaged elsewhere. He had accomplished nothing decisive when he heard of Mendizabal’s defeat, and had to reship his guns and take his departure before the victorious Caffarelli came up. He retired to Santander, and heard there that Wellington was leaving Madrid, and once more marching on Burgos. He determined to open up communications with the British army without delay, and on August 31 sent off Lieutenant Macfarlane to seek for the head of the approaching columns. That officer, skirting the flank of Clausel’s retreating host, reached Valladolid betimes, and explained to Wellington that the Santander road would be open and available for the transport of ammunition, guns, and even food, so soon as he should have driven the French past Burgos. And--as will be seen--it was so used during the unlucky siege of that fortress again and again--but not (as Popham recommended) for the bringing up of the heavy artillery that Wellington so much lacked.

By September 1st Caffarelli had patched up matters for a time on the side of Biscay, but though he had recovered Bilbao and preserved Guetaria, all the other coast-towns were out of his power save Santoña, and that important place was cut off from the nearest French garrison by a gap of some sixty miles. Even now Popham’s useful diversion had not ceased to have its effect. But its further working belongs to a later chapter.

So much for the annals of the war in northern Spain from June to August. The diversion which Wellington had planned had been brilliantly successful. A very different story must be told of the equally important scheme that he had concerted for keeping his enemies distracted on the eastern side of the Peninsula, by means of the Anglo-Sicilian expedition and the Spanish Army of Murcia.

Suchet, it will be remembered[722], had been stayed from further conquests after the fall of Valencia partly by the indirect results of Wellington’s operations on the Portuguese frontier--starting with the fall of Ciudad Rodrigo--partly by Napoleon’s action in drawing back to the Ebro the two divisions of Reille, and calling out of Spain the numerous Polish battalions serving in the Army of Aragon. But not the least of the hindering causes was a purely personal one--the long illness which kept Suchet confined to his bed for ten weeks in February, March, and April. By the time that he was in the saddle again a notable change had come over the aspect of the war all over the Peninsula. During his sickness his lieutenants, Habert and Harispe, maintained their position in front of the Xucar river, and observed the wrecks of the Valencian and Murcian divisions that had escaped from Blake’s disaster in January. The whole force remaining under Suchet--excluding the troops left behind in Catalonia and Aragon--was not above 15,000 men, and of these nearly 4,000 were locked up in garrisons, at Valencia, Saguntum, Peniscola, Morella, and other places. It is not surprising, therefore, that no farther advance was made against the Spaniards. Joseph O’Donnell, the successor of the unlucky Mahy, was able during the spring to reorganize some 12,000 men on the _cadres_ of his old battalions. In addition he had Roche’s reserve at Alicante, 4,000 strong, which had now been profiting for many months by the British subsidy and training, and was reckoned a solid corps. He had also Bassecourt’s few battalions in the inland--the troops that D’Armagnac had hunted in December and January in the district about Requeña[723]. Cartagena, the only fortress on the coast still in Spanish hands save Alicante, had been strengthened by the arrival of a British detachment[724]. Altogether there were some 20,000 enemies facing Suchet in April, and he regarded it as impossible to think of attacking Alicante, since he had not nearly enough men in hand to besiege a place of considerable size, and at the same time to provide a sufficient covering army against Joseph O’Donnell. So little was the Murcian army molested that General Freire, O’Donnell’s second-in-command, ignoring Suchet altogether, took advantage of Soult’s absence in Estremadura, at the time of the fall of Badajoz, to alarm eastern Andalusia. He occupied Baza on April 18th, and when driven away after a time by Leval, governor of Granada, turned instead against the coast-land of the South. On May 11th an expedition, aided by English war-ships from Alicante, landed near Almeria, and cleared out all the French garrisons from the small towns and shore batteries as far west as Almunecar. Already before this (on May 1-3) an English squadron had made a descent on Malaga, seized and destroyed the harbour-works, and carried off some privateers and merchant vessels from the port. But naught could be accomplished against the citadel of Gibalfaro. Soult did little or nothing to resent these insults, because he was at the time obsessed with his ever-recurring idea that Wellington was about to invade Andalusia, and his attention was entirely taken up with the movements of Hill and Ballasteros in the West and North, so that the East was neglected. Leval at Granada had a troublesome time, but was in no real danger, since Freire’s raids were executed with a trifling force.

[722] See above, p. 86.

[723] See above, p. 56.

[724] The 2/67th and a part of the foreign Regiment of de Watteville, also a British battery, from Cadiz.

Suchet was occupied at this time more with civil than with military affairs: for some time after his convalescence he was engaged in rearranging the administration of the kingdom of Valencia, and in raising the enormous war-contribution which Napoleon had directed him to exact--200,000,000 reals, or £2,800,000--in addition to the ordinary taxes. The Marshal in his _Mémoires_ gives a most self-laudatory account of his rule; according to his rose-coloured narrative[725], the imposts were raised with wisdom and benevolence, the population became contented and even loyal, the roads were safe, and material prosperity commenced at once to revive. Napier has reproduced most of Suchet’s testimonials to his own wisdom and integrity, without any hint that the Spanish version of the story is different. The Marshal who drove the civil population of Lerida under the fire of the cannon[726], and who signalized his entry into Valencia by wholesale executions of combatants and non-combatants[727], was not the benevolent being of his own legend. Since that legend has been republished in many a later volume, it may be well to give as a fair balance the version of an enemy--not of a Spaniard, but of a Prussian, that Colonel Schepeler whose authority on the war of Valencia we have so often had occasion to quote.

[725] See _Mémoires_, ii. pp. 283-99.

[726] See vol. iii. p. 307.

[727] See above, p. 75.

‘Napoleon Bonaparte looked upon Valencia as the prey of France, and Suchet did not fall behind in his oppressive high-handedness. The long-desired goal, the wealthy city, now lay open to their rapacity, and the riches that the clergy had denied to the needs of the nation went to fill the plunder-bag of the conqueror. The miraculous statue of Our Lady of Pity was stripped of her ancient jewelled robe: only a light mantle now draped her, and showed the cut of the nineteenth century. The silver apostles of the cathedral took their way to France with many other objects of value, and the Chapter was forced to pay ransom for hidden treasures. The magnanimous marshal imposed on the new French province, as a sort of “benevolence,” six million dollars (ten had been spoken of at first), with an additional million for the city of Valencia. The churches had to buy off their bells with another 60,000 dollars. Suchet, in his moderation, contented himself with exacting 500 dollars a day for his own table and household expenses.

‘Political persecution began with a decree of March 11, which ordered the judges of the local _Audiencia_ [Law Court] to meet as before, but to administer justice in the name of Napoleon _Empereur et Roi_. The patriots refused to serve and fled; whereupon their goods were confiscated, their families were harried, and when some of them were captured they were threatened with penal servitude or death. A decree drawn up in words of cold ferocity, declared every Spaniard who continued to oppose the French to be a rebel and a brigand, and therefore condemned to capital punishment. Several villages were punished with fire and sword, because they were too patriotic to arrest and deliver up insurgents. Contrary to the promise made at the capitulation in January[728], many patriots were arrested and executed, under the pretence that they had been concerned in the murder of Frenchmen in 1808, even though they might actually have saved the lives of certain of those unfortunates at that time.

[728] See p. 73 above.

‘Valencia produces little wheat: there was much lack of it, and the French would not accept rice. Their requisitions were exacted with cruel disregard of consequences, even from the poorest, and quickly brought back to the patriotic side the mutable Valencian people, who had already been sufficiently embittered when they found that they were annexed to France. All over the province there began to appear slaughter, rebellion, and finally guerrillero bands[729].’

[729] Schepeler, pp. 609-10.

The point which Schepeler makes as to Valencia being practically annexed to France--as shown by the administration of justice in the name of Napoleon, not of King Joseph--should be noted. It illustrates Suchet’s determination to consider himself as a French viceroy, rather than as the general of one of the armies recently placed by the Emperor under the King as Commander-in-Chief in the Peninsula. We have already noted the way in which he contrived to plead special orders from Paris, exempting him from the royal control, whenever Joseph tried to borrow some of his troops for use against Wellington[730]. At the same time it must be conceded that he had a much better excuse than Soult for his persistent disobedience to such orders--his whole available force was so small, that if he had sent 6,000 men to San Clemente or Ocaña, as Joseph directed, there would have been little or nothing left in Valencia save the garrisons, and the Spaniards from Alicante and Murcia could have taken their revenge for the disasters of the past winter[731]. He represented to the King that to draw off such a body of troops to La Mancha implied the abandonment of all his recent conquests, and that if something had to be evacuated, it was better that Soult should begin the process, since Andalusia was a more outlying possession than Valencia--’les provinces du sud devaient être évacuées avant celles de l’est.’ And here he was no doubt right: as we have been remarking again and again, the only solution for the situation created by Wellington’s successes was to concentrate a great mass of troops at all costs, and the Army of the South could best provide that mass. It had 50,000 men under arms at the moment--Suchet had not in Valencia more than 15,000.

[730] See above, pp. 304-5.

[731] This is Suchet’s own view, see his _Mémoires_, ii. p. 251.

Hitherto we have spoken of those parts of the east coast of Spain which lie south of the Ebro. But if the situation in Valencia had not altered much between February and June, the same was also the case in Catalonia. Since Eroles’s victory over Bourke at Roda in March[732] there had been much marching and counter-marching in that principality, but nothing decisive. Lacy, the unpopular captain-general, was at odds with the Junta, and especially with Eroles, the best of his divisional officers, who was the most influential man in Catalonia, owing to his local connexions and his untiring energy. Lacy was a stranger, an enemy of the ‘Somaten’ system, and a pronounced Liberal. The political tendencies of the Catalans were distinctly favourable to the other or ‘Servile’ party. The captain-general was also accused of nourishing jealousy against Sarsfield, his second-in-command: and it is certain that both that officer and Eroles believed him capable of any mean trick toward them. But though divided counsels and mutual suspicions often hindered the co-operation of the commanders and the people, all were equally bitter enemies of the invader, and none of them showed any signs of slackening in their grim resolve to hold out to the end. The Catalan army did not now count more than 8,000 men in the field, but its central position in the mountains of the interior, round which the French garrisons were dispersed in a long semicircle, gave it advantages that compensated to a certain extent for its lack of numbers. It could strike out at any isolated point on the circumference, and, whether its blow failed or succeeded, generally got off before the enemy had concentrated in sufficient numbers to do it much harm. On the other hand, Decaen, now commanding in Catalonia, and Maurice Mathieu, the governor of Barcelona, though they had some three times as many men under arms as Lacy, were reduced to a position that was little more than defensive. It is true that they occasionally collected a heavy column and struck into the inland: but the enemy avoided them, and replied by counter-attacks on depleted sections of the French circle of garrisons. On April 9th, for example, 4,800 men marched from Gerona against Olot: the local levies under Claros and Rovira skirmished with them, giving ground, and finally losing the town. But though they did not stop the advance of the enemy, Milans, with a larger force, moved on the important harbour of Mataro, and laid siege to the garrison there (April 22), a stroke which soon brought the bulk of French troops back from Olot to drive him off. At the same time Sarsfield’s division pressed in upon the garrison of Tarragona, and cut off its communications with Barcelona.

[732] See above, p. 98.

This forced Decaen to march to open the road, with all the men that Maurice Mathieu could spare from Barcelona (April 28). Letting them go by, Lacy at once renewed the attack on Mataro, bringing up the forces of Sarsfield and Milans, and borrowing four ship-guns from Commodore Codrington to batter the fort, in which the French had taken refuge after evacuating the town [May 3]. Decaen and Lamarque promptly turned back, and on the third day of the siege came hastily to break it up. The Spaniards dispersed in various directions, after burying the guns, which (much to Codrington’s regret) were discovered and exhumed by the enemy. The net result of all this marching and counter-marching was that much shoe-leather had been worn out, and a few hundred men killed or wounded on each side: but certainly no progress had been made in the conquest of Catalonia. Indeed, Manso, at the end of the campaign, established himself at Molins de Rey, quite close to Barcelona, and Sarsfield occupied Montserrat, so that between them they once more cut off the communications from Barcelona southward and westward. Both had to be driven off in June, in order that the roads might again be opened.

Early in July Lacy devised a scheme which made him more hated than ever in Catalonia. He concerted with some Spanish employés in the French commissariat service a plan for blowing up the powder magazine of the great fortress of Lerida, and arranged to be outside its walls on the day fixed for the explosion, and to storm it during the confusion that would follow. Eroles and Sarsfield both protested, pointing out that a whole quarter of the city must be destroyed, with great loss of life. Lacy replied that the results would justify the sacrifice, persisted in his scheme, and moved with every available man towards Lerida to be ready on the appointed day. He miscalculated his hours, however; and, though he left hundreds of stragglers behind from over-marching, his column arrived too late. The explosion took place on the 16th, with dreadful success; not only did a hundred of the garrison perish, but a much larger number of the citizens; many houses and one of the bastions fell. The governor Henriod, a very firm-handed man whose record in Lerida was most tyrannical[733], had been entirely unaware of the approach of the Spaniards, but proved equal to the occasion. He put his garrison under arms, manned the breach, and showed such a firm front, when Lacy appeared, that the captain-general, having tired troops and no cannon, refused to attempt the storm. He went off as quickly as he had come, having caused the death of several hundred of his countrymen with no profit whatever. If he was ready to adopt such terrible means, he should at least have had his plans correctly timed. The Catalans never forgave him the useless atrocity[734].

[733] See notes in Vidal de la Blache’s _L’Evacuation d’Espagne_, 1914, which reaches me just as this goes to press, for anecdotes concerning his doings.

[734] About the same time a still more dreadful plot was said to have been formed in Barcelona, with the knowledge and approval of Lacy--arsenic was to be mixed with the flour of the garrison’s rations by secret agents. [See Suchet’s _Mémoires_, ii. p. 256, and Arteche, xii. p. 353.] How far the plan was a reality is difficult to decide. There is a large file of papers in the Paris War Office concerning experiments carried out by a commission of army-doctors, in consequence of a sudden outbreak of sickness among the troops in July. One or two soldiers died, a great number were seized with vomiting and stomach-cramps; poison being suspected, the doctors took possession of the flour, attempted to analyse it, and tried its effects on a number of street dogs. A few of the animals died: most were violently sick, but got over the dose. Poison was not definitely proved, and dirty utensils and bad baking might conceivably have been the cause of the outbreak. Some Catalan writers say that there was a poisoning-plot, or I should have doubted the whole story. See the Appendix to Arteche, xii. p. 483.

Operations in Catalonia and Valencia were thus dragging on with no great profit to one side or the other, when Wellington’s great scheme for the Anglo-Sicilian diversion on the east coast began at last to work, and--as he had expected--set a new face to affairs. Unfortunately the expedition was conducted very differently from his desire. We have already shown how, by Lord William Bentinck’s perversity, it started too late, and was far weaker than was originally intended[735]. But on July 15th General Maitland arrived at Port Mahon with a fleet carrying three English[736] and two German battalions, and parts of three foreign regiments, with a handful of cavalry, and two companies of artillery. He sent messengers across Spain to announce to Wellington his arrival, and his purpose of landing in Catalonia, as had been directed. At Majorca he picked up Whittingham’s newly-organized Balearic division, and after some delay he set sail on July 28 for Palamos, a central point on the Catalan coast, off which he arrived on the morning of July 31st with over 10,000 men on board.

[735] See above, p. 347.

[736] 1/10th, 1/81st, 1/58th, 4th and 6th Line battalions K.G.L., and parts of the foreign battalions of De Roll, Dillon, and the Calabrian Free Corps. See table in Appendix XIV. The total was 248 officers, and 6,643 rank and file.

Owing to Bentinck’s unhappy hesitation in May and June, after the expedition had been announced and the troops ordered to prepare for embarkation, French spies in Sicily had found the time to send warning to Paris, and Suchet had been advised by the Minister of War that a fleet from Palermo might appear in his neighbourhood at any moment. He received his warning in the end of June, a month before Maitland’s arrival[737], and this turned out in the end profitable to the allied cause; for, though the fleet never appeared, he was always expecting it, and used the argument that he was about to be attacked by an English force as his most effective reply to King Joseph’s constant demands for assistance in New Castile. The arrival at Alicante of transports intended to carry Roche’s division to Catalonia, and of some vessels bearing the battering-train which Wellington had sent round for Bentinck’s use, was duly reported to him: for some time he took this flotilla to be the Anglo-Sicilian squadron. Hence he was expecting all through June and July the attack which (through Bentinck’s perversity) was never delivered. The threat proved as effective as the actual descent might have been, and Wellington would have been much relieved if only he could have seen a few of Suchet’s many letters refusing to move a man to support the King[738].

[737] Clarke’s dispatch with the information was dated June 9th.

[738] Two of them dated July 22 and August 12 did ultimately fall into his hands, but only after the victory of Salamanca. See below, pp. 617-18.

Suchet’s great trouble was that he could not tell in the least whether the Sicilian expedition would land in Catalonia or in Valencia. It might come ashore anywhere between Alicante and Rosas. He prepared a small movable central reserve, with which he could march northward if the blow should fall between Valencia and Tortosa, or southward--to reinforce Habert and Harispe--if it should be struck in the South. Decaen was warned to have a strong force concentrated in central Catalonia, in case the descent came in his direction, and Suchet promised him such assistance as he could spare. On a rumour that the Sicilian fleet had turned northward--as a matter of fact it was not yet in Spanish waters--the Marshal thought it worth while to make a rapid visit to Catalonia, to concert matters with Decaen. He marched by Tortosa with a flying column, and on July 10th met Decaen at Reus. Here he learned that there were no signs of the enemy to be discovered, and after visiting Tarragona, inspecting its fortifications, and reinforcing its garrison, returned southward in a more leisurely fashion than he had gone forth.

During Suchet’s absence from his Valencian viceroyalty the captain-general of Murcia took measures which brought about one of the most needless and gratuitous disasters that ever befell the ever-unlucky army of which he was in charge. Joseph O’Donnell knew that the Sicilian expedition was due, and he had been warned that Roche’s division would be taken off to join it; he was aware that Maitland’s arrival would modify all Suchet’s arrangements, and would force him to draw troops away from his own front. He had been requested by Wellington to content himself with ‘containing’ the French force in his front, and to risk nothing. But on July 18th he marched out from his positions in front of Alicante with the design of surprising General Harispe. He knew that Suchet had gone north, and was not aware of his return; and he had been informed, quite truly, that Harispe’s cantonments were much scattered. Unfortunately he was as incapable as he was presumptuous, and he entirely lacked the fiery determination of his brother Henry, the hero of La Bispal. According to contemporary critics he was set, at this moment, on making what he thought would be a brilliant descent on an unprepared enemy, without any reference to his orders or to the general state of the war[739]. And he wished to fight before Roche’s troops were taken from him, as they must soon be.

[739] See Schepeler, pp. 617 and 623.

Harispe had only some 5,000 men--his own division, with one stray battalion belonging to Habert[740], and Delort’s cavalry brigade[741]. He had one infantry regiment in reserve at Alcoy[742], another at Ibi[743], the third[744]--with the bulk of Delort’s horsemen--in and about Castalla, the nearest point in the French cantonments to Alicante. O’Donnell’s ambitious plan was to surround the troops in Castalla and Ibi by a concentric movement of several columns marching far apart, and to destroy them before Harispe himself could come up with his reserve from the rear. Bassecourt and his detachment from the northern hills was ordered to fall in at the same time on Alcoy, so as to distract Harispe and keep him engaged--a doubtful expedient since he lay many marches away, and it was obvious that the timing of his diversion would probably miscarry.

[740] Of the 44th Line.

[741] The 13th Cuirassiers and 24th Dragoons.

[742] The 116th Line, 2 battalions.

[743] The 1st Léger, 3 battalions.

[744] The 7th Line, 2 battalions.

O’Donnell marched in three masses: on the right Roche’s division went by Xixona with the order to surprise the French troops in Ibi. The main body, three weak infantry brigades under Montijo, Mijares, and Michelena, with two squadrons of cavalry and a battery, moved straight upon Castalla. The main body of the horse, about 800 strong, under General Santesteban, went out on the left on the side of Villena, with orders to outflank the enemy and try to cut in upon his rear. The whole force made up 10,000 infantry and 1,000 cavalry, not taking into account the possible (but unlikely) advent of Bassecourt, so that Harispe was outnumbered by much more than two to one. As an extra precaution all the transports ready in Alicante were sent out--with only one battalion on board--along the coast, to demonstrate opposite Denia and the mouth of the Xucar, in order to call off the attention of Habert’s division, which lay in that direction.

Having marched all night on the 20th July, the Spanish columns found themselves--in a very fatigued condition--in front of the enemy at four o’clock on the morning of the 21st. They were out of touch with each other, Roche being separated from the centre by the mountain-spur called the Sierra de Cata, and the cavalry having been sent very far out on the flank. General Mesclop was opposed to Roche at Ibi, with four battalions and a squadron of cuirassiers--General Delort, at Castalla, had only one squadron of the cuirassiers, two battalions, and a battery, but was expecting the arrival of the 24th Dragoons from the neighbouring town of Biar, some way to his right, and of the remaining two squadrons of the 13th Cuirassiers from Onil on his left. Meanwhile he evacuated Castalla, but took up a position on a hillside covered by a stream and a ravine crossed by a narrow bridge, with his trifling force. He had already sent orders to Mesclop at Ibi to come in to his aid, leaving only a rearguard to hold off as long as possible the Spanish column in front of him. The latter did as he was bid; he threw into the Castle of Ibi a company of the 44th and two guns, and left the rest of that battalion and a troop of cuirassiers to support them. With the rest of his force, the three battalions of the 1st Léger and the remaining troop of cuirassiers, he set off in haste for Castalla.

O’Donnell assailed Delort in a very leisurely way after occupying the town of Castalla--his troops were tired, and four of his six guns had fallen behind. But Montijo’s brigade and the two pieces which had kept up with it were developing an attack on the bridge, and Michelena and Mijares had passed the ravine higher up, when the French detached troops began to appear from all directions. The first to get up were the 400 men of the 24th Dragoons, who--screened by an olive wood--came in with a tremendous impact, and quite unforeseen, upon Mijares’s flank, and completely broke up his three battalions. They then, after re-forming their ranks, formed in column and charged across the narrow bridge in front of O’Donnell’s centre, though it was commanded by his two guns. An attack delivered across such a defile, passable by only two horses abreast, looked like madness--but was successful! The guns only fired one round each before they were ridden over, and the brigade supporting them broke up. Delort then attacked, with his two battalions and with the cuirassiers who had just come up from Onil. Montijo’s brigade, the only intact Spanish unit left, was thus driven from the field and scattered. The 6,000 infantry of O’Donnell’s centre became a mass of fugitives--only one regiment out of the whole[745] kept its ranks and went off in decent order. Of the rest nearly half were hunted down and captured in droves by the French cuirassiers and dragoons. Mesclop with the three battalions from Ibi arrived too late to take any part in the rout. Delort sent him back at once to relieve the detachment that he had left in front of Roche. The latter had driven it out of the village of Ibi after some skirmishing, when he saw approaching him not only Mesclop’s column, but Harispe coming up from Alcoy with the 116th of the Line. He at once halted, turned to the rear, and retired in good order towards Xixona: the enemy’s cavalry tried to break his rearguard but failed, and the whole division got back to Alicante without loss. The same chance happened to Santesteban’s cavalry, which, marching from Villena at 7 o’clock, had reached Biar, in the enemy’s rear, only when the fighting--which had begun at 4--was all over. O’Donnell tried to throw blame on this officer; but the fact seems to be that his own calculation of time and distance was faulty: he had sent his cavalry on too wide a sweep, separated by hills from his main body, and had kept up no proper communication with it. Nor does it appear that if Santesteban had come up to Biar a little earlier he would have been able to accomplish anything very essential. Bassecourt’s diversion, as might have been expected, did not work: before he got near Alcoy the main body had been cut to pieces: he retired in haste to Almanza on hearing the news.

[745] Cuenca, of Montijo’s brigade. Schepeler, p. 619.

O’Donnell’s infantry was so shattered that his army was reduced to as bad a condition as it had shown in January, after Blake’s original disaster at Valencia. He had lost over 3,000 men, of whom 2,135 were unwounded prisoners, three flags, and the only two guns that had got to the front. The survivors of the three broken brigades had dispersed all over the countryside, and took weeks to collect. It was fortunate that Roche’s division had reached Alicante intact, or that city itself might have been in danger. The French had lost, according to their own account, no more than 200 men[746]: only the two cavalry regiments, two battalions of the 7th Line, and one of the 44th had been put into

## action. As Suchet truly remarks in his _Mémoires_[747] the total

numbers engaged--3,000 men[748]--on his side were somewhat less than the casualty list of the enemy.

[746] This is the figure given by Suchet in his contemporary dispatch to King Joseph, of which a copy lies in the Scovell papers. In some French accounts the number is cut down to 70.

[747] Vol. ii. p. 260.

[748] 7th Line, about 1,200; 13th Cuirassiers and 24th Dragoons, about 1,000; one battalion 44th, about 650; artillery, &c., about 150 = 3,000 in all. The 1st Léger and 116th Line were practically not engaged.

The rout of Castalla put the Murcian army out of action for months--a lucky thing for Suchet, since the force of the allies at Alicante was just about to be increased by the arrival there of Maitland’s expedition, and if O’Donnell’s army had been still intact, a very formidable body of troops would have been collected opposite him. It remains to explain the appearance of the Anglo-Sicilians in this direction, contrary to the orders of Wellington, who had expressed his wish that they should land in Catalonia, join Lacy, and lay siege to Tarragona--an operation which he thought would force Suchet to evacuate Valencia altogether, in order to bring help to Decaen.

We left Maitland anchored in the Bay of Palamos on July 31. The moment that he appeared Eroles went on board his ship, to urge his immediate disembarkation, and to promise the enthusiastic assistance of the Catalans. The energetic baron gave a most optimistic picture of the state of affairs, he declared that the whole country would rise at the sight of the red-coats, that Tarragona was weakly held, and that the total force of the French, including Suchet’s column near Tortosa, was only 13,000 men. Lacy and Sarsfield appeared later, and gave much less encouraging information: they rated the enemy at a far higher figure than Eroles, and were right in so doing, for Decaen had some 25,000 men, and could by an effort have concentrated 15,000, exclusive of succours from Suchet. The Spanish Army of Catalonia could only furnish 7,000 foot and 300 horse, of whom many were so far off at the moment that Lacy declared that it would take six or eight days to bring them up. By the time that they were all arrived, the French would have concentrated also, and would be equal in numbers to the whole force that the allies could collect. Tarragona was reported to be in a better state of defence than Eroles allowed, and the engineers declared that it might take ten days to reduce it. But the greatest problem of all was that of provisions: Lacy declared that the country could furnish little or nothing: he could not undertake to keep his own small army concentrated for more than a week. The Anglo-Sicilians must be fed from the fleet, and he could provide no transport. Evidently the expedition would be tied down to the shore, and the siege of Tarragona was the only possible operation. Since the Anglo-Sicilian army could not manœuvre at large or retire into the inland, it would have to fight Decaen, to cover the investment of Tarragona, within a few days of its landing. On the other hand if, as Eroles promised, the _somatenes_ rose on every side at the news of the disembarkation, the outlying French troops might not be able to get up to join Decaen, the roads would be blocked, the enemy might never be able to concentrate, and the force about Barcelona, his only immediately available field army, was not more than 8,000 strong, and might be beaten.

There were those who said that Lacy never wished to see the expedition land, because he was jealous of Eroles, and thought that a general rising which ended in success would have meant the end of his own power and tenure of office[749]. It is at least certain that the views which he expressed caused Maitland much trouble, and made him to flinch from his original idea of landing without delay and attacking Tarragona, according to Wellington’s desire. The English general took refuge in a council of war--the usual resource of commanders of a wavering purpose. His lieutenants all advised him to refuse to land, on the ground that his forces were too small and heterogeneous, that Lacy could give no prompt assistance, and that there was no sign as yet of the general rising which Eroles promised. Moreover, some of the naval officers told him that anchorage off the Catalan coast was so dangerous, even in summer, that they could not promise him that the army could be taken safely on board in case of a defeat. To the intense disgust of Eroles and the other Catalan leaders, but not at all to Lacy’s displeasure, Maitland accepted the advice of his council of war, and resolved to make off, and to land farther south. The original idea was to have come ashore somewhere in the midst of the long coast-line south of the Ebro, between Tortosa and Valencia, with the object of breaking Suchet’s line in the middle. But the news of Joseph O’Donnell’s gratuitous disaster at Castalla, which obviously enabled the Marshal to use his whole army against a disembarking force, and the suggestion that Alicante itself might be in danger, induced Maitland in the end to order his whole armament to steer southward. He arrived at Alicante on August 7th, and commenced to send his troops ashore--both his own 6,000 men and Whittingham’s 4,000 auxiliaries of the Balearic division. Since Roche was already there, with his troops in good order, there were 14,000 men collected in Alicante, over and above the wrecks of O’Donnell’s force. If only the Murcians had been intact, the mass assembled would have caused Suchet serious qualms, since it would have outnumbered the French corps in Valencia very considerably, and there was in it a nucleus of good troops in Maitland’s British and German battalions. The news of Salamanca had also come to hand by this time, and had transformed the general aspect of affairs in Suchet’s eyes: King Joseph was again demanding instant help from him, in the hope of retaining Madrid, and had called in (without his knowledge or consent) the division of Palombini from Aragon, and the garrison of Cuenca[750]. If Wellington should advance--as he actually did--against the King, and should drive him from his capital, it was possible that the main theatre of the war might be transferred to the borders of Valencia.

[749] This seems to have been Codrington’s view, see his _Memoirs_, i. p. 278, and he knew Lacy and the Catalans well.

[750] See above, pp. 487 and 488.

The Marshal therefore resolved to concentrate: he ordered Habert and Harispe to fall back behind the Xucar with their 8,000 men, abandoning their advanced positions in front of Alicante, and placed them at Jativa; here he threw up some field-works and armed a _tête-de-pont_ on the Xucar at Alberique. He ordered Paris’s brigade to come down from northern Aragon to Teruel, and he warned the generals in Catalonia that he might ask for reinforcements from them.

Maitland therefore, after his landing, found that the French had disappeared from his immediate front. He was joined by Roche, and by the 67th regiment from Cartagena, and proposed to drive Harispe from Castalla and Ibi. But he marched against him on August 16th-18th, only to find that he had already retired behind the Xucar. Farther than Monforte he found himself unable to advance, for want of transport and food. For the expedition from Sicily had not been fitted out for an advance into the inland. It had been supposed by Bentinck that the troops would be able to hire or requisition in Spain the mules and carts that they would require for a forward movement. But the country-side about Alicante was already exhausted by the long stay of the Murcian army in that region; and O’Donnell--before Maitland had come to know the difficulties of his position, got from him a pledge that he would not take anything from it either by purchase or by requisition. The British general had hired mules to draw his guns, but found that he could not feed them on a forward march, because the resources of the district were denied him. He himself had to stop at Elda, Roche at Alcoy, because the problem of transport and food could not be solved. All that he could do was to feel the French line of outposts behind the Xucar with a flying column composed of his own handful of cavalry--200 sabres--and a detachment of Spanish horse lent him by Elio, the successor of O’Donnell [August 20th-21st].

But even the thought of farther advance had now to be given up, for the news arrived that King Joseph had evacuated Madrid on the 14th, and was marching on Valencia with the 15,000 men that he had collected. To have tried any further attack on Suchet, when such an army was coming in from the flank to join the Marshal, would have been insane. The French force in this region would be doubled in strength by the King’s arrival. Wherefore Maitland drew back his own division to Alicante, and brought Roche back to Xixona, not far in front of that fortress, expecting that he might ere long be pushed back, and perhaps besieged there. Wellington in the end of the month, having the same idea, sent him elaborate directions for the defence of the place, bidding him to hold it as long as possible, but to keep his transports close at hand, and to re-embark if things came to the worst[751].

[751] See Wellington to Maitland, _Dispatches_, ix. p. 386, dated Aug. 30.

On the 25th King Joseph’s army and its vast convoy of French and Spanish refugees, joined Suchet’s outposts at Almanza, and the dangerous combination which Maitland and Wellington had foreseen came to pass. But what was still more threatening for the army at Alicante was the rumour that Soult was about to evacuate Andalusia, and to bring the whole of the Army of the South to Valencia. This would mean that nearly 80,000 French troops would ere long be collected within striking distance of the motley force over which Maitland and Elio now held command, and it seemed probable that Soult in his march might sweep over the whole country-side, disperse the Spanish forces on the Murcian border, and perhaps besiege and take Cartagena and Alicante as a _parergon_ on the way. We have seen in