SECTION XX OPERATIONS IN THE EAST AND SOUTH OF SPAIN DURING THE SPRING AND SUMMER OF 1810 CHAPTER I

SUCHET AND AUGEREAU IN ARAGON, VALENCIA, AND CATALONIA, MARCH–JULY 1810

Though Suchet had successfully pacified the plains of Aragon during the autumn of 1809, and though Augereau in the last month of that year had received the surrender of the much-enduring garrison of Gerona, the position of the French in North-Eastern Spain was still far from satisfactory. It was not yet possible for the 3rd and the 7th Corps to combine their operations. While the broad strip of territory in Western Catalonia reaching from the foot of the Pyrenees to the sea—whose places of strength were Lerida, Tarragona, and Tortosa—still remained intact, Augereau and Suchet could still communicate only by the circuitous route through France: a letter from Saragossa to Barcelona took a fortnight or more to arrive at its destination. It was high time that they should endeavour to get into touch with each other, by cutting through the Spanish line of defence. Both of the corps-commanders had now a comparatively free hand, and could assemble a considerable force for offensive operations. Suchet’s position was the happier: he had no Spanish army opposed to him at the moment. His only opponents were—to the West the bold guerrillero Mina the Younger on the borders of Navarre; to the South Villacampa with the dilapidated remains of his band on the side of the Sierra de Albaracin and Molina; and to the East the governor of Lerida, who kept a flying force under Colonel Perena on foot between the Cinca and the Segre, for the encouragement of the insurgents of Eastern Aragon. None of[p. 283] these three opponents could do any serious mischief to the 3rd Corps, which had risen by the New Year of 1810 to a strength of 24,000 men[329], now that its drafts from France were beginning to arrive. Mina was the most tiresome of the three: he was a young man of untiring energy, and kept the borders of Navarre in a perpetual ferment, till he was finally put down by mere force of numbers. For while Suchet sent a number of columns under General Harispe to hunt him, Lagrange’s division of the newly arrived 8th Corps came down from the side of Pampeluna, and swept the valleys of the Arga and the Aragon from the other quarter. Pursued from one refuge to another by 12,000 men, Mina ended by bidding his followers disperse. He did not appear again as a combatant till the 8th Corps had passed on into Castile in the month of February 1810.

Villacampa and Perena being very weak adversaries, Suchet could now dispose of the greater part of his corps for offensive operations. Two lines of attack were open to him: the more natural and obvious one was to attack Lerida or Tortosa, while Augereau and the 7th Corps moved against them from the other side. This was Suchet’s own intention, and he had received dispatches from Paris which approved of the plan. But another objective was pointed out to him from another quarter. It will be remembered that while King Joseph and Soult were planning the details of their Andalusian expedition, one of the schemes which they discussed was the use of the 3rd Corps as a flanking detachment to cover their attack on Seville[330]: it was to march on Valencia while Mortier marched on Badajoz. The plan was rejected in favour of the direct frontal march against the passes of the Sierra Morena. Nevertheless, when the defiles had been passed, and the King had entered Cordova in triumph, he recurred to his original scheme, and on January 27 Soult sent orders to Suchet directing him to make a dash at Valencia, and assuring him that the demoralization of the Spaniards was so great that he would probably enter the city without meeting with much resistance. Similar orders, as we have already seen,[p. 284] were sent to Ney, who was directed at the same moment to make himself master of Ciudad Rodrigo[331], which Soult and the King imagined to be likely to surrender at the first summons.

Suchet, though doubting (as he says) the wisdom of the plan of campaign imposed upon him, prepared to obey, and concentrated 8,000 men, consisting of Laval’s division and half that of Musnier, at Teruel, on the borders of Valencia, while Habert, with six battalions more, was to move in a separate parallel column from Alca?iz, and to strike down on to the sea-coast of Valencia by way of Morella. On their way to Teruel some of Laval’s columns made a side-stroke at Villacampa, and drove off his band into the mountains of New Castile.

The march on Valencia had already begun, when Suchet received at Teruel, on March 1, the dispatch from the Emperor which announced to him that the kingdom of Aragon had been made an independent ‘military government’ under his charge[332], and that he was for the future to take no orders from Madrid, but to seek all his instructions from Paris. By the same dispatch Suchet was informed that the Emperor approved of the plan for directing the 3rd Corps against Lerida and the rear of Catalonia which he had himself always advocated. But since Habert’s column had already disappeared in the mountains, and could not be recalled, and since his own advanced guard had actually crossed the borders of Valencia, Suchet thought that it was now too late to turn back, and that he must endeavour to carry out Soult’s orders, even though they were now countermanded by the Emperor. Accordingly he marched by Sarrion, on the road which leads from Teruel to Segorbe, and from thence to Murviedro on the Valencian coast, where Habert was to join him.

The force opposed to Suchet was that Valencian army which the governor—or rather the dictator—José Caro had so persistently refused to lend to Blake in the preceding autumn for the relief of Gerona. It was about 12,000 strong at the moment, and as some nine months had elapsed since the disaster of Belchite, where the Valencian troops had fared so badly, the regiments had been recruited up to their old strength, and had been thoroughly reorganized and re-equipped. Nevertheless,[p. 285] Caro did not wish to risk them in the open field, and, when Suchet’s approach was reported to him, ordered the troops to prepare to retreat within the walls of the city of Valencia, which had been so successfully defended against Moncey two years before. He was probably right in his decision: having no great superiority of numbers over the French invading columns, he would have been risking over much if he had offered battle on the frontier, even in an advantageous position. Behind the fortifications of Valencia, which had been improved and enlarged during the last two years, he could at least be certain of making a long and formidable resistance. If the enemy appeared without a great battering-train, and unprepared for a regular siege, the garrison would be absolutely safe. Accordingly, Caro sent out only a single brigade to observe and retard Suchet’s advance: this force, assisted by some armed peasants, tried to hold the defile of Alventosa against the French (March 4), but was beaten out of the position with ease, lost four guns, and retired by Segorbe and Murviedro on to Valencia, without offering further resistance. Suchet, following in the wake of the Spanish detachment, reached Murviedro on March 5, and there met Habert, whose march by Morella and Castellon-de-la-Plana had been absolutely unopposed.

A day later the French appeared in front of the fortifications of Valencia. Having surveyed the outworks, and summoned the place in vain, Suchet found that there was no more to be done. His bolt was shot: the city intended to defend itself: it was full of troops, and the whole population had been armed. Caro had given notice that the defence was to be conducted in the style of Saragossa, and as a foretaste of his intentions court-martialled and hanged several persons who were accused of being traitors[333]. After lying four days in front of Valencia, Suchet marched off again on the night of March 10-11, having[p. 286] convinced himself that he had been sent on a fool’s errand by Soult and King Joseph. For the expedition had only been undertaken on the hypothesis that the enemy was demoralized and ready to surrender—which was clearly not the case. No siege train had been brought, and without heavy artillery Valencia was impregnable. The retreat of the French was hastened by the fact that their communications with Aragon had already been cut off. Villacampa, undismayed by his defeat in February, had returned to the front the moment that the invading columns had passed on to the Valencian coast. He had blockaded the garrison of Teruel, cut off two small columns which were moving in its neighbourhood, and captured four guns and some 300 prisoners. At the same moment, though Suchet was unaware of the fact, Colonel Perena, with a detachment from Lerida, had made a demonstration against Monzon, and when the garrison of Eastern Aragon concentrated to defend it, had fallen on Fraga and burnt the bridge there—the only crossing of the Cinca which was in French hands. Mina, too, warned of the departure of the main body of the 3rd Corps for the South, had reassembled some of his levies, and begun to render the road from Saragossa to Pampeluna once more unsafe.

Suchet, therefore, on withdrawing from Valencia, found plenty of occupation awaiting him in Aragon. Villacampa was once more chased into the wilds of the Sierra de Albaracin: and several columns were detached in pursuit of Mina. This time the young partisan was unfortunate—he was caught between one of Suchet’s detachments and a party sent out by the governor of Navarre, and taken prisoner on March 31, with some scores of his followers. His place, however, was taken ere long by his uncle Espoz y Mina, who rallied the remnant of the guerrilleros of North-Western Aragon, and continued the struggle with an energy as great, and a success far greater, than that of his nephew.

Meanwhile, the months of February and March had been wasted, so far as the 3rd Corps was concerned, and it was not till April that Suchet was able to take in hand the enterprise which he himself had approved, and to which the Emperor had now given his consent—the siege of Lerida. A few weeks after his return to Saragossa he received from Paris two imperial[p. 287] letters denouncing his late campaign as presumptuous, objectless, and altogether deserving of high condemnation[334].

While the Valencian expedition had been in progress Marshal Augereau had been trying his hand at the Catalan problems which Verdier and St. Cyr had found so puzzling. His luck had been bad—and his faults had been many—for of all the commanders-in-chief who successively endeavoured to subjugate the untameable principality he was decidedly the least capable. When Gerona fell, the main body of his army was let loose for operations in the open field. But Verdier’s divisions, on whom the burden of the siege work had fallen, were in much too dilapidated a condition for active service[335]. Leaving them to recover their strength and gather up their convalescents, in cantonments between Figueras and Gerona, Augereau moved southwards in January, at the head of the troops of Souham and Pino, whose two divisions still counted 12,000 bayonets with the eagles, though they left behind them 5,000 sick in the hospitals of Figueras and Perpignan.

The Marshal’s main duty at this moment was to clear the road from Figueras to Barcelona, which the Spanish fortress of Gerona had so effectually blocked during the whole year 1809. All further advances were suspended till this operation should have been completed, by the Emperor’s special orders. Duhesme, indeed, isolated in Barcelona with some 6,000 men, communicating with France only by sea, and surrounded by a discontented population, which was reduced to semi-starvation by the blockade kept up by the British fleet, was in a perilous condition. There had been many who thought that if Blake, in October 1809, had marched against Barcelona instead of trying to throw succours into Gerona, he would have forced Augereau to abandon the siege of the latter place. For Duhesme would have been in such danger, since he was neither strong enough to fight in the open nor to man efficiently the immense circuit of the walls of Barcelona, that Augereau might[p. 288] probably have judged that it was better to save the capital of Catalonia than to continue the interminable blockade of the smaller town that was holding out so obstinately against him. Be this as it may, Augereau set Souham’s and Pino’s columns on the march early in January 1810, after having first executed at their head a sort of battue of the somatenes of the mountains around Gerona, as far as Ripol and Campredon. His progress was a reign of terror: he had issued on December 28th a proclamation stating that all irregulars taken with arms in their hands were to be treated as simple highway robbers, and hung without any form of trial[336]. He was as good as his word, and all the prisoners taken by the flying columns of Souham and Pino were suspended on a line of gallows erected on the high-road between Gerona and Figueras. These atrocities turned the war in Catalonia into a struggle even more ferocious than that of 1809, for the natives, very naturally, retaliated upon every French straggler that they caught, and the Spanish regular officers had great difficulty in saving the lives even of large bodies who had laid down their arms upon formal terms of surrender.

Two roads lead from Gerona and the valley of the Ter to Barcelona—that by Vich and that by Hostalrich. They unite near Granollers, twenty miles outside Barcelona. Augereau marched with the Italian division of Pino by the second road, while he sent Souham by the other. Both met with opposition: the small castle of Hostalrich was still held by the Spaniards—it was the only fortified place in central Catalonia which was now in their hands, and commanded, though it did not actually block, the main road to Barcelona. On arriving before it, the Marshal ordered Mazzuchelli’s Italian brigade to form the siege, trusting that the place would fall in a few days; but finding the resistance more serious than he had expected, he finally continued his march, with the remainder of the Italian division, towards Granollers and Barcelona. He had ordered Duhesme to come out to meet him at the former place, with as many of his troops as he could spare from garrison duty.
Map ofCcatalonia

Enlarge  CATALONIA

[p. 289]Souham meanwhile, on the eastern road, had entered Vich on January 11. He found that it had just been evacuated by the main body of the Spanish Army of Catalonia, whose head quarters had been maintained in that town all through the later months of the siege of Gerona. This force, still numbering about 7,000 men, was no longer under the command of its old general. Blake had resigned at Christmastide, after a lively altercation with the Junta of Catalonia, who laid the blame for the fall of Gerona on his shoulders, while he maintained that they were to blame for not having ordered the general levée en masse of the principality at an earlier date[337]. The command then passed in quick succession to the two senior general officers of the Catalan army, first to the Marquis of Portago, and then, after he had fallen ill, to General Garcia Conde, Governor of Lerida. But these interim commanders-in-chief were replaced in a few weeks by a younger and much more active leader, Henry O’Donnell—the man who had conducted the second convoy for Gerona, and had cut his way out of the place with such splendid address and resolution. He had only just been raised to the rank of Major-General, yet was now entrusted by the Central Junta with the control of the entire army of Catalonia. As he was setting to work to reorganize it, Souham arrived in front of his head quarters. Judging it unwise to offer battle at the head of demoralized troops, O’Donnell evacuated Vich, and fell back to the mountains above it. Souham followed him with an advanced guard as far as the pass called the Col de Suspina, where the Spaniard suddenly turned upon the pursuing force, and hurled it back with loss upon the main body of the division (January 12, 1810). He refused, however, to face Souham next day, and retreated towards Manresa; thereupon the French general also turned back, thinking it profitless to leave the Vich-Barcelona road, and to plunge into the hills in pursuit of an evasive enemy. He ended by making his way across the mountains to join the Marshal.

Meanwhile, Augereau, with Pino’s Italian brigades, reached Granollers, and came into touch with Duhesme’s troops, just in time to find that a disaster had preceded his arrival. The Governor of Barcelona, in obedience to the orders sent him, had[p. 290] marched out on January 16 with three battalions and 250 cuirassiers in order to meet the Marshal. He waited four days at Granollers, and then left the brigade under the charge of Colonel Guétry, and returned in person to Barcelona, recalled by rumours of a threatened attack from the side of Tarragona. On the morning after his departure Guétry’s force, which believed itself in complete security, was suddenly surprised in its camps by a detachment sent out by O’Donnell. That enterprising officer had heard that Augereau was still twenty miles away, and that Guétry had scattered his men in the three villages of Santa Perpetua, Mollet, and Granollers, in order to cover them from the bitter January weather. Accordingly, he resolved to risk an attack on this unwary detachment. Two brigades—about 4,000 men—under the Marquis of Campo Verde, made a forced march across the mountains, and fell upon the villages at dawn. The battalion of the 112th in Santa Perpetua was completely cut to pieces or captured—only two men are said to have escaped. The 7th and the cuirassiers at Mollet fared somewhat better, but were driven back to Barcelona with heavy loss. Part of the third battalion, one of the 5th Italian Line regiment, which was posted at Granollers, escaped destruction by throwing itself into a fortified convent, and held out for two days, till it was saved by the approach of Augereau along the high-road. Altogether, Duhesme’s division was thinned by the loss of 1,000 men on January 21-22: Guétry had been taken prisoner, with some 600 men more and two guns. The Spaniards disappeared the moment that Augereau came in sight, and rejoined O’Donnell in the hills.

Augereau entered Barcelona on the twenty-fourth, and at once asserted his authority by deposing Duhesme from the governorship and sending him home to France. They were old enemies, and the general’s friends regarded his disgrace as a display of spite on the part of his superior[338]. But as all the Spanish narratives describe the eighteen-months’ dictatorship of Duhesme as having been as much distinguished for private[p. 291] rapacity as for public oppression, it is probable that the Marshal’s action was wholly justifiable. The Emperor, however, refused to sanction the prosecution of the general on the charges laid against him, remarking that such proceedings would give too much pleasure to the Catalans, ‘il y avait bien autre chose à faire que de réjouir les Espagnols par cette réaction[339].’

Having concentrated Pino’s and Souham’s troops at Barcelona, Augereau would have proceeded to advance against the Catalans, and lay siege to Tarragona, but for one fact—the magazines of the city were almost empty, and no food could be procured for the army. Indeed, after a very few days it was necessary for the Marshal to retrace his steps, in order to bring up an enormous convoy for the revictualling of the place, which was being collected at Perpignan and Figueras. All that he had done by his march was to open up the road, and to muzzle the fortress of Hostalrich, which was still being blockaded by the Italians. Accordingly, on February 1 his two divisions marched back each by the way that it had originally taken—Souham to Vich, where he halted, Pino to Gerona, where the convoy began presently to gather, escorted thither in detachments by a large body of reinforcements which had just come up from France. For the Emperor had strengthened the Army of Catalonia by a division of troops of the Confederation of the Rhine, under General Rouyer, and by a Neapolitan brigade—some 8,000 men in all. But the long train of carts and mules came in slowly, and March began before Augereau was ready to move.

Meanwhile, his lieutenant Souham had been exposed to a sudden and unexpected peril. O’Donnell had discovered that the division at Vich was completely isolated and did not count much more than 5,000 sabres and bayonets. Having reorganized his own field army at Moya, near Manresa, and brought it up to a strength of 7,000 foot and 500 horse by calling a few troops from Tarragona, he directed the somatenes of Northern Catalonia to muster on the other side of Vich, so as to fall on Souham from the rear. The indefatigable miquelete leaders Rovira and Milans got together between 3,000 and 4,000 men, despite of all their previous losses and defeats. On February 19 these levies thrust in the pickets of the French division on the eastern[p. 292] side, but Souham did not see his danger till, on the following morning, he found O’Donnell’s regular troops pouring down into the plain of Vich in three columns, and challenging him to a battle in the open. Since the day of Valls the Catalan army had never tried such a bold stroke.

The French general was greatly outnumbered—he had but 4,000 infantry and 1,200 horse to oppose to O’Donnell’s 7,500 regulars and the 3,500 miqueletes. The action fell into two separate parts—while Rovira and Milans bickered with two battalions left to guard the town of Vich, Souham fought a pitched battle against the Spanish main body with the eight battalions and two and a half regiments of cavalry[340] which constituted the remainder of his force. It was a fierce and well-contested fight: O’Donnell took the offensive, and his men displayed an unwonted vigour and initiative. Unhappily for this enterprising general his small body of horsemen was utterly unable to restrain or to cope with the superior French cavalry. Twice the battle was turned by the charge of Souham’s squadrons: after the first repulse O’Donnell rallied his beaten right wing, threw in all his reserves, and tried to outflank the shorter French line on both sides. The enemy was losing heavily and showing signs of yielding, when the whole of his cavalry made a second desperate charge on the Spanish right. It was completely successful; O’Donnell’s turning column, composed of the Swiss regiments of Kayser[341] and Traxler, was broken, and the larger part of it captured. Thereupon, the Spanish general, who had displayed undaunted courage throughout the day, and headed several charges in person, thought it time to retire. He fell back on the mountains, leaving behind him 800 killed and wounded, and 1,000 prisoners. The French[p. 293] had suffered at least 600 casualties[342], including Souham himself, desperately wounded in the head, and had been within an ace of destruction: but for their superiority in cavalry, the day was lost.

On March 13, Augereau had at last collected his vast convoy at Gerona—there were more than a thousand waggons laden with flour, besides pack-mules, caissons, carriages, and other vehicles of all sorts. He marched in person to escort it, with the Italians—who were now under Severoli, Pino having gone home on leave,—and the newly arrived German division of Rouyer. Verdier was left behind—as in January—to defend the Ampurdam from the incursions of the miqueletes of Rovira and Milans. Meanwhile, Souham’s division, which had passed into the hands of General Augereau, the Marshal’s brother, since its old chief had been invalided to France, pursued a line of march parallel to that of the main force. Moving from Vich by the Col de Suspina and Manresa, it came down into the valley of the Llobregat, on the same day that Pino’s and Rouyer’s troops reached it by the other route. The Marshal and the main column had made their way past Hostalrich, which was found still unsubdued, and still blockaded by Mazzuchelli’s brigade, which had been left opposite it in January. These troops were ordered to join their division, a mixed detachment under Colonel Devaux being left in their stead to watch the castle. The Marshal then took up his residence in the palace at Barcelona, and had himself proclaimed Governor of Catalonia with great state, in consonance with the imperial decree of February 8, which had taken the principality out of the hands of King Joseph, and made its administrator responsible to Paris alone, and not to Madrid.

Augereau established himself permanently at Barcelona, and proceeded during the next two months to act rather as a viceroy than as a commander-in-chief. The conduct of military operations he handed over to his brother, to the unbounded disgust of the other generals. In what proportions the responsibility[p. 294] for the disasters which followed should be distributed between the two Augereaus it is hard to say. But Napoleon, very naturally and reasonably, placed it on the Marshal’s shoulders, and wrote that ‘Ce n’est point en restant dans les capitales éloignées de l’armée que des généraux en chef peuvent acquérir de la gloire ou mériter mon estime[343].’ The force assembled on the Llobregat was now a very large one, consisting of three divisions, those of Augereau, Severoli, and Rouyer. It numbered nearly 20,000 men, for along with the convoy there had marched a mass of drafts for the old regiments of the 7th Corps, which brought their battalions up to full strength[344].

There were only two rational plans of campaign open to the French: the one was to march on Tarragona with a siege-train, and to complete the conquest of Catalonia by the capture of its greatest fortress. The other was to mask Tarragona, and to strike across the principality westward, in order to get into communication with Suchet and the Army of Aragon, by way of Igualada, Cervera, and Lerida. If the two corps could meet, Northern Catalonia would be completely isolated from Tarragona, Tortosa, and Valencia. It was this latter plan which Augereau had been ordered to carry out by his master. He was directed to march on Lerida, and to join Suchet before that place. By a dispatch of February 19 he had been informed that on March 1 the Army of Aragon would have arrived before Lerida, and would be forming its siege[345]. This prophecy was false: for Suchet, as we have seen, had gone off on his Valencian expedition, and was at Teruel on the appointed day. It was impossible to direct from Paris a combined movement depending on accurate timing for its success. But Augereau should have attempted to carry out the order sent him by his master with proper zeal and dispatch. This he failed to do: his own inclination was to strike a blow at Tarragona, and the movements of his corps show that this was the operation which he had determined to carry out. Instead of marching in person[p. 295] with his main body, on Cervera and Lerida, he directed his brother to take his own and Severoli’s divisions and to move by Villafranca on Reus, a large town twelve miles to the north-west of Tarragona, and suitable as a base of operations against that city (March 29). A battalion and a half was dropped at Villafranca, to keep open the communications between Reus and Barcelona, while a brigade of Rouyer’s newly arrived Germans, under the ever-unlucky Schwartz—the vanquished of Bruch and Esparraguera[346]—was placed as a sort of flank-guard at Manresa. On March 27 a summons was sent in to Tarragona demanding surrender. This was, of course, refused—the town was full of troops, for Henry O’Donnell had just strengthened its garrison by retiring into it himself, with the 6,000 men who represented the remains of his field army.

The Spanish general, contemplating the position of affairs with a wary eye, had convinced himself that he could stop General Augereau’s further movements by striking at his line of communication with Barcelona. This he proceeded to do in the most skilful and successful fashion. Before the enemy closed in upon Tarragona, he sent out a picked force under General Juan Caro, with orders to attack Villafranca and Manresa without a moment’s delay. This daring stroke was completely successful: Caro stormed Villafranca at dawn on March 30, and took or captured the whole of the 800 men posted there. He was wounded, and had to hand over his command to Campo Verde, who, after some preliminary skirmishing lasting for three days, completed the little campaign by driving Schwartz out of Manresa, after a heavy fight, on April 5, in which the German brigade lost 30 officers and 800 men[347]. The somatenes turned out to hunt the routed force as it retired on Barcelona, and inflicted many further losses, so that Schwartz only brought back a third of his brigade to the Marshal.

The touch between the two divisions at Reus and the garrison of Barcelona was thus completely severed. Moreover, Augereau[p. 296] feared lest the next move of the raiding force might be an attack on the isolated force under Colonel Devaux which was blockading Hostalrich. He therefore sent out a brig with one of his aides-de-camp on board, bearing orders to his brother and Severoli to return at once to his head quarters with their troops. By good fortune the messenger reached his destination without being intercepted by English cruisers, and the two divisions started on their march back to Barcelona on April 7. Only two days before, Severoli, more by luck than by skill, had got for a moment into touch with the Army of Aragon. An exploring column of two battalions under Colonel Villatte, which he had sent out westward to Falcet, in the direction of the Ebro, met—entirely by chance—a similar detachment of Musnier’s division of the 3rd Corps, which had advanced to Mora. On interchanging news of their respective armies, Villatte learnt from his colleague that Suchet was at this moment marching on Lerida, where, according to the Emperor’s instructions to Marshal Augereau, he ought to have arrived more than a month before. Villatte brought back the information to Severoli, but the latter could make no use of it while he was under orders to return without delay to Barcelona[348]. He and General Augereau accomplished their retreat in three days, much harassed on the way, not only by the somatenes of the mountains but also by O’Donnell, who followed them with a detachment of the garrison of Tarragona. They took some small revenge on him, however, at Villafranca, where they turned on the pursuers and inflicted a sharp check on their advanced guard, which was pressing in, with more courage than discretion, on a force which outnumbered it by four to one. On April 9 the whole army was encamped outside the walls of Barcelona. Counting the garrison, Augereau had now more than 20,000 men in hand, but, not contented with this concentration, he sent orders back to the Ampurdam, ordering Verdier to bring forward as many troops as possible to Gerona, and from thence to push forward down the high-road as far as Granollers, so as to succour Devaux, if the Spaniards should show any intention of relieving Hostalrich.

The Marshal’s next move lost him the favour of Napoleon,[p. 297] and led to his removal in disgrace to Paris. On April 11 he issued orders to Severoli to march on Hostalrich, and there to take over the conduct of the siege from Devaux, while at the same time he himself moved back to Gerona with his brother’s French division, and an immense train, consisting partly of the empty carts of the convoy which he had brought south in March, partly of confiscated property of all sorts from Barcelona[349]. Augereau’s excuse for this retrograde movement, which abandoned all Central Catalonia to the enemy, was that his army would have exhausted the magazines of Barcelona if he had kept it concentrated for ten days more, and that he had no other way of feeding it, since the activity of the somatenes made the dispatch of foraging detachments utterly impossible. The moral effect of the move was deplorable: after taking in hand an offensive movement against Tarragona, he had allowed himself to be checked by an enemy hopelessly inferior in numbers, had lost over 3,000 men in petty combats, and then had retired to the base from which he had started in January. Three months had been wasted, and nothing had been gained save that the small castle of Hostalrich was now in a desperate condition for want of food, and must fall if not speedily succoured. What rendered the position of Augereau the more shameful was that he had now in front of him only a skeleton enemy. For O’Donnell at this moment was distracted by the advance of Suchet against Lerida, and had been forced to draw off towards the borders of Aragon two of the four divisions into which he had reorganized his little field army. Facing Augereau there was only Campo Verde’s division of regulars—not over 5,000 strong—and the bands of the mountains. It was against such an enemy that the Marshal had concentrated 25,000 men—including Verdier’s force—between Hostalrich and Gerona.

The food-problem, on which those who defend Augereau lay all the stress possible[350], was no doubt a very real one.[p. 298] The feeding of Barcelona, with its garrison and its large civil population, by means of convoys brought from France, was no easy task. But it seems clear that with the large force at his disposition—the 7th Corps was over 50,000 strong[351]—Augereau could have organized convoy-guards sufficiently powerful to defy the Spaniards, without abandoning the offensive position in front of Barcelona, which was all-important to him. He might have set aside 10,000 men for that purpose, and still have kept a strong field army together. Had he done so, O’Donnell could not have dared to march with 8,000 men, the pick of the Catalan army, to relieve Lerida, while his old enemies were massed, awaiting an impossible attack, a hundred miles away.

Meanwhile, the one task which was completed by the 7th Corps during the months of April and May was the reduction of Hostalrich. The little castle was a place of strength, perched high above the abandoned town to which it belonged. It was garrisoned by two battalions, one of Granadan regulars from Reding’s old division, the other of local Catalan levies[352], under Colonel Juliano Estrada, an officer of high spirit and commendable obstinacy. To the first summons made to him in January by Augereau, he had replied that ‘Hostalrich was the child of Gerona, and would know how to emulate the conduct of the mother-city.’ Nor was this high-flown epigram belied by the after-conduct of the garrison. Hostalrich held out from January 16 to May 12 without receiving either a convoy from without or any reinforcements. Twice the local miqueletes attempted to pass succours into the castle, but on each occasion they were driven off, after having barely succeeded in communicating with Estrada. Their convoys never got close enough to enter the gates. At last, on May 12, the provisions of the place were completely exhausted: thereupon Estrada took in hand a scheme which was only tried by one other governor during the whole Peninsular War—the Frenchman Brennier at Almeida in May 1811. He put every able-bodied man of the garrison under arms at midnight, issued[p. 299] silently from the castle, and charged at the besiegers’ lines, with the intention of cutting his way through to Campo Verde’s head quarters in the mountains. He was fortunate enough to pierce the Italian outposts at the first rush. But two whole brigades were presently at his heels, his guides lost their way among cliffs and ravines, and he was overtaken. Turning back with his rearguard to fend off the pursuers, he was wounded and taken prisoner. Ten other officers and some 300 soldiers shared his fate: the rest of the garrison dispersed in the hills, and reached Vich in small parties to the number of 800 men. If only their gallant commander had escaped with them, the exploit would have been an exact parallel to the Almeida sortie of 1811.

The reduction of Hostalrich and of a small fort on the Isles of Las Medas near the mouth of the Ter, were the last achievements of the Army of Catalonia while it was under the command of Augereau. On April 24 the Emperor had resolved to remove him from his post, and had ordered Marshal Macdonald, Duke of Tarentum, to relieve him. In the Moniteur it was merely stated that the elder Marshal had been forced by ill-health to resign his charge. But the real causes of his displacement are plainly stated in Napoleon’s letter to the Minister of War. They were the blow which he had given to the imperial prestige by his retreat to Gerona, and his disobedience in having failed to march on Lerida in aid of Suchet. It must be confessed that the wrath of the Emperor was justifiably roused. Its final manifestation took the shape of a minute directing that when the Moniteur published the news of the successes at Hostalrich and Las Medas, the name of Augereau was to be kept out of the paragraph, and only those of his subordinates mentioned[353].

While the new commander of the 7th Corps was on his way from Italy to Perpignan, and for some time after his arrival on May 22, there was a complete suspension of active operations. None of the French generals in the principality could do more than await orders, and keep the neighbourhood of his cantonments safe from the irrepressible miqueletes. The whole interest[p. 300] of the French operations in North-Western Spain during this period turns on the movements of the Army of Aragon.

On returning from his futile Valencian expedition, Suchet had found awaiting him at Saragossa orders to commence without delay the siege of Lerida, and promising that Augereau would ‘contain’ the Spanish army of Catalonia, and would ultimately open up communication with him by way of Cervera or Momblanch. In obedience to the imperial mandate, the commander of the 3rd Corps collected more than half his troops for the expedition against Lerida. This concentration entailed the abandonment of the rough country of Southern Aragon to Villacampa and his bands; for having determined to leave only Laval’s division behind him to protect the central parts of the province, Suchet had to evacuate the region of Teruel, Albaracin and Montalvan. Laval’s two brigades remained in the valley of the Jiloca, between Saragossa and Daroca, watching the debouches from the mountains. The other two divisions of the Army of Aragon were destined for the siege of Lerida. Musnier, with one column, marching down from Caspe and Alca?iz[354], crossed the Ebro at Flix, and approached Lerida from the south. Habert, with his division, coming from the side of Saragossa, made for the same goal by way of Alcubierre and Monzon. It might seem a dangerous experiment to march the two isolated columns across the Spanish front, for a junction at a point so remote from their respective starting-points. But, for the moment, there was no hostile force of any importance in their neighbourhood. Severoli and Augereau being still at Reus, when Musnier and Habert started on their way, O’Donnell could spare no troops for the relief of Lerida. It was not till they had retired, and the Duke of Castiglione had withdrawn all his forces to Hostalrich and Gerona, that the Captain-General found himself free for an expedition to the West. In front of Musnier, while he was marching on Lerida, there were nothing but trifling detachments sent out for purposes of observation by the Governor of Tortosa. In front of Habert there were only 2,000 men under Colonel Perena, thrown out from Lerida to observe the open[p. 301] country between the Segre and the Cinca. This was the sole field-force which remained from the old Spanish Army of Aragon, with the exception of the two dilapidated regiments that formed the nucleus of Villacampa’s band.

On April 10, Suchet arrived at Monzon with a brigade of infantry, and the six companies of artillery and four of engineers which were destined for the siege of Lerida. On the thirteenth he was before the walls of the place on the west bank of the Segre, while on the next day Habert appeared on the other bank, driving before him Colonel Perena and his battalions. This second French column had passed the Segre at Balaguer, from which it had expelled the Spanish brigade. Perena, having the choice between returning into the mountains or strengthening the garrison of Lerida, had taken the latter alternative. On the same day Musnier came up from the South, and joined the army with the force that had started from Alca?iz and had crossed the Ebro at Flix. Suchet had now concentrated the whole of the 13,000 men[355] whom he destined for the siege.

The town of Lerida, then containing some 18,000 inhabitants, lies in a vast treeless plain on the western bank of the broad and rapid Segre. Its topography is peculiar: out of the dead level of the plain rise two steep and isolated hills. One of these, about 800 yards long by 400 broad, and rising 150 feet above the surrounding flats, is crowned by the citadel of the town; the other, somewhat longer but rather narrower, is no other than the hill about which Caesar and Afranius contended in the old Civil Wars of Rome, when Lerida was still Ilerda. It lies about three-quarters of a mile from the first-named height. In 1810 its culminating summit was covered by a large work called Fort Garden, and its southern and lower end was protected by the two small redoubts called San Fernando and El Pilar. The town occupies the flat ground between the citadel-height and the river. It has no trans-pontine suburb, but only a strong[p. 302] tête-du-pont on the other side of the Segre. The strength of the place lay in the fact that its whole eastern front was protected by the river, its whole western front by the high-lying citadel. The comparatively short southern front was under the guns of Fort Garden, which commanded all the ground over which an enemy could approach this side of the town. There only remained, unprotected by outer works, the northern front, and this was the point which a skilful enemy would select as his objective. Another fact which much favoured the defence was that the besieger was forced to make very long lines of investment, since he had to shut in not only the town, but the whole plateau on which lie Fort Garden and its dependent works. Moreover, he must keep a force beyond the Segre, to block the tête-du-pont, and this force must be a large one, because the east was the quarter from which succours sent from Catalonia must certainly appear. The garrison, on the other hand, was quite adequate to man the works entrusted to it; the entry into the town of Perena’s Aragonese battalions had brought it up to a strength of over 8,000 men. The artillery were the weak point—there were only 350 trained men to man over 100 pieces. The Governor was Major-General Garcia Conde, who owed his promotion to his brilliant achievements at Gerona in the preceding September.

Suchet’s 13,000 men were insufficient for a close and thorough investment of such a large space as that covered by the Lerida-Garden fortifications. He was forced to leave a considerable part of its southern front watched by cavalry posts alone. Of his total force, Musnier, with six battalions and the greater part of the cavalry—about 4,000 men in all, was beyond the river. Three battalions blocked the tête-du-pont, the rest formed a sort of covering force, destined to keep off the approach of enemies from the side of Catalonia. On the hither side of the Segre the three brigades of Habert, Buget, and Vergès were encamped facing the western and north-western fronts of Lerida. They counted only thirteen battalions, being weakened by 1,200 men who had been detached to garrison Monzon and Fraga. The communication between Musnier’s division and the rest of the army was by a flying bridge thrown across the Segre two miles above the town; it was guarded by a redoubt. The artillery[p. 303] park was placed at the village of San Rufo, opposite the north-western angle of the town, from which it is distant about 2,500 yards. It was evident from the first that this would be the point of attack selected by the French engineers.

The army had barely taken up its position when, on April 19, rumours were spread abroad that O’Donnell was on his way to relieve Lerida. They must have been pure guesses, for the Spanish general on that day was still at Tarragona and had made no movement. They were, however, sufficiently persistent to induce Suchet to take off Musnier’s covering corps, strengthened by three other battalions, on a mission of exploration, almost as far as Tarrega, on the Barcelona road. No Spanish force was met, nor could any information be extracted from the inhabitants. The absence of the troops on the left bank of the Segre was soon detected by Garcia Conde, who—to his great regret three days later—sent messengers to inform O’Donnell that the French lines to the east of the city were now almost unoccupied.

The fiery O’Donnell was already on his way when this news reached him. The moment that he was sure that Augereau had left Barcelona and was retreating northward, he had resolved to fly to the relief of Lerida with the two divisions of which he could dispose. On the twentieth he marched from Tarragona by Valls and Momblanch, with 7,000 regular infantry, 400 horse, a single battery, and some 1,500 miqueletes. On the twenty-third he had reached Juneda, twelve miles from Lerida, without having seen a French picket or having been detected himself. He intended to drive off the few battalions blocking the bridge-head, and to open up communication with the besieged town that same day. Unfortunately for him, Suchet and Musnier had returned on the previous night from their expedition to Tarrega, and were now lying at Alcoletge, three miles north of the bridge-head, with seven battalions and 500 cuirassiers.

Pushing briskly forward along the high-road, in a level treeless plateau, destitute of cover of any kind, the leading Spanish division—that of Ibarrola—came into contact about midday with the small force guarding the bridge-head, which was now in charge of Harispe, Suchet’s chief-of-the-staff. To the surprise of the Spaniards, who had expected to see these three battalions and two squadrons of hussars retire in haste, the French showed[p. 304] fight. Harispe was aware that Musnier could come up to his aid within an hour, and was ready for battle. His hussars charged the leading regiment of Ibarrola’s column, and threw it into disorder, while his infantry formed up in support. At this moment the Spanish general received news that Musnier was moving in towards his flank; he therefore retired, to get the support of O’Donnell, who was following him with the second division. But at the ruined village of Margalef, six miles from Lerida, the column of Musnier came up with Ibarrola, and he was so hard pressed that he formed line, with 300 cavalry on his right wing and his three guns on his left. The position was as bad a one as could be found for the Spanish division—a dead flat, with no cover. Musnier, on reaching the front, flung his cavalry, 500 cuirassiers of the 13th regiment, at the Spanish right wing. The horsemen placed there broke, and fled without crossing sabres, and the French charge fell on the flank battalion of the infantry, which was caught while vainly trying to form square. It was ridden down in a moment, and the horsemen then rolled up the whole line, regiment after regiment. A great part of Ibarrola’s corps was captured, and O’Donnell, who came on the field with the division of Pirez just as the disaster took place, could do no more than retire in good order, covering the scattered remnants of his front line. A Swiss battalion, which he told off as his rearguard, was pierced by the cuirassiers and for the most part captured. In this disastrous affair the Spaniards lost 500 killed and wounded and several thousand prisoners[356], as also four flags and the half-battery belonging to Ibarrola’s division. The French, of whom the cavalry alone were seriously engaged, are[p. 305] said by Suchet to have had no more than 23 killed and 82 wounded[357].

It is clear that O’Donnell must take the blame for the ruin that fell upon his little field army. He should not have been caught with his two divisions marching with an interval of four miles between them. Nor ought he to have been ignorant of the return of Musnier’s force, considering that he had 400 cavalry, who should have been exploring the whole country-side for miles around, instead of riding in a mass along with Ibarrola’s infantry. The carelessness shown was unpardonable: relying, apparently, on Garcia Conde’s dispatch—now two days old—concerning the weakness of the French beyond the Segre, the Spanish general was caught moving as if for an unopposed entry into Lerida, instead of in battle order. It may be added that he would have done well to collect more men before advancing against Suchet—8,000 bayonets and sabres were too small a force to tackle the main body of the 3rd Corps, even if they were aided by a sortie from Lerida, as O’Donnell had intended. It would have been well to make some endeavour to get troops from the Valencian army, whose 12,000 men were absolutely idle at the moment. It is true that José Caro and the Valencian Junta were very chary of sending their men outside their own border. But for such a great affair as the relief of Lerida a brigade or two might have been borrowed. There was a considerable Valencian force, at the moment, in the neighbourhood of Alca?iz.

After the combat of Margalef Suchet summoned Lerida for a second time, and offered to allow the Governor to see the prisoners and guns which he had taken from O’Donnell. Garcia Conde very rightly answered that he relied on his own forces alone, and should fight to the end. Accordingly, the regular siege began. The trenches were opened opposite the north front of the town on April 29; several days of heavy rain hindered the completion of the first parallel, but on May 7 the breaching batteries were ready and the bombardment began. The front attacked—the Carmen and Magdalena bastions—was weak: it was not protected by any flanking fires, and had neither a ditch[p. 306] nor a covered way. It was bound to succumb before the very heavy fire directed against it, unless the defence should succeed in beating down the fire of the breaching batteries. The Spaniards did their best, bringing up every gun that could be mounted, and replacing each injured piece as it was disabled. But the end was obvious from the first: the walls were not strong enough to resist the attack.

Meanwhile, Suchet made two successive assaults on a part of the defences very far distant from the main front of attack—the two isolated redoubts which stood on the south end of the extramural plateau, of which Fort Garden formed the main protection. He wished to gain a footing on this high ground, both because he could from thence molest the south front of the town, and because he wished to prevent the Spaniards from using the plateau as a place of refuge after the fall of the city, which he regarded as inevitable. The first attempt to take the works by escalade, on the night of April 23-24, was a disastrous failure: the Pilar was occupied for a moment, but the attack on San Fernando failed, and the dominating fire from it drove the stormers out of the smaller work, after they had held it for a few hours.

The second escalade was more successful: it was carried out on the night of May 12-13, and ended in the storming of both works. Only a small part of their garrison succeeded in escaping into Fort Garden: the besieged lost 300 men, the successful assailants only 120. The greater part of the plateau was now in the hands of the French.

On the next day, May 13, the engineers announced that the two breaches in the north front of the town walls were practicable, and that same evening Lerida was stormed. The breaches were carried with no great difficulty, but the garrison made a stubborn resistance for some time, behind traverses and fortified houses in the rear of them. When these were carried, the city was at Suchet’s mercy; but it was not at the city alone that he was aiming—he wished to master the citadel also. During the last siege of Lerida, that by the Duke of Orleans in 1707, the high-lying castle had held out for many days after the town had been lost. Suchet’s way of securing his end was effective but brutal. On the whole, it was the greatest atrocity perpetrated by any[p. 307] combatant, French, Spanish, or English, during the whole Peninsular War. When his troops had entered the streets, he directed columns towards each gate, and having secured possession of them all, so as to make escape into the open country impossible, bade his troops push the whole non-combatant population of Lerida uphill into the citadel, where the beaten garrison was already taking refuge. ‘The soldiery,’ as he writes, with evident complacency and pride in his ingenuity, ‘were set in a concentric movement to push the inhabitants, along with the garrison, towards the upper streets and the citadel. They were dislodged by musketry fire from street after street, house after house, in order to force them into the castle. That work was still firing, and its discharges augmented the danger and the panic of the civil population, as they were thrust, along with the wreck of the garrison, into the ditch and over the drawbridge. Pressed on by our soldiers, they hastily poured into the castle yard, before the Governor had time to order that they should not be allowed to enter.’ The castle being crammed with some fifteen thousand men, women, and children, Suchet gave orders to bombard it with every available mortar and howitzer. ‘Every shell,’ he writes, ‘that fell into the narrow space containing this multitude, fell on serried masses of non-combatants no less than of soldiery. It had been calculated that the Governor and the most determined officers would be influenced by the presence of these women, children, old folk, and unarmed peasants. As General Suchet had flattered himself would be the case, the scheme had a prompt and decisive effect.’ On the 14th at midday, Garcia Conde, unable to stand the slaughter any longer, hoisted the white flag.

It is difficult to see how the forcing of thousands of non-combatants, by means of musketry fire, on to the front of the enemy’s line of defence, differs in any way from the device, not unknown among African savages and Red Indians[358], of attacking under cover of captured women and children thrust in upon the weapons of their fathers and husbands. The act places that polished writer and able administrator Louis-Gabriel Suchet on the moral level of a king of Dahomey. He acknowledges that the plan was deliberately thought out, and that scores of his[p. 308] victims perished not in the subsequent bombardment, but by being shot down by his own men while the crowd was being collected and hunted forward[359]. Historians have denounced the atrocities committed by the French rank and file at Tarragona or Oporto, by the English rank and file at Badajoz or St. Sebastian, but the cultured general who worked out this most effective plan for the reduction of a hostile citadel has never had his due meed of shame. Napier’s remark that, ‘though a town taken by assault is considered the lawful prey of a licentious soldiery, yet this remnant of barbarism does not warrant the driving of unarmed and helpless people into a situation where they must perish,’ seems a sufficiently mild censure, when all the circumstances are taken into account[360].

The total number of Spanish troops surrendered by Garcia Conde in the Citadel and Fort Garden, or captured by the French during the storm, amounted to over 7,000 men, of whom about 800 were wounded lying in the hospital. It was calculated that 1,200 or 1,500 more had perished during the siege, and that about 500 of the civil population had fallen victims to Suchet’s barbarous device. The French losses during the whole series of operations had been 1,100 killed and wounded. The Spaniards declared that Garcia Conde had betrayed Lerida, and ought never to have surrendered. But the only ground for this accusation was that, after a short captivity, he did homage to King Joseph, and became an Afrancesado. Though he had shown dash and courage at Gerona, it is clear that he lacked the firmness of governors such as Mariano Alvarez, or Andrés Herrasti. His defence of Lerida had not been particularly skilful nor particularly resolute: with over 8,000 men within the walls he ought to have been able to hold out longer against Suchet’s 13,000. The most blameworthy part of his arrangements was his neglect to retrench the breaches, and to form a strong second line of temporary works behind them. Alvarez, under similar conditions, and with a garrison far less strong in comparison to the besieging army, held out for months after his walls had been breached, simply because he treated them only as an outer line of resistance, and was prepared to fight on behind his inner defences.

[p. 309]

After the fall of Lerida, Suchet, having received through France the news that the whole of Augereau’s army was collected in the neighbourhood of Hostalrich and Gerona, made up his mind that he had better attempt nothing ambitious until the 7th Corps was in a position to help him. But there was a small task close to his hand, which was well worth undertaking in a moment of enforced leisure. This was the siege of Mequinenza, the only fortress left in Spanish hands on the eastern side of Aragon. It was a small place, but not without its strategic importance, for not only did it cover the junction of the Segre and the Ebro, but it was the highest point open for navigation on the last-named river. Any one holding Mequinenza can use the Ebro as a high-road, except in times of drought, and this was an advantage of no small importance, since the country from thence to the sea chances to be singularly destitute of roads of any kind. A few months later, Suchet found the place invaluable, as he was able to prepare his battering-train for the siege of Tortosa within its walls, and then to send all the heavy material down-stream with the minimum of trouble. Mequinenza was a small place, consisting of a few hundred houses along the river bank enclosed by a weak and old-fashioned wall, but dominated by a strong castle, which towers 500 feet above the water at the end of a spur of the Sierra de Montenegre. The garrison consisted of about 1,000 men, under a Colonel Carbon.

The very day after Lerida had fallen, Suchet sent a brigade to invest Mequinenza: more troops followed after an interval. It was clear that the taking of the town would offer small difficulty: but the castle was another matter. Its defences were independent of those of the place below, and its site was so lofty and rocky that it could not be battered from any ground accessible by existing roads. Indeed, it could only be approached along the crest of the Sierra, of which it forms the last lofty point. The main interest of the siege of Mequinenza lies in the fact that, in order to reduce the castle, the French engineers had to build a road practicable for heavy guns in zig-zags up the side of the Sierra. They had arrived in front of the place on May 15: by June 1 the road was completed: it was a piece of hard work, as in order to utilize the easiest possible slopes it had been made no less than five miles long. But when the guns had once been[p. 310] got up on to the crest of the mountain, Mequinenza—town and castle alike—was doomed. The town was stormed on June 5; it might have been captured long before, but there was little use in taking possession of it until the attack upon the castle had been begun. Before the storm, Colonel Carbon had wisely ordered all the large river craft in the place, eleven in number, to run down-stream to Tortosa; he was well aware that Suchet wanted to open up the navigation of the Ebro, and was resolved that he should find no vessels ready for him. Two of the craft ran ashore and were captured; the other nine got off clear.

When the three batteries erected on the summit of the Sierra opened on the castle, the walls began to crumble at once. At the end of eight days its front towards the French trenches was a mass of shapeless ruins. Carbon then surrendered, without waiting for an assault, which must undoubtedly have proved successful. There was nothing particularly obstinate, or, on the other hand, particularly discreditable, in the defence. The castle was not a modern fortress; its sole strength had lain in its inaccessible position; and when the French had climbed up on to a level with it, nothing more could be done.

Having mastered Lerida and Mequinenza, and obtained a firm footing in the plain of Western Catalonia, Suchet had now harder work before him. His first necessity was to clear the country behind him of the insurgents, who had swarmed down from the hills to attack the troops left behind in Central Aragon, while the main army had been concentrated before Lerida. Villacampa had half destroyed a column of 350 men at Arandija on the Xalon on May 14. Catalan somatenes had attacked the garrison of the valley of Venasque, on the very frontier of France, on May 16. Valencian bands had besieged the castle of Alca?iz for many days. But all the enemies of Suchet had to fly, as soon as his main army became once more free for field service. No attempt was made to oppose a serious resistance to his movable columns: the insurgents fled to right and left: the most extraordinary proof of their demoralization was that, on June 13, General Montmarie, at the head of one of these columns, found the strong fort of Morella, within the Valencian border, absolutely unoccupied. He seized and garrisoned it, knowing that this place, which commands the[p. 311] mountain-road from Aragon to Valencia, was of immense strategical importance.

When June came round, Suchet had once more mastered all Aragon, and was free for work outside its borders. A march against Valencia, on the one side, or against Tortosa and Tarragona on the other, was equally possible. But at this moment Napoleon was set on proceeding ‘methodically’—in Eastern Spain no less than in Portugal—and the orders issued to the Commander of the 3rd Corps were, to proceed against Tortosa, and cut off Valencia from Catalonia by its capture. The longer and more difficult advance against Valencia itself was relegated to a distant date; there must be no more fiascos like that of February. Accordingly, Suchet was ordered to follow up the capture of Lerida and Mequinenza by the reduction of Tortosa[361]. He was informed at the same time that Macdonald, the new commander of the Army of Catalonia, was ordered to attack Tarragona. In this way, the restless O’Donnell would find his hands full at home, and would not be able to spare time or men for the relief of Tortosa. Only from the Valencian Army, which had always been badly led, and managed with the most narrow particularism by José Caro and his Junta, need Suchet expect any trouble.

But until Macdonald was prepared to lead down his corps to the neighbourhood of Tarragona and the lower Ebro, Suchet felt that his own enterprise must be held back, or at least conducted with caution. And although Macdonald had received his orders in May, on taking over Augereau’s post, it was some time before he was able to execute them. June and July had passed, and August had arrived, before he appeared in the regions where Suchet had been told to expect him, and at last placed himself in communication with the Army of Aragon. This long delay was entirely due to the burden imposed upon him by the necessities of Barcelona, with its large garrison and its vast civil population. The provisions which Augereau had introduced into the city in the early spring were exhausted, and his successor was forced to replace them by escorting thither two new convoys of great size in June and July. So active were the miqueletes that Macdonald found himself obliged to[p. 312] take with him his French and his Italian division as escort on each occasion. For a first-class convoy is a vast affair, stretching out to long miles in a mountain country, and requiring to be strongly guarded at every point of its unwieldy length. The Emperor was bitterly disappointed at Macdonald’s delays: it was hard to convince him that the problem of food-supply in Catalonia was almost insoluble, and that neither Barcelona nor the field army could be permanently fed by seizing the resources of the valley of the Llobregat, or the Campo of Tarragona, and making war support war in the regular style. Catalonia had never fed its whole population even in time of peace, but had largely depended on corn brought up from Aragon or imported by sea. But the British Mediterranean squadron now made the transport of food to Barcelona by water impossible, while, till the 3rd Corps and the 7th Corps should finally meet, it was impossible to supply the latter from the Aragonese base. While neither of the normal ways of procuring a regular supply of provisions was practicable, nothing remained but the weary task of escorting convoys from Perpignan and Figueras to Barcelona. This Macdonald accomplished thrice, in June, July, and August: it was only after the third journey that he left the storeshouses of Barcelona so thoroughly replenished that he was able to think of further offensive operations, and to concert the long-delayed junction with Suchet and the 3rd Corps. The Marshal in his autobiography confesses that he hated the task set him, and found it entirely out of the lines of his experience. He began his career by repealing Augereau’s hanging edicts, and endeavoured to introduce more humane methods of warfare. But his own men were out of hand, and the Spaniards being as fierce as ever, his task was a very hard one.

Meanwhile, Henry O’Donnell was granted three months in which to reorganize the Catalan army, which had suffered so severely at Margalef. His untiring energy enabled him to raise his force by the end of July to 22,000 men, organized in five divisions, with which he formed a double front, facing Suchet on the west and Macdonald on the east. One division under Campo Verde lay northward in the mountains, with head quarters at Cardona, and detachments pushed forward as far as[p. 313] Urgel and Olot: its main task was to harass the French garrison of the Ampurdam, and to threaten Macdonald’s rear every time that he moved forward towards Barcelona. A second division lay facing that city, on the lower Llobregat, with detachments at Montserrat and Manresa, which kept up the communications with Campo Verde and the North. Two more divisions lay westward, in the direction of Falcet and Borja Blanca, watching Suchet, and prepared to oppose any serious attempt on his part to close in upon Tortosa. The fifth division, or reserve, formed a central mass of troops, ready to reinforce either the detachments which watched Suchet or those which watched Macdonald; it lay south of Tarragona, at the Col de Alba. Thus posted O’Donnell waited for further developments, and continued to drill and exercise his new levies, to dismiss inactive and incompetent officers, to collect magazines, and to quarrel with the local Junta. For this active and intelligent, if reckless and high-handed, young officer was on even worse terms with the Catalan authorities than his predecessor, Blake. The main cause of quarrel was that O’Donnell wished to strengthen his regular troops, by drafting into the depleted cadres of Reding’s and Coupigny’s old battalions every recruit that he could catch. He held that the French could only be brought to a final check by disciplined troops. The Junta, on the other hand, believed in guerrilla warfare, and preferred to call out from time to time the miqueletes and somatenes, who drifted back to their homes whenever a crisis was over. Both sides had much to say for themselves: the Catalans could point out that the regular troops had been beaten times out of number, while their own irregulars had achieved many small successes, and done the French much harm. O’Donnell, on the other hand, was quite right in holding that guerrilla operations, much as they might incommode the enemy, would never deal him a fatal blow or finish the war. He dreamed of recruiting his army up to a force of 50,000 men at some happy future date, and of delivering a stroke with superior numbers which should destroy the French hold on Catalonia. Meanwhile, much as they quarrelled, the Irish-Spanish general and the local Junta were both too good haters of France to allow their disputes to prevent a vigorous resistance from[p. 314] being kept up in the principality. O’Donnell carried his point to the extent of persuading the Junta to allow him to form many of the ever-changing tercios of miqueletes into ‘legions,’ which were to be kept permanently embodied[362], and to count as part of the regular army. Hence came the large rise in his muster-rolls.