vii. 377); they are to the effect that Garcia Conde had intended

to start _at dusk_ on the third, but, hearing firing on the side of Palau, deferred his exit and took another road. If he was starting at 7 or 8 o’clock at night on the third, he cannot have been warned by the morning bugles at 4 o’clock on the previous morning. See St. Cyr, p. 234, and Napier, ii. 245, for the French story, which the latter takes over whole from the former. Belmas and Vacani do not give the tale, though they have a full narration of the escape of Garcia Conde.

Before he could consider his position safe, Verdier had to complete the lines of investment: this he did on September 5 by driving off the intermediate posts which Alvarez had thrown out from the Capuchin heights, to link the town with the garrison in the hermitage of Nuestra Señora de los Angeles. Mazzuchelli’s brigade stormed the hermitage itself on the following day, with a loss of about eighty men, and massacred the greater part of the garrison. On that same day, however, the French suffered a small disaster in another part of the environs. General Joba, who had been sent with three battalions to clear the road to Figueras from the bands of Claros and Rovira, was beaten and slain at San Gregorio by those chiefs. But the miqueletes afterwards retired to the mountains, and the road became intermittently passable, at least for large bodies of men.

It was not till September 11, however, that Verdier recommenced the actual siege, and bade his batteries open once more upon Gerona. The eleven days of respite since Blake interrupted the bombardment on September 1 had been invaluable to the garrison, who had cleared away the débris from the foot of the breaches, replaced the damaged artillery on the front of attack, and thrown up interior defences behind the shattered parts of the wall. They had also destroyed all the advanced trenches of the besiegers, which had to be reconstructed at much cost of life. In four days Verdier had recovered most of the lost ground, when he was surprised by a vigorous sally from the gate of San Pedro: the garrison, dashing out at three p.m., stormed the three nearest breaching batteries, spiked their guns, and filled in all the trenches which were advancing towards the foot of the walls. Four days’ work was thus undone in an hour, and it was only on September 19 that Verdier had reconstructed his works, and pushed forward so far towards his objective that he considered an assault possible. He then begged St. Cyr to lend him a brigade of fresh troops, pleading that the siege-corps was now so weak in numbers, and so demoralized by its losses, that he did not consider that the men would do themselves justice at a storm. The losses of officers had been fearful: one battalion was commanded by a lieutenant, another had been reduced to fifty men; desertion was rampant among several of the foreign corps. Of 14,000 infantry[48] of the French, Westphalian, and Italian divisions less than 6,000 now remained. So far as mere siegecraft went, as he explained to St. Cyr, ‘the affair might be considered at an end. We have made four large practicable breaches, each of them sufficient to reduce the town. But the troops cannot be trusted.’ St. Cyr refused to lend a man for the assault, writing with polite irony that ‘every general has his own task: yours is to take Gerona with the resources placed at your disposal by the government for that object, and the officers named by the government to conduct the siege[49].’ He added that he considered, from its past conduct, that the morale of the siege-corps was rather good than bad. He should not, therefore, allow the covering army to join the assault; but he would lend the whole of Pino’s division to take charge of Monjuich and the camps, during the storm, and would make a demonstration against the Mercadal, to distract the enemy from the breaches. With this Verdier had to be content, and, after making two final protests, concentrated all his brigades save those of the Westphalian division, and composed with them four columns, amounting to some 3,000 men, directing one against each of the four breaches. That sent against the platform of San Cristobal was a mere demonstration of 150 men, but the other three were heavy masses: the Italians went against Santa Lucia, the French brigade against the southern breach in the ‘Redoubt of the Germans,’ the Berg troops against the northern one. A separate demonstration was made against the Calvary fort, whose unsubdued fire still flanked the breaches, in the hope that its defenders might be prevented from interfering in the main struggle.

[48] Verdier did not exaggerate: see Appendix no. 2 at end of this volume, showing that his three divisions had lost 8,161 men out of 14,044 by September 15.

[49] See the acrid correspondence between St. Cyr and Verdier in Appendices nos. 37-8, 40-6 of Belmas, vol. ii.

Alvarez, who had noted the French columns marching from all quarters to take shelter, before the assault, in the trenches on the slopes of Monjuich and in front of Pedret, had fair warning of what was coming, and had done his best to provide against the danger. The less important parts of the enceinte had been put in charge of the citizens of the ‘Crusade,’ and the picked companies of every regiment had been told off the breaches. The Englishman, Ralph Marshall, was in charge of the curtain of Santa Lucia, William Nash, the Spanish-Irish colonel of Ultonia, commanded at the two breaches under La Gironella: Brigadier Fournas, the second-in-command of the garrison, had general supervision of the defences; he had previously taken charge of Monjuich during the great assault in August. Everything had been done to prepare a second line of resistance behind the breaches; barricades had been erected, houses loopholed, and a great many marksmen disposed on roofs and church towers, which looked down on the rear-side of the gaps in the wall.

At four o’clock in the afternoon of September 19 the three columns destined for the northern breaches descended from Monjuich on the side of San Daniel, crossed the Galligan, and plunged into the hollow at the foot of the ‘Redoubt of the Germans.’ At the same moment the fourth column started from the ruins of the tower of San Juan to attack the curtain of Santa Lucia. The diversion against the Calvary fort was made at the same moment, and beaten off in a few minutes, so that the fire of this work was not neutralized during the assault according to Verdier’s expectation. The main assault, nevertheless, was delivered with great energy, despite the flanking fire. At the two points of attack under La Gironella the stormers twice won, crossed, and descended from the breach, forcing their way into the ruined barracks behind. But they were mown down by the terrible musketry fire from the houses, and finally expelled with the bayonet. At the Santa Lucia curtain the Italians scaled the breach, but were brought up by a perpendicular drop of twelve feet behind it--the foot of the wall in this quarter chancing to be much higher than the level of the street below. They held the crest of the breach for some time, but were finally worsted in a long and furious exchange of fire with the Spaniards on the roofs and churches before them, and recoiled. The few surviving officers rallied the stormers, and brought them up for a second assault, but at the end of two hours of hard fighting all were constrained to retire to their trenches. They had lost 624 killed and wounded, including three colonels (the only three surviving in the whole of Verdier’s corps) and thirty other officers. The Spanish loss had been 251, among them Colonel Marshall, who was mortally wounded at his post on the Santa Lucia front.

Verdier accused his troops of cowardice, which seems to have been unjust. St. Cyr wrote to the Minister of War to express his opinion that his subordinate was making an excuse to cover his own error, in judging that a town must fall merely because there were large breaches in its walls[50]. ‘The columns stopped for ninety minutes on the breaches under as heavy a fire as has ever been seen. There was some disorder at the end, but that is not astonishing in view of the heavy loss suffered before the retreat. I do not think that picked grenadiers would have done any better, and I am convinced that the assault failed because the obstacles to surmount were too great.’ The fact was that the Spaniards had fought with such admirable obstinacy, and had so well arranged their inner defences, that it did not suffice that the breaches should have been perfectly practicable. At the northern assault the stormers actually penetrated into the buildings behind the gaps in the ruined wall, but could not get further forward[51]. In short, the history of the siege of Gerona gives a clear corroboration of the old military axiom that no town should ever surrender merely because it has been breached, and justifies Napoleon’s order that every governor who capitulated without having stood at least one assault should be sent before a court martial. It refutes the excuses of the too numerous commanders who have surrendered merely because there was a practicable breach in their walls, like Imaz at Badajoz in 1811. If all Spanish generals had been as wary and as resolute as Mariano Alvarez, the Peninsular War would have taken some unexpected turns. The moral of the defences of Tarifa, Burgos, and San Sebastian will be found to be the same as that of the defence of Gerona.

[50] ‘Il paraît que l’on a employé la ressource, malheureusement trop usitée en pareil cas, de dire que les troupes n’ont pas fait leur devoir, ce qui produit de justes réclamations de leur part.’ (St. Cyr to the Minister, Sept. 24, 1809.)

[51] The not unnatural suggestion that the German and Italian troops may have failed to display such desperate courage as the native French in the assault seems to be refuted by their losses, which were hardly smaller in proportion. Of 1,430 native French of the 7th and 56th Line and 32nd Léger, 328 were put out of action; of 1,400 Berg, Würzburg, and Italian troops, 296. The difference in the percentage is so small that it is clear that there was no great difference in conduct.

The effect of the repulse of September 19 on the besieging army was appalling. Verdier, after writing three venomous letters to the Emperor, the War Minister, and Marshal Augereau[52], in which he accused St. Cyr of having deliberately sacrificed the good of the service to his personal resentments, declared himself invalided. He then went off to Perpignan, though permission to depart was expressly denied him by his superior: his divisional generals, Lecchi and Morio, had already preceded him to France. Disgust at the failure of the storm had the same effect on the rank and file: 1,200 men went to the hospital in the fortnight that followed the assault, till by October 1 the three divisions of the siege-corps numbered little more than 4,000 bayonets--just enough to hold Monjuich and the camps by the great dépôts at Pont Mayor and Sarria. The store of ammunition in the park had been used up for the tremendous bombardment poured upon the breaches from the 15th to the 17th of September. A new supply was wanted from Perpignan, yet no troops could be detached to bring it forward, for the miqueletes were again active, and on September 13 had captured or destroyed near Bascara a convoy guarded by so many as 500 men.

[52] See especially Verdier to Augereau, no. 53, and to the Minister, no. 61, of Belmas’s Appendices.

St. Cyr, left in sole charge of the siege by Verdier’s departure, came to the conclusion that it was useless to proceed with the attack by means of trenches, batteries, and assaults, and frankly stated that he should starve the town out, but waste no further lives on active operations. He drew in the covering corps closer to Gerona, so that it could take a practical part in the investment, put the wrecks of Lecchi’s troops--of whom less than 1,000 survived--into Pino’s division, and sent the French brigade of Verdier’s old division to guard the line between Bascara and the Frontier. Thus the distinction between the siege-corps and the covering troops ceased to exist, and St. Cyr lay with some 16,000 men in a loose circle round Gerona, intent not on prosecuting advances against the walls, but only on preventing the introduction of further succours. He was aware that acute privations were already being suffered by the Spaniards: Garcia Conde’s convoy had brought in not much more than eight days’ provisions for the 5,000 men of the reinforced garrison and the 10,000 inhabitants who still survived. There was a considerable amount of flour still left in store, but little else: meat, salt and fresh, was all gone save horseflesh, for Alvarez had just begun to butcher his draught horses and those of his single squadron of cavalry. There was some small store of chocolate, tobacco, and coffee, but wine and aguardiente had run out, so had salt, oil, rice, and--what was most serious with autumn and winter approaching--wood and charcoal. All the timbers of the houses destroyed by the bombardment had been promptly used up, either for fortification or for cooking[53]. Medical stores were wholly unobtainable: the chief hospital had been burnt early in the siege, and the sick and wounded, laid in vaults or casemates for safety, died off like flies in the underground air. The seeds of pestilence were spread by the number of dead bodies of men and animals which were lying where they could not be reached, under the ruins of fallen houses. The spirit alike of garrison and troops still ran high: the repulse of the great assault of September 19, and the cessation of the bombardment for many days after had encouraged them. But they were beginning to murmur more and more bitterly against Blake: there was a general, if erroneous, opinion that he ought to have risked a battle, instead of merely throwing in provisions, on September 1. Alvarez himself shared this view, and wrote in vigorous terms to the Junta of Catalonia, to ask if his garrison was to perish slowly by famine.

[53] The very interesting list of the prices of commodities at the commencement and the end of the siege, drawn up by Dr. Ruiz, one of the Gerona diarists, may be found on p. 579 of Arteche’s vol. vii. Note the following--the real (20 to the dollar) = 2½_d._ :--

_Reals._ _In June._ _In Sept._

Wheat flour, the qr. 80 112 Barleymeal, the qr. 30 56 Oatmeal, the qr. 48 80 Coffee, the lb. 8 24 Chocolate, the lb. 16 64 Oil, the measure 2½ 24 Salt fish, the lb. 2¼ 32 Cheese, the lb. 4 40 Wood, the arroba (32 lb.) 5 48 Charcoal, ditto 3½ 40 Tobacco, the lb. 24 100 A fowl 14 320 Rice, the lb. 1½ 32 Fresh fish from the Ter, the lb. 4 36

Thus while flour and meal had not doubled in value, coffee had gone up threefold, chocolate and tobacco fourfold, cheese and fuel tenfold, and the other commodities far more.

Blake responded by a second effort, less happily planned than that of September 1. He called together his scattered divisions, now about 12,000 strong, and secretly concentrated them at La Bispal, between Gerona and the sea. He had again got together some 1,200 mules laden with foodstuffs, and a large drove of sheep and oxen. Henry O’Donnell, an officer of the Ultonia regiment, who had been sent out by Alvarez, marched at the head of the convoy with 2,000 picked men; a division of 4,000 men under General Wimpfen followed close behind to cover its rear. Blake, with the rest, remained at La Bispal: he committed the egregious fault of omitting to threaten other parts of the line of investment, so as to draw off St. Cyr’s attention from the crucial point. He trusted to secrecy and sudden action, having succeeded in concentrating his army without being discovered by the French, who thought him still far away beyond Hostalrich. Thus it came to pass that though O’Donnell struck sharply in, defeated an Italian regiment near Castellar, and another three miles further on, and reached the Constable fort with the head of the convoy, yet the rest of Pino’s division and part of Souham’s concentrated upon his flank and rear, because they were not drawn off by alarms in other quarters. They broke in between O’Donnell and his supports, captured all the convoy save 170 mules, and destroyed the leading regiment of Wimpfen’s column, shooting also, according to the Spanish reports, many scores of the unarmed peasants who were driving the beasts of burden[54]. About 700 of Wimpfen’s men were taken prisoners, about 1,300 killed or wounded, for little quarter was given. The remnant recoiled upon Blake, who fell back to Hostalrich next day, September 27, without offering to fight. The amount of food which reached the garrison was trifling, and Alvarez declared that he had no need for the additional mouths of O’Donnell’s four battalions, and refused to admit them into the city. They lay encamped under the Capuchin fort for some days, waiting for an opportunity to escape.

[54] See Toreno, ii, and Arteche, vii. 412.

After having thus wrecked Blake’s second attempt to succour Gerona, and driven him from the neighbourhood, St. Cyr betook himself to Perpignan, in order, as he explained to the Minister of War[55], to hurry up provisions to the army at the front, and to compel the officers at the base to send forward some 3,000 or 4,000 convalescents fit to march, whose services had been persistently denied him[56]. Arrived there he heard that Augereau, whose gout had long disappeared, was perfectly fit to take the field, and could have done so long before if he had not preferred to shift on to other shoulders the responsibility for the siege of Gerona. He was, on October 1, at the baths of Molitg, ‘destroying the germs of his malady’ as he gravely wrote to Paris,--amusing himself, as St. Cyr maintains in his memoirs. Convinced that the siege had still a long time to run, and eager to do an ill turn to the officer who had intrigued to get his place, St. Cyr played on the Marshal precisely the same trick that Verdier had played on himself a fortnight before. He announced that he was indisposed, wrote to congratulate Augereau on his convalescence, and to resign the command to his hands, and departed to his home, without waiting for an answer, or obtaining leave from Paris--a daring act, as Napoleon was enraged, and might have treated him hardly. He was indeed put under arrest for a short time.

[55] See St. Cyr to the Minister, Belmas, ii. Appendix no. 67.

[56] Augereau to the Minister, ibid., Oct. 8.

From the first to the eleventh of October Souham remained in charge of the army, but on the twelfth Augereau appeared and took command, bringing with him the mass of convalescents who had been lingering at Perpignan. Among them was Verdier, whose health became all that could be desired when St. Cyr had disappeared. The night following the Marshal’s arrival was disturbed by an exciting incident. Henry O’Donnell from his refuge on the Capuchin heights, had been watching for a fortnight for a good chance of escape. There was a dense fog on the night of the 12th-13th: taking advantage of it O’Donnell came down with his brigade, made a circuit round the town, crossed the Oña and struck straight away into the plain of Salt, which, being the most open and exposed, was also the least guarded section of the French lines of investment. He broke through the chain of vedettes almost without firing, and came rushing before dawn into Souham’s head-quarters camp on the heights of Aguaviva. The battalion sleeping there was scattered, and the general forced to fly in his shirt. O’Donnell swept off his riding-horses and baggage, as also some prisoners, and was out of reach in half an hour, before the rallying fractions of the French division came up to the rescue of their chief. By six o’clock the escaping column was in safety in the mountains by Santa Coloma, where it joined the miqueletes of Milans. For this daring exploit O’Donnell was made a major-general by the Supreme Junta. His departure was a great relief to Alvarez, who had to husband every mouthful of food, and had already put both the garrison and the townsfolk on half-rations of flour and horseflesh.

Augereau was in every way inferior as an officer to St. Cyr. An old soldier of fortune risen from the ranks, he had little education or military science; his one virtue was headlong courage on the battlefield, yet when placed in supreme command he often hesitated, and showed hopeless indecision. He had been lucky enough to earn a great reputation as Napoleon’s second-in-command in the old campaigns of Italy in 1796-7. Since then he had made his fortune by becoming one of the Emperor’s most zealous tools and flatterers. He was reckoned a blind and reckless Bonapartist, ready to risk anything for his master, but spoilt his reputation for sincerity by deserting him at the first opportunity in 1814. He was inclined to a harsh interpretation of the laws of war, and enjoyed a doubtful reputation for financial integrity. Yet he was prone to ridiculous self-laudatory proclamations and manifestos, written in a bombastic strain which he vainly imagined to resemble his master’s thunders of the _Bulletins_. Scraps of his address to the citizens of Gerona may serve to display his fatuity--

‘Unhappy inhabitants--wretched victims immolated to the caprice and madness of ambitious men greedy for your blood--return to your senses, open your eyes, consider the ills which surround you! With what tranquillity do your leaders look upon the graves crammed with your corpses! Are you not horror-struck at these cannibals, whose mirth bursts out in the midst of the human hecatomb, and who yet dare to lift their gory hands in prayer towards the throne of a God of Peace? They call themselves the apostles of Jesus Christ! Tremble, cruel and infamous men! The God who judges the actions of mortals is slow to condemn, but his vengeance is terrible.... I warn you for the last time, inhabitants of Gerona, reflect while you still may! If you force me to throw aside my usual mildness, your ruin is inevitable. I shall be the first to groan at it, but the laws of war impose on me the dire necessity.... I am severe but just. Unhappy Gerona! if thy defenders persist in their obstinacy, thou shalt perish in blood and flame.

(Signed) AUGEREAU.’

Stuff of this sort was not likely to have much effect on fanatics like Alvarez and his ‘Crusaders.’ If it is so wrong to cause the deaths of men--they had only to answer--Why has Bonaparte sent his legions into Spain? On the Marshal’s line of argument, that it is wrong to resist overwhelming force, it is apparently a sin before God for any man to attempt to defend his house and family against any bandit. There is much odious and hypocritical nonsense in some of Napoleon’s bulletins, where he grows tender on the miseries of the people he has conquered, but nothing to approach the maunderings of his copyist.

Augereau found the army about Gerona showing not more than 12,000 bayonets fit for the field--gunners and sappers excluded. The men were sick of the siege, and it would seem that the Marshal was forced, after inspecting the regiments and conferring with the generals, to acquiesce in St. Cyr’s decision that any further assaults would probably lead to more repulses. He gave out that he was resolved to change the system on which the operations had hitherto been conducted, but the change amounted to nothing more than that he ordered a slow but steady bombardment to be kept up, and occasionally vexed the Spaniards by demonstrations against the more exposed points of the wall. It does not appear that either of these expedients had the least effect in shaking the morale of the garrison. It is true that during October and November the hearts of the Geronese were commencing to grow sick, but this was solely the result of starvation and dwindling numbers. As to the bombardment, they were now hardened to any amount of dropping fire: on October 28 they celebrated the feast of San Narciso, their patron, by a procession all round the town, which was under fire for the whole time of its progress, and paid no attention to the casualties which it cost them.

Meanwhile, when the second half of October had begun, Blake made the third and last of his attempts to throw succours into Gerona. It was even more feebly carried out than that of September 26, for the army employed was less numerous. Blake’s force had not received any reinforcement to make up for the men lost in the last affair, a fact that seems surprising, since Valencia ought now to have been able to send him the remainder of the regiments which had been reorganized since the disasters of June. But it would seem that José Caro, who was in command in that province, and the local Junta, made excuses for retaining as many men as possible, and cared little for the danger of Gerona, so long as the war was kept far from their own frontier. It was, at any rate, with no more than 10,000 or 12,000 men, the remains of his original force, that Blake once more came forward on October 18, and threatened the blockading army by demonstrations both from the side of La Bispal and that of Santa Coloma. He had again collected a considerable amount of food at Hostalrich, but had not yet formed a convoy: apparently he was waiting to discover the weakest point in the French lines before risking his mules and his stores, both of which were by now very hard to procure. There followed a fortnight of confused skirmishing, without any battle, though Augereau tried with all his might to force on a general engagement. One of his Italian brigades was roughly handled near La Bispal on the twenty-first, and another repulsed near Santa Coloma on the twenty-sixth, but on each occasion, when the French reinforcements came up, Blake gave back and refused to fight. On November 1 the whole of Souham’s division marched on Santa Coloma, and forced Loygorri and Henry O’Donnell to evacuate it and retire to the mountains. Souham reported that he had inflicted a loss of 2,000 men on the Spaniards, at the cost of eleven killed and forty-three wounded on his own side! The real casualty list of the two Spanish divisions seems to have been somewhat over 100 men[57].

[57] See Souham’s dispatch, striving to make the combat into a very big business, in Belmas, ii, Appendix no. 72, and cf. Arteche, vii. pp. 430-1.

Nothing decisive had taken place up to November 7, when Augereau conceived the idea that he might make an end of Blake’s fruitless but vexatious demonstrations, by dealing a sudden blow at his magazines in Hostalrich. If these were destroyed it would cost the Spaniards much time to collect another store of provisions for Gerona. Accordingly Pino marched with three brigades to storm the town, which was protected only by a dilapidated mediaeval wall unfurnished with guns, though the castle which dominated it was a place of considerable strength, and proof against a _coup de main_. Only one of Blake’s divisions, that of Cuadrado, less than 2,000 strong, was in this quarter, and Augereau found employment for the others by sending some of Souham’s troops against them. The expedition succeeded: while Mazzuchelli’s brigade occupied the attention of Cuadrado, the rest of the Italians stormed Hostalrich, which was defended only by its own inhabitants and the small garrison of the castle. The Spaniards were driven up into that stronghold after a lively fight, and all the magazines fell into Pino’s hands and were burnt. At a cost of only thirty-five killed and sixty-four wounded the food, which Blake had collected with so much difficulty, was destroyed[58]. Thereupon the Spanish general gave up the attempt to succour Gerona, and withdrew to the plain of Vich, to recommence the Sisyphean task of getting together one more convoy. It was not destined to be of any use to Alvarez and his gallant garrison, for by the time that it was collected the siege had arrived at its final stage.

[58] See Pino’s and Augereau’s dispatches in Belmas’s Appendices, nos. 73 and 74.

The Geronese were now reaching the end of their strength: for the first time since the investment began in May some of the defenders began to show signs of slackening. The heavy rains of October and the commencement of the cold season were reducing alike troops and inhabitants to a desperate condition. They had long used up all their fuel, and found the chill of winter intolerable in their cellars and casemates. Alvarez, though reduced to a state of physical prostration by dysentery and fever, was still steadfast in heart. But there was discontent brewing among some of his subordinates: it is notable, as showing the spirit of the time, that the malcontents were found among the professional soldiers, not among the citizens. Early in November several officers were found holding secret conferences, and drawing up an address to the local Junta, setting forth the desperate state of the city and the necessity for deposing the governor, who was represented as incapacitated for command by reason of his illness: it was apparently hinted that he was going mad, or was intermittently delirious[59]. Some of the wild sayings attributed to Alvarez during the later days of the siege might be quoted as a support for their representations. To a captain who asked to what point he was expected to retire, if he were driven from his post, it is said that he answered, ‘to the cemetery.’ To another officer, the first who dared to say that capitulation was inevitable because of the exhaustion of the magazines, he replied, ‘When the last food is gone we will start eating the cowards, and we will begin with you.’ Though aware that their conspiracies were known, the malcontents did not desist from their efforts, and Alvarez made preparations for seizing and shooting the chiefs. But on the night of November 19 eight of them, including three lieutenant-colonels[60], warned by a traitor of their approaching fate, fled to Augereau’s camp. Their arrival was the most encouraging event for the French that had occurred since the commencement of the siege. They spoke freely of the exhaustion of the garrison, and said that Alvarez was mad and moribund.

[59] Alvarez’s letter to Blake of Nov. 3 printed in Arteche’s Appendix, no. 18 of his vol. vii, gives this account of the first discovery of plots.

[60] Of whom two, strangely enough, had been specially mentioned for courage at the September assault.

It was apparently this information concerning the desperate state of the garrison which induced Augereau to recommence active siege operations. He ordered up ammunition from Perpignan to fill the empty magazines, and when it arrived began to batter a new breach in the curtain of Santa Lucia. On December 2 Pino’s Italians stormed the suburb of La Marina, outside the southern end of the town, a quarter hitherto unassailed, and made a lodgement therein, as if to open a new point of attack. But this was only done to distract the enemy from the real design of the Marshal, which was nothing less than to cut off the forts on the Capuchin heights from Gerona by seizing the redoubts, those of the ‘Chapter’ and the ‘City,’ which covered the steep upward path from the walls to the group of works on the hilltop. At midnight on December 6 the voltigeur and grenadier companies of Pino’s division climbed the rough southern face of the Capuchin heights, and surprised and escaladed the ‘Redoubt of the City,’ putting the garrison to the sword. Next morning the batteries of the forts above and the city below opened a furious fire upon the lost redoubt, and Alvarez directed his last sally, sending out every man that he could collect to recover the work. This led to a long and bloody fight on the slopes, which ended most disastrously for the garrison. Not only was the sortie repulsed, but in the confusion the French carried the Calvary and Chapter redoubts, the other works which guarded the access from Gerona to the upper forts. On the afternoon of December 7 the communication with them was completely cut off, and as their garrisons possessed no separate magazines, and had been wont to receive their daily dole from the city, it was clear that they must be starved out. They had only food for forty-eight hours at the moment[61].

[61] Napier (ii. 249) says that the sortie was so far successful that the Geronese opened the way for the garrison of the Constable fort to escape into the city. But I can find no authority for this in either the French or the Spanish narratives, see especially Vacani.

The excitement of the sally had drained away the governor’s last strength: he took to his bed that evening, was in delirium next day, and on the morning of the ninth received the last sacraments of the Church, the doctors having declared that his hours were numbered. His last conscious act was to protest against any proposal to surrender, before he handed over the command to the senior officer present, General Juliano Bolivar. Had Alvarez retained his senses, it is certain that an attempt would have been made to hold the town, even when the starving garrisons of the forts should have surrendered. But the moment that his stern hand was removed, his successor, Bolivar, called together a council of war, to which the members of the Junta, no less than the officers commanding corps, were invited. They voted that further resistance was impossible, and sent out Brigadier-General Fournas, the man who had so well defended Monjuich, to obtain terms from Augereau. On the morning of the tenth the Marshal received him, and dictated a simple surrender, without any of the favourable conditions which Fournas at first demanded. His only concession was that he offered to exchange the garrison for an equal number of the unhappy prisoners from Dupont’s army, now lying in misery on the pontoons at Cadiz, if the Supreme Junta concurred. But the bargain was never ratified, as the authorities at Seville were obdurate.

On the morning of December 11 the survivors of the garrison marched out, and laid down their arms on the glacis of the Mercadal. Only 3,000 men came forth; these looked like living spectres, so pale, weak, and tattered that ‘the besiegers,’ as eye-witnesses observed, ‘felt ashamed to have been held at bay so long by dying men.’ There were 1,200 more lying in the hospitals. The rest of the 9,000 who had defended the place from May, or had entered with Garcia Conde in September, were dead. A detailed inspection of figures shows that of the 5,723 men of Alvarez’s original command only 2,008 survived, while of the 3,648 who had come later there were still 2,240 left: i. e. two-thirds of the old garrison and one-third of the succours had perished. The mortality by famine and disease far exceeded that by the sword: 800 men had died in the hospitals in October, and 1,300 in November, from mere exhaustion. The town was in a dreadful state: about 6,000 of the 14,000 inhabitants had perished, including nearly all the very young and the very old. 12,000 bombs and 8,000 shells had been thrown into the unhappy city: it presented a melancholy vista of houses roofless, or with one or two of the side-walls knocked in, of streets blocked by the fallen masonry of churches or towers, under which half-decayed corpses were partially buried. The open spaces were strewn with broken muskets, bloody rags, wheels of disabled guns and carts, fragments of shells, and the bones of horses and mules whose flesh had been eaten. The stench was so dreadful that Augereau had to keep his troops out of the place, lest infection should be bred among them. In the magazines nothing was found save a little unground corn; all the other provisions had been exhausted. There were also 168 cannon, mostly disabled; about 10,000 lb. of powder, and a million musket cartridges. The military chest handed over contained 562 reals--about 6_l._ sterling.

Augereau behaved very harshly to the garrison: many feeble or diseased men were made to march to Perpignan and perished by the way. The priests and monks of the ‘Crusade’ were informed that they were combatants, and sent off with the soldiery. But the fate of the gallant Governor provokes especial indignation. Alvarez did not die of his fever: when he was somewhat recovered he was forwarded to Perpignan, and from thence to Narbonne, where he was kept for some time and seemed convalescent. Orders then came from Paris that he was to be sent back to Spain--apparently to be tried as a traitor, for it was alleged that in the spring of 1808 he had accepted the provisional government installed by Murat. He was separated from his aide-de-camp and servants, and passed on from dungeon to dungeon till he reached Figueras. The day after his arrival at that place he was found dead, on a barrow--the only bed granted him--in the dirty cellar where he had been placed. It is probable that he perished from natural causes, but many Spaniards believed that he had been murdered[62].

[62] For details of this disgraceful cruelty, see Arteche’s ‘Elogio’ on Alvarez in the proceedings of the Madrid Academy. The Emperor Napoleon himself must bear the responsibility, as it was by orders from Paris that Alvarez was sent back from France to Figueras. Apparently he was to be tried at Barcelona, and perhaps executed. There is no allusion to the matter in the _Correspondance de Napoléon_.

Great as the losses of the garrison of Gerona had been, they were far exceeded, both positively and proportionately, by those of the besieging army. The French official returns show that on June 15 the three divisions charged with the attack, those of Verdier, Morio, and Lecchi, had 14,456 bayonets, and the two divisions of the covering army, those of Souham and Pino, 15,732: there were 2,637 artillerymen and engineers over and above these figures. On December 31, twenty days after the surrender, and when the regiments had been joined by most of their convalescents, the three siege-divisions counted 6,343 men, the covering divisions 11,666, and the artillery and engineers, 2,390.

This shows a loss of over 13,000 men; but on examination the deficit is seen to be even larger, for two new battalions from France had just joined Verdier’s division in December, and their 1,000 bayonets should be deducted from his total. It would seem, then, that the capture of Gerona cost the 7th Corps about 14,000 men, as well as a whole campaigning season, from April to December. The attack on Catalonia had been brought to a complete standstill, and when Gerona fell the French occupied nothing but the ruined city, the fortresses of Rosas and Figueras hard by the frontier, and the isolated Barcelona, where Duhesme, with the 6,000 men of his division, had been lying quiescent all the summer and autumn. Such a force was too weak to make detachments to aid St. Cyr or Augereau, since 4,000 men at least were needed for the garrison of the citadel and the outlying forts, and it would have been hopeless for the small remainder to take the field. Duhesme only conducted one short incursion to Villafranca during the siege of Gerona. In the last months of the year Barcelona was again in a state of partial starvation: the food brought in by Cosmao’s convoy in the spring had been exhausted, while a second provision-fleet from Toulon, escorted by five men-of-war, had been completely destroyed in October. Admiral Martin surprised it off Cape Creus, drove ashore and burnt two line-of-battle ships and a frigate, and captured most of the convoy. The rest took refuge in the harbour of Rosas, where Captain Halliwell attacked them with the boats of the squadron and burnt them all[63].

[63] For details, see James’s _Naval History_, v. pp. 142-5.

While Gerona was enduring its last month of starvation, those whose care it should have been to succour the place at all costs were indulging in a fruitless exchange of recriminations, and making preparations when it was all too late. Blake, after retiring to Vich on November 10, informed the Junta of Catalonia that he was helpless, unless more men could be found, and that they must find them. Why he did not rather insist that the Valencian reserves should be brought up, and risk stripping Tarragona and Lerida of their regular garrisons, it is hard to say. This at any rate would have been in his power. The Catalan Junta replied by summoning a congress at Manresa on November 20, to which representatives of every district of the principality were invited. The congress voted that a levy _en masse_ of all the able-bodied men from seventeen to forty-five years of age should be called out[64], and authorized a loan of 10,000,000 reals for equipping them. They also wrote to Seville, not for the first time, to demand reinforcements from the Central Junta. But the battle of Ocaña had just been fought and lost, and Andalusia could not have spared a man, even if there had been time to transport troops to Tarragona. All that the Catalans received was honorary votes of approval for the gallant behaviour of the Geronese. The levy _en masse_ was actually begun, but there was an insuperable difficulty in collecting and equipping the men in winter time, when days were short and roads were bad. The weeks passed by, and Gerona fell long before enough men had been got together to induce Blake to try a new offensive movement. Why was the congress not called in September rather than in November? Blake had always declared that he was too weak to risk a battle with the French for the raising of the siege, but till the last moment the Catalans contented themselves with arguing with him, and writing remonstrances to the Central Junta, instead of lending him the aid of their last levies.

[64] The Proclamation of Nov. 29 ordering this levy, written in a very magniloquent style, may be found in Belmas, Appendix no. 81.

One or two points connected with this famous siege require a word of comment. It is quite clear that St. Cyr during its early stages did not try his honest best to help Verdier. During June and July his covering army was doing no good whatever at Vich: he pretended that he had placed it there in order to ward off possible attacks by Blake. But it was matter of public knowledge that Blake was far away in Aragon, engaged in his unhappy campaign against Suchet, and that Coupigny, left at Tarragona with a few thousand men, was not a serious danger. St. Cyr could have spared a whole division more for the siege operations, without risking anything. If he had done so, Gerona could have been approached on two sides instead of one, the Mercadal front might have been attacked, and the loose blockade, which was all that Verdier could keep up, for want of more men, might have been made effective. But St. Cyr all through his military career earned a reputation for callous selfishness and habitual leaving of his colleagues in the lurch. On this occasion he was bitterly offended with Verdier, for giving himself the airs of an equal, and corresponding directly with the Emperor. There can be no doubt that he took a malicious pleasure in seeing his failures. It is hardly disguised in his clever and plausible _Journal des Opérations de l’Armée de Catalogne en 1808-1809_[65].

[65] Napoleon’s comments on the operations of his generals are always interesting, though sometimes founded on imperfect information, or vitiated by predispositions. Of St. Cyr’s campaign he writes [Disp. no. 16,004] to Clarke, his Minister of War:

‘Il faut me faire un rapport sérieux sur la campagne du général Gouvion Saint-Cyr en Catalogne: (1) Sur les raisons qui l’ont porté à évacuer cette province, lorsque Saragosse était prise et sa jonction faite avec le maréchal Mortier. (2) Sur ce qu’il s’est laissé attaquer par les Espagnols, et ne les a jamais attaqués, et sur ce que, après les avoir toujours battus par la valeur des troupes, il n’a jamais profité de la victoire. (3) Sur ce qu’il a, par cet esprit d’égoisme qui lui est particulier, compromis le siège de Gérone: sur ce qu’il n’a jamais secouru suffisamment l’armée assiégeante, l’a au contraire attirée à lui, et a laissé ravitailler la ville. (4) Sur ce qu’il a quitté l’armée sans permission, sous le vain prétexte de maladie.’

The first point seems unjust to St. Cyr. From his position in front of Tarragona, after Valls, he had no real chance of combining his operations with the army of Aragon. But the other three charges seem well founded.

Verdier, on the other hand, seems to have felt all through that he was being asked to perform a task almost impossible, when he was set to take Gerona with his own 14,000 men, unaided by the covering army. His only receipt for success was to try to hurry on the matter by delivering desperate blows. Both the assault on Monjuich on July 8 and that on the city on September 19 were premature; there was some excuse for the former: Verdier had not yet realized how well Alvarez could fight. But the second seems unpardonable, after the warning received at Monjuich. If the general, as he declared before delivering his assault, mistrusted his own troops, he had no right to order a storm at all, considering his experience of the way in which the Spaniards had behaved in July. He acted on the fallacious theory that a practicable breach implies a town that can be taken, which is far from being the case if the garrison are both desperate and ingenious in defending themselves. The only way to deal with such a resolute and capable adversary was to proceed by the slow and regular methods of siegecraft, to sap right up to the ditch before delivering an assault, and batter everything to pieces before risking a man. This was how Monjuich was actually taken, after the storm had failed. Having neither established himself close under the walls, nor subdued the flanking fires from the Calvary and Chapter redoubts, nor ascertained how far the Spaniards had prepared inner defences for themselves, he had no right to attack at all.

As to Blake, even after making all possible allowances for the fact that he could not trust his troops--the half-rallied wrecks of Maria and Belchite--for a battle in the field, he must yet be pronounced guilty of feebleness and want of ingenuity. If he could never bring up enough regulars to give him a chance of facing St. Cyr, the fault was largely his own: a more forcible general would have insisted that the Valencian reserves should march[66], and would have stripped Lerida and Tarragona of men: it could safely have been done, for neither Suchet nor Duhesme was showing any signs of threatening those points. He might have insisted that the Catalan Junta should call out the full levy of _somatenes_ in September instead of in November. He might also have made a better use of the irregulars already in the field, the bands of Rovira, Milans, and Claros. These miqueletes did admirable service all through the siege, by harassing Verdier’s rear and cutting off his convoys, but they were not employed (as they should have been) in combination with the regulars, but allowed, as a rule, to go off on excursions of their own, which had no relation to the main objects of Blake’s strategy. The only occasion on which proper use was made of them was when, on September 1, they were set to threaten Verdier’s lines, while Garcia Conde’s convoy was approaching Gerona. It may be pleaded in the Spanish general’s defence that it was difficult to exact obedience from the chiefs: there was a distinct coolness between the regulars and the irregulars, which sometimes led to actual quarrels and conflicts when they met. But here again the reply is that more forcible captain-generals were able to control the miqueletes, and if Blake failed to do so, it was only one more sign of his inadequacy. It is impossible to avoid the conclusion that he mismanaged matters, and that if in his second and third attempts to relieve Gerona he had repeated the tactics of his first, he would have had a far better chance of success. On September 1 only did he make any scientific attempt to distract the enemy’s attention and forces, and on that occasion he was successful. Summing things up, it may be said that he was not wrong to refuse battle with the troops that he had actually brought up to Gerona: they would undoubtedly have been routed if he had risked a general engagement. His fault was that he did not bring up larger forces, when it was in his power to do so, by the exercise of compulsion on the Catalan and Valencian Juntas. But these bodies must share Blake’s responsibilities: they undoubtedly behaved in a slack and selfish fashion, and let Gerona perish, though it was keeping the war from their doors for a long eight months.

[66] The Valencian troops at Maria were eleven battalions, viz. Savoia (three), 1st and 3rd Cazadores de Valencia (two), America (two), Voluntarios de Valencia (one), 1st of Valencia (three). Of these only Savoia (now two batts. only) and Voluntarios de Valencia turned up for the relief of Gerona. Along with them came two fresh regiments, 2nd Cazadores of Orihuela, and Almanza, which had not been at Maria. But these were Murcian, not Valencian, troops.

All the more credit is due to Alvarez, considering the way in which he was left unsuccoured, and fed with vain promises. A less constant soul would have abandoned the defence long before: the last two months of resistance were his sole work: if he had fallen sick in October instead of December, his subordinates would have yielded long before. But it is not merely for heroic obstinacy that he must be praised. Every detail of the defence shows that he was a most ingenious and provident general: nothing was left undone to make the work of the besiegers hard. Moreover, as Napier has observed, it is not the least of his titles to merit that he preserved a strict discipline, and exacted the possible maximum of work from soldier and civilian alike, without the use of any of those wholesale executions which disgraced the defence of Saragossa. His words were sometimes truculent, but his acts were just and moderate. He never countenanced mob-law, as did Palafox, yet he was far better obeyed by the citizens, and got as good service from them as did the Aragonese commander. He showed that good organization is not incompatible with patriotic enthusiasm, and is far more effective in the hour of danger than reckless courage and blind self-sacrifice.

SECTION XVII: CHAPTER IV

THE AUTUMN CAMPAIGN OF 1809: TAMAMES, OCAÑA, AND ALBA DE TORMES

As early as August 30, when Wellington had not fully completed his retreat from Almaraz and Jaraicejo to Badajoz and Merida, the central Junta had already begun to pester him and his brother, the Ambassador at Seville, with plans for a resumption of the offensive in the valley of the Tagus. On that day Martin de Garay, the Secretary of State, wrote to represent to Wellesley that he had good reason to believe that the troops of Victor, Mortier, and Soult were making a general movement to the rear, and that the moment had arrived when the allied armies in Estremadura and La Mancha should ‘move forward with the greatest activity, either to observe more closely the movements of the enemy, or to attack him when circumstances may render it expedient[67].’ The French movement of retreat was wholly imaginary, and it is astonishing that the Spanish Government should have been so mad as to believe it possible that ‘their retrograde movement may have originated in accounts received from the North, which compel the enemy either to retire into the interior of France, or to take up a position nearer to the Pyrenees.’ On a groundless rumour, of the highest intrinsic improbability, they were ready to hurl the newly-rallied troops of Eguia and Venegas upon the French, and to invite Wellington to join in the advance. Irresponsible frivolity could go no further. But the Junta, as has been already said, were eager for a military success, which should cause their unpopularity to be forgotten, and were ready to seize on any excuse for ordering their troops forward. This particular rumour died away--the French were still in force on the Tagus, and, as a matter of fact, the only movement northwards on their part had been the return of Ney’s corps to Salamanca. But though the truth was soon discovered, the Junta only began to look out for new excuses for recommencing active operations.

[67] De Garay to Wellesley in _Wellesley Dispatches_, p. 92.

Wellington, when these schemes were laid before him, reiterated his refusal to join in any offensive campaign, pointed out that the allied forces were not strong enough to embark on any such hazardous undertaking, and bluntly expressed his opinion that ‘he was much afraid, from what he had seen of the proceedings of the Central Junta, that in the distribution of their forces they do not consider military defence and military operations so much as political intrigue, and the attainment of petty political objects.’ He then proceeded to make an estimate of the French armies, to show their numerical superiority to the allies; in this he very much under-estimated the enemy’s resources, calculating the whole force of the eight corps in Spain at 125,000 men, exclusive of sick and garrisons not available for active service. As a matter of fact there were 180,000 men, not 125,000, with the Eagles at that moment, after all deductions had been made, so that his reasoning was far more cogent than he supposed[68]. But this only makes more culpable the obstinate determination of the Junta to resume operations with the much inferior force which they had at their disposal.

[68] Wellington to Wellesley, from Merida, Sept. 1, 1809.

Undismayed by their first repulse, the Spanish ministers were soon making new representations to Wellesley and Wellington, in order to induce them to commit the English army to a forward policy. They sent in repeated schemes for supplying Wellington with food and transport on a lavish scale[69]; but he merely expressed his doubts as to whether orders that looked admirable on paper would ever be carried out in practice. He consented for the present to remain at Badajoz, as long as he could subsist his army in its environs, but warned the Junta that it was more probable that he would retire within the Portuguese border, for reasons of supply, than that he would join in another campaign on the Tagus.

[69] See the details in Wellesley to Canning, Sept. 2, 1809.

Despite of all, the government at Seville went on with its plans for a general advance, even after they recognized that Wellington was not to be moved. A grand plan of operations was gradually devised by the War-Minister Cornel and his advisers. Stated shortly it was as follows. The army in La Mancha, which Venegas had rallied after the disaster of Almonacid, was to be raised to a strength of over 50,000 men by the drafting into it of a full two-thirds of Cuesta’s old army of Estremadura. On September 21 Eguia marched eastwards up the Guadiana, with three divisions of infantry and twelve or thirteen regiments of cavalry, to join Venegas[70]. The remaining force, amounting to two divisions of infantry and 2,500 cavalry, was left in Estremadura under the Duke of Albuquerque, the officer to whom the government was obliged to assign this army, because the Junta of Badajoz pressed for his appointment and would not hear of any other commander. He was considered an Anglophil, and a friend of some of the Andalusian malcontents, so the force left with him was cut down to the minimum. All the old regular regiments were withdrawn from him, save one single battalion, and he was left with nothing save the newly-raised volunteer units, some of which had behaved so badly at Talavera[71]. His cavalry was soon after reduced by the order to send a brigade to join the Army of the North, so that he was finally left with only five regiments of that arm or about 1,500 sabres. Of his infantry, about 12,000 strong, over 4,000 were absorbed by the garrison of Badajoz, so that he had only 8,000 men available for service in the field.

[70] His head quarters moved from Truxillo on the seventeenth, were at La Serena on the twenty-first, and joined the army of La Mancha about October 1.

[71] See the list of Albuquerque’s army in Appendix no. 2. There had been twenty-one regular battalions in Cuesta’s army in June. Twenty of these marched off with Eguia, leaving only one (4th Walloon Guards) with Albuquerque.

Eguia, on the other hand, carried with him to La Mancha some 25,000 men, the picked corps of the Estremaduran army; and, as the remains of Venegas’s divisions rallied and recruited after Almonacid, amounted to rather more than that number, the united force exceeded 50,000 sabres and bayonets. With this army the Junta intended to make a direct stroke at Madrid, while Albuquerque was directed to show himself on the Tagus, in front of Almaraz and Talavera, with the object of detaining at least one of the French corps in that direction. It was hoped, even yet, that Wellington might be induced to join in this demonstration. If once the redcoats reappeared at the front, neither Soult nor Mortier could be moved to oppose the army of La Mancha. Meanwhile Ney and the French corps in Leon and Old Castile were to be distracted by the use of a new force from the north, whose composition must be explained. The Junta held that the last campaign had failed only because the allies had possessed no force ready to detain Soult and Ney. If they had not appeared at Plasencia, Wellington, Cuesta, and Venegas would have been able to drive King Joseph out of his capital. Two months later the whole position was changed, in their estimation, by the fact that Spain once more possessed a large ‘Army of the Left,’ which would be able to occupy at least two French corps, while the rest of the allies marched again on Madrid. That such a force existed did indeed modify the aspect of affairs. La Romana had been moved to Seville to become a member of the Junta, but his successor, the Duke Del Parque, was collecting a host very formidable as far as numbers went. The old army of Galicia had been reformed into four divisions under Martin de la Carrera, Losada, Mahy, and the Conde de Belveder--the general whose name was so unfortunately connected with the ill-fought combat of Gamonal. These four divisions now comprised 27,000 men, of whom more than half were newly-raised Galician recruits, whom La Romana had embodied in the depleted cadres of his original battalions, after Ney and Soult had evacuated the province in July. A few of the ancient regiments that had made the campaign of Espinosa had died out completely--their small remnants having been drafted into other corps[72]. On the other hand there were a few new regiments of Galician volunteers--but La Romana had set his face against the creation of such units, wisely preferring to place his new levies in the ranks of the old battalions of the regular army[73]. In the main, therefore, the new ‘Army of the Left’ represented, as far as names and cadres went, Blake’s original ‘Army of Galicia[74].’ It had the same cardinal fault as that army, in that it had practically no cavalry whatever: the single dragoon regiment that Blake had owned (La Reina) having been almost completely destroyed in 1808[75]. Each division had a battery; the guns, of which La Romana’s army had been almost destitute in the spring, had been supplied from England, and landed at Corunna during the summer.

[72] The only regiments of Blake’s original army that seem to be completely dead in October 1809 are 2nd of Catalonia, Naples, Pontevedra, Compostella. Naples had been drafted into Rey early in 1809. Of the others I can find no details.

[73] The new Galician regiments which appear in the autumn of 1809 are Monforte de Lemos, Voluntarios de la Muerte, La Union, Lovera, Maceda, Morazzo.

[74] For the full muster-roll of Del Parque’s army in October, see Appendix no. 4.

[75] Some small fraction of it reappeared in the campaign of 1809.

But the Galician divisions, though the most numerous, were not the only units which were told off to the new ‘Army of the Left.’ Asturias had been free of invaders since Ney and Bonnet retired from its borders in June 1809. The Central Junta ordered Ballasteros to join the main army with the few regular troops in the principality, and ten battalions of the local volunteers, a force of over 9,000 men. The Asturian Junta, always very selfish and particularist in its aims, made some protests but obeyed. Nine of its less efficient regiments were left behind to watch Bonnet.

Finally the Duke Del Parque himself had been collecting fresh levies about Ciudad Rodrigo, while the plains of Leon lay abandoned by the French during the absence of Ney’s corps in the valley of the Tagus. Including the garrison of Rodrigo he had 9,000 men, all in new units save one old line battalion and one old militia regiment[76]. Deducting the 3,500 men which held the fortress, there were seven battalions--nearly 6,000 bayonets--and a squadron or two of horse available for the strengthening of the field army. These were now told off as the ‘5th Division of the Army of the Left’; that of Ballasteros was numbered the 3rd Division.

[76] One battalion of Majorca, and the Militia battalion of Segovia.

The Galician, Asturian, and Leonese divisions had between them less than 500 horsemen. To make up for this destitution the Central Junta directed the Duke of Albuquerque to send off to Ciudad Rodrigo, via the Portuguese frontier, a brigade of his cavalry. Accordingly the Prince of Anglona marched north with three regiments[77], only 1,000 sabres in all, and joined Del Parque on September 25. Thus at the end of that month the ‘Army of the Left’ numbered nearly 50,000 men--all infantry save 1,500 horse and 1,200 gunners. But they were scattered all over North-Western Spain, from Oviedo to Astorga, and from Astorga to Ciudad Rodrigo, and had to be concentrated before they could act. Nor was the concentration devoid of danger, for the French might fall upon the Asturians or the Leonese before they had joined the Galician main body. As a matter of fact the 50,000 never took the field in one mass, for Del Parque left a division under Mahy to protect Galicia, and, when these regiments and the garrison of Rodrigo were deducted, he had but 40,000 in all, including sick and men on detachment. This, nevertheless, constituted a formidable force--if it had been in existence in July, Soult and Ney could never have marched against Wellington with their whole strength, and the Talavera campaign might have had another end. But the troops were of varying quality--the Leonese division was absolutely raw: the Galicians had far too many recruits with only two months’ training in their ranks, the Estremaduran cavalry had a bad record of disasters. A general of genius might have accomplished something with the Army of the Left--but Del Parque, though more cautious than many of his compeers, was no genius.

[77] Borbon, Sagunto, and Granaderos de Llerena, 1,053 sabres in October. These regiments had newly rejoined the Estremaduran army from the rear.

The Junta had a deeply-rooted notion that if sufficient pressure were applied to Wellesley and Wellington, they would permit Beresford’s Portuguese army, now some 20,000 strong, to join Del Parque for the advance into the plains of Leon. They had mistaken their men: Wellington returned as peremptory a refusal to their request for the aid of the Portuguese troops as to their demand that his own British army should advance with Albuquerque to the Tagus[78].

[78] See Wellington to Wellesley, from Badajoz, Oct. 30, 1809.

Nothing could be more hazardous than the plan finally formulated at the Seville War Office for the simultaneous advance of the armies of La Mancha, the North, and Estremadura. Even if it had been energetically supported by Wellington and Beresford, it would have been rash: converging operations by several armies starting from distant bases against an enemy concentrated in their midst are proverbially disastrous. In this particular plan three forces--numbering in all about 110,000 men, and starting from points so far apart as Ciudad Rodrigo, Truxillo, and the Passes by La Carolina, were to fall upon some 120,000 men, placed in a comparatively compact body in their centre. A single mistake in the timing of operations, the chance that one Spanish army might outmarch another, or that one of the three might fail to detain any hostile force in its front (as had happened with Venegas during the Talavera Campaign) was bound to be ruinous. The French had it in their power to deal with their enemies in detail, if the least mischance should occur: and with Spanish generals and Spanish armies it was almost certain that some error would be made.

Meanwhile the Junta made their last preparation for the grand stroke, by deposing Venegas from the command of the united army in La Mancha. Eguia held the interim command for a few days, but was to be replaced by Areizaga, an elderly general who had never commanded more than a single division, and had to his credit only courage shown in a subordinate position at the battle of Alcañiz. He was summoned from Lerida, and came hastily to take up his charge.

The sole advantage which the Spaniards possessed in October 1809 was that their enemy did not expect to be attacked. A month after Talavera matters had apparently settled down for the whole autumn, as far as the French generals could calculate. With the knowledge that the Austrian War was over, and that unlimited reinforcements could now be poured into Spain by his brother, King Joseph was content to wait. He had refused to allow Soult to make his favourite move of invading Portugal in the end of August, because he wished the Emperor to take up the responsibility of settling the next plan of campaign, and of determining the number of new troops that would be required to carry it out. The French corps, therefore, were in a semicircle round Madrid: Soult and Mortier in the central Tagus Valley at Plasencia and Talavera, Victor in La Mancha, with Sebastiani supporting him at Toledo and Aranjuez, Ney at Salamanca, Dessolles and the Royal Guard as a central reserve in the capital. This was a purely defensive position, and Joseph intended to retain it, till the masses of troops from Germany, with the Emperor himself perchance at their head, should come up to his aid. It does not seem to have entered into his head that the enemy would again take the offensive, after the fiasco of the Talavera campaign, and the bloody lesson of Almonacid.

In September and the early days of October the French hardly moved at all. Ney left his corps at Salamanca, and went on a short leave to Paris on September 25, so little was any danger expected in the plains of Leon. The charge of the 6th corps was handed over to Marchand, his senior divisional general. There was an even more important change of command pending--Jourdan had been soliciting permission to return to France ever since July. He had been on excellent terms with King Joseph, but found it hard to exact obedience from the marshals--indeed he was generally engaged in a controversy either with Victor or with Soult. The Emperor was not inclined to allow him to quit Spain, but Jourdan kept sending in applications to be superseded, backed by medical certificates as to his dangerous state of health. Finally he was granted leave to return, by a letter which reached him on October 25, just as the new campaign was beginning to develop into an acute phase. But he gladly handed over his duties to Soult, who thus became ‘major-general’ or chief of the Staff to King Joseph, and departed without lingering or reluctance for France, glad to be quit of a most invidious office[79].

[79] For Jourdan’s personal views, see his _Mémoires_, ed. Grouchy, p. 282.

Before Jourdan’s departure there had been some small movements of the French troops: hearing vague rumours of the passage eastward of Eguia’s army, King Joseph ordered a corresponding shift of his own troops towards that quarter. Soult and the 2nd Corps were ordered from Plasencia to Oropesa and Talavera, there relieving Mortier and the 5th Corps, who were to push up the Tagus toward Toledo. This would enable Victor to call up Sebastiani’s cavalry and two of his infantry divisions from Toledo into La Mancha. Having thus got together some 25,000 men, Victor advanced to Daimiel, and pressed in the advanced posts of the main Spanish army on October 15. Eguia, who was still in temporary command, since Areizaga had not yet arrived, made no attempt to stand, but retired into the passes of the Sierra Morena. This apparent timidity of the enemy convinced the Marshal that nothing dangerous was on hand in this quarter. He drew back his army into cantonments, in a semicircle from Toledo to Tarancon, leaving the cavalry of Milhaud and Paris out in his front.

Nothing more happened in La Mancha for a fortnight: but on the other wing, in the kingdom of Leon, matters came to a head sooner. About the middle of September the bulk of the Galician army, the divisions of Losada, Belveder and La Carrera, had moved down the Portuguese frontier via Alcanizas, and joined Del Parque at Ciudad Rodrigo. On the twenty-fifth of the same month the Prince of Anglona, with the cavalry brigade from Estremadura, also came in to unite himself to the Army of the Left. Del Parque had thus 25,000 infantry and 1,500 horse concentrated. He had still to be joined by Ballasteros and the Asturians, who had to pick their way with caution through the plains of Leon. Mahy and the 4th division of the Galicians had been left in the passes above Astorga, to cover the high-road into Galicia. He had a vanguard in Astorga, under Santocildes, and the town, whose walls had been repaired by the order of La Romana, was now capable of making some defence.

Facing Del Parque and his lieutenants there were two distinct forces. The 6th Corps, now under Marchand, was concentrated at Salamanca. Having received few or no drafts since its return from Galicia it was rather weak--its twenty-one battalions and four cavalry regiments only counted at the end of September some 13,000 bayonets and 1,200 sabres[80] effective--the sick being numerous. In the north of Leon and in Old Castile Kellermann was in charge, with an independent force of no great strength: his own division of dragoons, nearly 3,000 sabres, was its only formidable unit. The infantry was composed of three Swiss battalions, and four or five French battalions, which had been left in garrisons in Old Castile when the regiments to which they belonged went southward in the preceding winter[81]. The whole did not amount to more than 3,500 bayonets. The dragoons were very serviceable in the vast plains of Leon, but it was with difficulty, and only by cutting down garrisons to a dangerous extent, that Kellermann could assemble a weak infantry brigade of 2,000 men to back the horsemen.

[80] See the table given by Sprünglin on p. 366 of his _Mémoires_.

[81] Apparently Kellermann had at this moment a battalion each of the 2nd, 3rd, and 4th Swiss, a battalion of the Garde de Paris, one each of the 12th Léger and 32nd Line, and one or two of the 122nd.

It was nevertheless on Kellermann’s side, and by the initiative of the French, that the first clash took place in north-western Spain. Hearing vague reports of the movement of the Galician divisions towards Ciudad Rodrigo, Kellermann sent General Carrié, with two regiments of dragoons and 1,200 infantry, to occupy Astorga, being ignorant apparently that it was now garrisoned and more or less fortified. Carrié found the place occupied, made a weak attack upon it on October 9, and was beaten off. He was able to report to his chief that the Spaniards (i. e. Mahy’s division) were in some force in the passes beyond.

At much the same moment that this fact was ascertained Del Parque began to move: he had been lying since September 24 at Fuente Guinaldo in the highland above Ciudad Rodrigo. On October 5 he made an advance as far as Tamames, on the by-road from Rodrigo to Salamanca which skirts the mountains, wisely avoiding the high-road in the more level ground by San Martin del Rio and Castrejon. He had with him his three Galician divisions and his 1,500 horse, but he had not brought forward his raw Leonese division under Castrofuerte, which still lay by Rodrigo. On hearing of the duke’s advance Marchand sent out reconnaissances, and having discovered the position of the Spaniards, resolved at once to attack them. On October 17 he started out from Salamanca, taking with him his whole corps, except the two battalions of the 50th regiment, which were left to garrison the town.

On the afternoon of the next day Marchand came in sight of the enemy, who was drawn up ready to receive him on the heights above Tamames. The French general had with him nineteen battalions, some 12,000 bayonets--his 1,200 horse, and fourteen guns. Del Parque had 20,000 Galician infantry, Anglona’s cavalry, and eighteen guns: his position was so strong, and his superiority in infantry so marked, that he was probably justified in risking a battle on the defensive.

Tamames, an unwalled village of moderate size, lies at the foot of a range of swelling hills. Its strategical importance lies in the fact that it is the meeting-place of the two country roads from Ciudad Rodrigo to Salamanca _via_ Matilla, and from Ciudad Rodrigo to Bejar and the Pass of Baños _via_ Nava Redonda. Placed there, Del Parque’s army threatened Salamanca, and had a choice of lines of retreat, the roads to Rodrigo and to the passes into Estremadura being both open. But retreat was not the duke’s intention. He had drawn up his army on the heights above Tamames, occupying the village below with a battalion or two. On the right, where the hillside was steeper, he had placed Losada and the 2nd Division: on the left, where the ridge sinks down gently into the plain, was Martin de la Carrera with the Vanguard Division. The Conde de Belveder’s division--the third--formed the reserve, and was drawn up on the reverse slope, behind La Carrera. The Prince of Anglona’s cavalry brigade was out on the extreme left, partly hidden by woods, in the low ground beyond the flank of the Vanguard.

Marchand, arriving on the ground in the afternoon after a march of fourteen miles from Matilla, was overjoyed to see the enemy offering battle, and attacked without a moment’s hesitation. His arrangements much resembled those of Victor at Ucles--though his luck was to be very different. It was clear that the Spanish left was the weak point, and that the heights could be turned and ascended on that side with ease. Accordingly Maucune’s brigade (six battalions in all)[82] and the light cavalry, strengthened by one regiment of dragoons, were ordered to march off to the right, to form in a line perpendicular to that of Del Parque, and break down his flank. When this movement was well developed, Marcognet’s brigade (six battalions)[83] was to attack the Spanish centre, to the east of the village of Tamames, while the 25th Léger (two battalions) was to contain the hostile right by a demonstration against the high and difficult ground in that direction. Marchand kept in reserve, behind his centre, the 27th and 59th of the Line (six battalions) and his remaining regiment of dragoons. The vice of this formation was that the striking force--Maucune’s column--was too weak: it would have been wise to have strengthened it at the expense of the centre, and to have made a mere demonstration against the heights above the village of Tamames, as well as on the extreme French left.

[82] 6th Léger (two batts.), 69th Line (three batts.), and one battalion of _voltigeurs réunis_.

[83] 39th and 76th of the Line.

Maucune accomplished his flank march undisturbed, deployed in front of La Carrera’s left and advanced against it. The Spanish general threw back his wing to protect himself, and ordered his cavalry to threaten the flank of the advancing force. But he was nearly swept away: when the skirmishing lines were in contact, the French brigadier ordered his cavalry to charge the centre of the Spanish division: striking in diagonally, Lorcet’s Hussars and Chasseurs broke La Carrera’s line, and captured the six guns of his divisional artillery. Almost at the same moment Anglona’s cavalry came in upon Maucune’s flank; but being opposed by two battalions of the 69th in square, they received but one fire and fled hastily to the rear. Maucune then resumed his march up the hill, covering his flank with his horsemen, and pushing La Carrera’s broken line before him. But at the head of the slope he met Belveder’s reserve, which let the broken troops pass through their intervals, and took up the fight steadily enough. The French were now opposed by triple numbers, and the combat came to a standstill: Maucune’s offensive power was exhausted, and he could no longer use his cavalry on the steep ground which he had reached.

Meanwhile, on seeing their right brigade opening the combat with such success, the two other French columns went forward, Marcognet against the Spanish centre, Anselme of the 25th Léger against the extreme right. But the ground was here much steeper: Losada’s Galician division stood its ground very steadily, and Marcognet’s two regiments made an involuntary halt three-quarters of the way up the heights, under the full fire of the two Spanish batteries there placed and the long line of infantry. The officers made several desperate attempts to induce the columns to resume their advance, but to no effect. They fell in great numbers, and at last the regiments recoiled and descended the hill in disorder. Losada’s battalions pursued them to the foot of the slope, and the Spanish light troops in the village sallied out upon their flank, and completed their rout. Marcognet’s brigade poured down into the plain as a disordered mass of fugitives, and were only stayed when Marchand brought up the 27th and 59th to their rescue. Del Parque wisely halted the pursuing force before it came into contact with the French reserves, and took up again his post on the heights.

Meanwhile the 25th Léger, on the extreme French right, had not pressed its attack home, and retreated when the central advance was repulsed. Maucune, too, seeing the rout to his left, withdrew from the heights under cover of his cavalry, carrying off only one of the Spanish guns that he had taken early in the fight, and leaving in return a disabled piece of his own on the hill.

The battle was fairly lost, and Marchand retired, under cover of his cavalry along the Salamanca road. The enemy made no serious attempt to pursue him in the plain, where his horsemen would have been able to act with advantage. The French had lost 1,300 or 1,400 men, including 18 officers killed, and a general (Lorcet) and 54 officers wounded[84]. Marcognet’s brigade supplied the greater part of the casualties; the 76th lost its eagle, seven officers killed and fifteen wounded: the 39th almost as many. The cavalry and Maucune’s brigade suffered little. The very moderate Spanish loss was 713 killed and wounded, mostly in La Carrera’s division.

[84] Marchand in his dispatch says 1,300 men in all were lost, and a gun; he makes no mention of the eagle. His aide-de-camp, Sprünglin, who has a good account of the battle in his _Mémoires_ (pp. 370-1), gives the total of 1,500. The Spaniards exaggerated the loss to 3,000.

This was the first general action since Baylen in which the Spaniards gained a complete victory. They had a superiority of about seven to four in numbers, and a good position; nevertheless the troops were so raw, and the past record of the Army of the Left was so disheartening, that the victory reflects considerable credit on the Galicians. The 6th Corps was reckoned the best of all the French units in Spain, being entirely composed of old regiments from the army of Germany. It is not too much to say that Ney’s absence was responsible for the defeat of his men. Marchand attacked at three points, and was weak at each. The Marshal would certainly have massed a whole division against the Spanish left, and would not have been stopped by the stout resistance made by Belveder’s reserve. A demonstration by a few battalions would have ‘contained’ Losada’s troops on the left, where the ground was too unfavourable for a serious attack[85].

[85] ‘La perte de cette affaire fut entièrement due à la faute que fit le Général Marchand de multiplier ses attaques, et de s’engager par petits paquets. Tout le monde se mêlait de donner son avis, et on remarquait l’absence de M. le Maréchal,’ says Sprünglin in p. 371 of his _Mémoires_.

On the 19th of October the beaten army reached Salamanca by a forced march. Marchand feared that the enemy would now manœuvre either by Ledesma, so as to cut him off from Kellermann and the troops in the north, or by Alba de Tormes, so as to intercept his communication with Madrid. In either case he would have to retreat, for there was no good defensive ground on the Tormes to resist an army coming from the west. As a matter of fact Del Parque moved by Ledesma, for two reasons: the first was that he wished to avoid the plains, fearing that Kellermann might have joined the 6th Corps with his cavalry division. The second was that, by moving in this direction, he hoped to make his junction with Ballasteros, who had started from the Asturias to join him, and had been reported to have moved from Astorga to Miranda del Duero, and to be feeling his way south-eastward. The juncture took place: the Asturian division, after an unsuccessful attempt to cut off the garrison of Zamora on the seventeenth, had marched to Ledesma, and met the main army there. Del Parque had now 28,000 men, and though still very weak in cavalry, thought himself strong enough to march on Salamanca. He reached it on October 25 and found it evacuated. Marchand, learning that Kellermann was too far off to help him, and knowing that no reinforcements from Madrid could reach him for many days, had evacuated the town on the previous evening. He retired towards Toro, thus throwing up his communications with Madrid in order to make sure of joining Kellermann. This seems doubtful policy, for that general could only aid him with 4,000 or 5,000 men, and their joint force would be under 20,000 strong. On the other hand, by retiring on Peñaranda or Medina de Campo, and so approaching the King’s army, he could have counted on picking up much larger reinforcements, and on resuming the struggle with a good prospect of success.

As a matter of fact Jourdan, on hearing of the disaster of Tamames, had dispatched, to aid the 6th Corps, Godinot’s brigade of Dessolles’ division, some 3,500 bayonets, from Madrid, and Heudelet’s division of the 2nd Corps, about 4,000 strong, from Oropesa, as well as a couple of regiments of cavalry. He made these detachments without scruple, because there was as yet no sign of any activity on the part of the Spanish armies of La Mancha and Estremadura. A week later he would have found it much more hazardous to weaken his front in the valley of the Tagus. These were the last orders issued by Jourdan, who resigned his post on October 31, while Soult on November 5 arrived at Madrid and replaced him as chief of the staff to King Joseph.

Del Parque, not unnaturally elated by his victory, now nourished ambitious ideas of clearing the whole of Leon and Old Castile of the enemy, being aware that the armies of La Mancha and Estremadura ought now to be on the move, and that full occupation would be found ere long for the French corps in the valley of the Tagus. He ordered up his 5th Division, the raw Leonese battalions of Castrofuerte, from Ciudad Rodrigo, and made vehement appeals to the Portuguese Government to lend him the whole of Beresford’s army for a great advance up the Douro. The Regency, though much pressed by the Spanish ambassador at Lisbon, gave a blank refusal, following Wellington’s advice to have nothing to do with offensive operations in Spanish company[86]. But part of Beresford’s troops were ordered up to the frontier, not so much to lend a moral support to Del Parque’s advance[87] as to be ready to defend their own borders in the event of his defeat. Showing more prudence than Wellington had expected, Del Parque did not push forward from Salamanca, when he became certain that he would have to depend on his own forces alone. Even after the arrival of his reserves from Rodrigo he remained quiet, only pushing out reconnaissances to discover which way the enemy had gone. He had, in fact, carried out his part in the Central Junta’s plan of campaign, by calling the attention of the French to the north, and distracting troops thither from the King’s army. It was now the time for Albuquerque and Areizaga to take up the game, and relieve him. Marchand meanwhile had retired across the Douro, and taken up an extended line behind it from Zamora to Tordesillas--a front of over forty miles--which it would have been impossible to hold with his 13,000 men against a heavy attack delivered at one point. But he was hardly in position when Kellermann arrived, took over the command, and changed the whole plan of campaign (November 1). He had left two battalions to guard Benavente, two to hold Valladolid, and had only brought up his 3,000 dragoons and 1,500 infantry. Seeing that it was absolutely necessary to recover the line of communication with Madrid, he ordered the 6th Corps to leave Zamora and Toro, mass at Tordesillas, and then cross the Douro to Medina del Campo, the junction point of the roads from Madrid, Segovia, Valladolid, and Toro. To this same place he brought up his own small force, and having received Godinot’s brigade from Madrid, had thirty-four battalions and eighteen squadrons concentrated--about 23,000 men. Though not yet joined by the other troops from the south--Heudelet’s division--he now marched straight upon Salamanca in two columns, one by Cantalapiedra, the other by Fuente Sauco, intending to offer battle to Del Parque.

[86] Del Parque’s demands had begun as early as the end of September, see Wellington to Castlereagh, Badajoz, Sept. 29, _Dispatches_, v. 200-1, and cf. Wellington to Forjaz, Oct. 15, ibid. 223.

[87] Wellington to Beresford, Nov. 16, 1809.

But the duke, much to the surprise of every one, utterly refused to fight, holding the plain too dangerous for an army so weak in cavalry as his own, and over-estimating the enemy’s force at 36,000 men[88]. He retired from Salamanca, after having held it less than a fortnight, on November 5, and took not the road to Ciudad Rodrigo but that to Bejar and the Pass of Baños, as if he were about to pass the mountains into Estremadura[89]. This was an excellent move: the French could not pursue him in force without evacuating Old Castile and Leon, which it would have been impossible for them to contemplate. For when Kellermann had concentrated his troops to strike at Salamanca, there was nothing left behind him in the vast upland save a battalion or two at Benavente, Valladolid, and Burgos. Mahy, from Galicia, and the Asturians might have overrun the whole region unopposed. As it was, the whole of the provinces behind the Douro showed signs of bursting out into insurrection. Julian Sanchez, the Empecinado, and other guerrillero chiefs, whose names were soon to be famous, raised large bands during the absence of the normal garrisons, and swept the country-side, capturing convoys and cutting the lines of communication between Vittoria, Burgos, and Valladolid. Porlier came down with a flying column from the Asturias, assaulted Palencia, and threatened Burgos. The French governors on every side kept reporting their perilous position, when they could get a message through to Madrid[90].

[88] He sent this estimate to Wellington, see the latter to Beresford, Badajoz, Nov. 16.

[89] The Junta afterwards contemplated bringing him down to join Albuquerque, via Plasencia, which was free of French troops, since Soult had moved to Oropesa. But this does not seem to have been thought of so early as Nov. 5.

[90] See Soult to Clarke, from Madrid, Nov. 6, for these movements.

Realizing that he must cover his rear, or the whole of Old Castile would be lost to the insurgents, Kellermann, after occupying Salamanca on November 6, left the 6th Corps and Godinot’s brigade distributed between Ledesma, Salamanca, and Alba de Tormes, watching Del Parque, and returned in haste with his own troops to the Douro. He commenced to send out flying columns from Valladolid to deal with the guerrilleros, but did not work too far afield, lest he might be called back by a new forward movement on the part of the Army of the Left. But in a few days he had to recast all his arrangements, for--as Del Parque had calculated--the campaign in La Mancha had just opened, and the position of the French in Leon and Old Castile was profoundly affected by the new developments.

In the south, as we have already explained, the Junta designed Albuquerque’s army of Estremadura to be a mere demonstrating force, while Areizaga’s 55,000 men were to strike the real blow. The Estremaduran troops, as was proper, moved early to draw the attention of the enemy. Albuquerque’s first division under Bassecourt--6,000 infantry and 600 horse--was on the Tagus from Almaraz to Meza de Ibor: his second division under St. Juan and the rest of his cavalry--some 4,000 in all--were moving up from Truxillo. Bassecourt began by sending a small force of all arms across the river at Almaraz, to drive in Soult’s outposts and spread reports abroad in all directions that he was acting as the vanguard to Wellington’s army, which was marching up from Badajoz. Unfortunately the full effect that he desired was not produced, because deserters informed Soult that the British Army was still quiescent on the Guadiana[91]. The French made no movement, and left the 2nd Corps alone to watch Albuquerque.

[91] Soult to Clarke, from Madrid, Nov. 6. The deserters were a body of twenty-one men of the Walloon Guards, who had enlisted from Dupont’s prisoners in order to get a chance of escaping: they reached Oropesa on Oct. 25.

Meanwhile Areizaga, within a few days of assuming the command of the army of La Mancha, commenced his forward movement. On November 3, having concentrated his eight divisions of infantry and his 5,700 horse at Santa Cruz de Mudela, at the foot of the passes, he gave the order to advance into the plains. The head quarters followed the high-road, with the train and three divisions: the rest, to avoid encumbering the _chaussée_, marched by parallel side-roads, but were never more than ten miles from their Commander-in-Chief: at any rate Areizaga avoided the sin of dispersion. His army was the best which had been seen under the Spanish banners since Tudela. The men had all been furnished with new clothes and equipment since August, mainly from English stores landed at Cadiz. There were sixty guns, and such a body of cavalry as had never yet been collected during the war. The value of the troops was very unequal; if there were many old battalions of the regular army, there were also many new units composed of half-trained Andalusian levies. The cavalry included the old runaways of Medellin, and many other regiments of doubtful value. The morale was on the whole not satisfactory. ‘I wish I had anything agreeable to communicate to you from this army’ wrote Colonel Roche, a British officer attached to Areizaga’s staff, to Wellington. ‘The corps which belonged to the original army of La Mancha are certainly in every respect superior to those from Estremadura, and from everything that I can learn none of those abuses which were to be lamented in the army of Estremadura existed here--or, at least, in a much less degree. But nothing can exceed the general discontent, dissatisfaction, and demoralization of the mass of the people and of the army. How can anybody who has the faculty of reason separate the inefficiency, intrigue, bad organization, and consequent disasters of the army from the source of all those evils in the Junta? There is not a man of the least reflection who, as things now stand, has a hope of success; and this is the more melancholy, because the mass of the people are just as inveterate in their resentment and abhorrence of the French as at the first hour of the revolution[92].’ The fact seems to have been that the superior officers doubted the wisdom of taking the offensive according to the Junta’s orders, and had no confidence in Areizaga, who was only known as a fighting general, and had no reputation for skill. The rank and file, as Arteche remarks, were disposed to do their duty, but had no confidence in their luck[93]. Their government and not their generals must take the major part of the blame for the disaster that followed.

[92] Roche to Wellington, from Santa Cruz de la Mudela, _Wellington Supplementary Dispatches_, vi. 394. Cf. also the same to the same, vi. 414.

[93] _Historia de la Guerra de la Independencia_, vii. 283.

Areizaga was well aware that his best chance was to strike with extreme boldness and vigour, and to dash into the midst of the French before they could concentrate. Hence his march was at first conducted with great rapidity and decision; between the 3rd and the 8th of November he made nearly fifteen miles a day, though the roads were somewhat broken up by the autumn rains. On the eighth he reached La Guardia, eighty miles from his starting-point, and his advanced cavalry under General Freire had its first skirmish with a brigade of Milhaud’s dragoons at Costa de Madera, near Dos Barrios. The Spanish horse deployed in such numbers that the French were compelled to move off in haste and with some loss, though they had beaten off with ease the first two or three regiments which had gone forward against them.

The Spanish advance had been so rapid and so unexpected that Soult and King Joseph had been taken completely by surprise. On November 6 the Marshal had reported to Paris that ‘the troops on the Tagus and in La Mancha are up to the present unmolested, and as, from all I can learn, there is no prospect of the enemy making any offensive movement on that side, I intend to form from them a strong flying column to hunt the brigands in the direction of Burgos[94].’ Only four days later he had to announce that an army of at least 40,000 men was close in front of Aranjuez, and not more than thirty-five miles from Madrid, and that he was hurrying together troops from all quarters to make head against them. At the moment indeed, there was nothing directly between Areizaga’s vanguard at La Guardia and the Spanish capital, save the Polish division of the 4th Corps stationed at Aranjuez, and Milhaud’s five regiments of dragoons at Ocaña. If the Spaniard had pushed on for three days more at his starting pace, he might have crossed the Tagus, and have forced King Joseph to fight, close in front of Madrid, with an imperfectly assembled army. On the ninth and tenth Leval’s Germans were in march from Toledo to Aranjuez to join Sebastiani’s Poles: Mortier’s first division was hurrying from Talavera to Toledo, and his second division was making ready to follow. The 2nd Corps, despite Albuquerque’s demonstration in front of Almaraz, was preparing to quit Oropesa, in order to replace Mortier’s men at Talavera. Victor, in the meanwhile, with the First Corps, was lying in front of Toledo at Ajofrin, with his cavalry at Mora and Yebenes: he reported that no hostile force had come his way, but that he had ascertained that a large army had marched past his front along the great _chaussée_ from Madridejos to Aranjuez. He was in a position to attack it in rear and flank, if there was a sufficient force gathered in its front to justify him in closing[95].

[94] Soult to Clarke, Madrid, Nov. 6.

[95] Soult to Clarke, Madrid, Nov. 10.

But on reaching La Guardia, Areizaga seemed suddenly to realize the dangers of his movement. No doubt it was the news that Victor was almost in his rear that paralysed him, but he halted on the ninth, when a bold advance would certainly have enabled him to seize Aranjuez, by evicting the small force under Milhaud and Sebastiani. For three fatal days, the 9th, 10th, and 11th of November, the Spanish main body remained halted in a mass at La Guardia, as if for the special purpose of allowing the enemy to concentrate. On the eleventh Areizaga at last began to move again: he sent forward the whole of his cavalry, supported by Zayas and his Vanguard division, to press back the force in his front. They found Milhaud’s five regiments of dragoons ranged in line of battle before the small town of Ocaña, and supported by Sebastiani’s Polish infantry. Freire advanced, using his triple superiority of numbers to turn both flanks of the French cavalry; Milhaud, after some partial charges, retired behind the Poles, who formed a line of six battalion squares. The Spanish horse made a half-hearted attempt to attack them, but were repelled by their rolling fire before they came to close quarters, and drew back. It was now four o’clock in the afternoon, and the Spanish infantry was only just beginning to come up. Zayas and Freire agreed that it was too late to begin a second attack, and put off fighting till the next morning. But during the night the French evacuated Ocaña and retired to Aranjuez, wisely judging that it would be insane to wait for the arrival of the Spanish main body. They had lost about fifty men, Freire’s cavalry just over two hundred.

Next day [November 12], Areizaga brought up the whole of his army to Ocaña, and his cavalry reconnoitred up to the gates of Aranjuez and the bridge of Puente La Reyna. Sebastiani made ready to defend them, and having been joined by the German division from Toledo, wrote to Soult to say that he would resist to the last extremity, in order to gain time for the arrival of Victor’s corps and the other troops which were marching up from the west and north[96]. The attack which he expected was never delivered. Areizaga, nervous about the presence of the 1st Corps on his flank, had resolved to shift his army eastward to get further away from it. Abandoning his line of communication by La Guardia and Madridejos, he marched his whole force by cross-roads parallel to the Tagus up to La Zarza, and seized the fords of Villamanrique, twenty-five miles above Aranjuez, on the Madrid-Albacete road. If Victor, as he supposed, had been manœuvring on his flank, this movement would have cut him off from his base in Andalusia, and have left him only the mountains of Murcia as a line of retreat. But, as a matter of fact, the 1st Corps was no longer at Ajofrin or Mora, but had been called behind the Tagus, so that his retreat was safer than he supposed.

[96] Sebastiani to Soult, night of the twelfth-thirteenth, from Aranjuez.

Soult and King Joseph, meanwhile, had been completing their concentration. They had written to Kellermann, ordering him to send back to Madrid without delay the brigade of Dessolles’ division under Godinot which had been lent him, and to spare them as well one infantry brigade of the 6th Corps. These troops were too far off to be available at once; but of the remainder of their units the Royal Guard and Spanish battalions of King Joseph, with Dessolles’ remaining brigade, were moved out to support Sebastiani. Victor had been brought back across the Tagus, and was also marching on Aranjuez. Mortier’s corps was concentrated at Toledo, while the 2nd Corps was in motion from Oropesa to Talavera, having discovered no signs of a serious advance on the part of Albuquerque. The care of Madrid was handed over to the incomplete French division of the 4th Corps[97], some of whose battalions were dispersed at Guadalajara, Alcala, Segovia, and other garrisons. Paris’s light cavalry of the same corps was also at this moment watching the roads to the east of Madrid.

[97] It had still no divisional general, and was officially known by the name of ‘Sebastiani’s division’--regiments 28th, 32nd, 58th, 75th.

On the twelfth Areizaga threw Lacy’s division across the Tagus, and laid down two pontoon bridges near Villamanrique, so as to be able to bring over his whole army in the shortest possible time. But the thirteenth, fourteenth, and fifteenth were days of storm, the river rose high, and the artillery and train stuck fast on the vile cross-roads from Ocaña over which they were being brought. In consequence less than half the Spanish army was north of the Tagus on November 15, though the advance cavalry pushed on to the line of the Tajuna, and skirmished with Paris’s _chasseurs_ about Arganda. It seemed nevertheless that Areizaga was committed to an advance upon Madrid by the high-road from Albacete, wherefore Soult blew up the bridges of Aranjuez and Puente la Reyna, and ordered Victor to march from Aranjuez on Arganda with the 1st Corps, nearly 20,000 men, purposing to join him with the King’s reserves and to offer battle on the Tajuna, while Mortier and Sebastiani’s Poles and Germans should fall upon the enemy’s flank. But this plan was foiled by a new move upon Areizaga’s part; he now commenced a retreat as objectless as his late advance. Just as Victor’s cavalry came in touch with his front, he withdrew his whole army across the Tagus, destroyed his bridges, and retired to La Zarza on the seventeenth, evidently with the intention of recovering his old line of communication with Andalusia, via Ocaña and Madridejos.

The moment that this new departure became evident, Soult reversed the marching orders of all his columns save Victor’s, and bade them return hastily to Aranjuez, where the bridge was repaired in haste, and to cross the Tagus there, with the intention of intercepting Areizaga’s line of retreat and forcing a battle on him near Ocaña. Victor, however, had got so far to the east that it would have wasted time to bring him back to Aranjuez, wherefore he was directed to cross the river at Villamanrique and follow hard in Areizaga’s rear.

On the morning of the eighteenth Milhaud’s and Paris’s cavalry, riding at the head of the French army, crossed the Tagus at Aranjuez, and pressing forward met, between Ontigola and Ocaña, Freire’s horsemen moving at the head of Areizaga’s column, which on this day was strung out between La Zarza and Noblejas, marching hastily westward towards the high-road. The collision of Milhaud and Freire brought about the largest cavalry fight which took place during the whole Peninsular War. For Milhaud and Paris had eight regiments, nearly 3,000 men, while three of Freire’s four divisions were present, to the number of over 4,000 sabres. On neither side was any infantry in hand.

Sebastiani, who had come up with the light cavalry of his corps, was eager for a fight, and engaged at once. Charging the Spanish front line with Paris’s light horse, he broke it with ease: but Freire came on with his reserves, forming the greater part of them into a solid column--an odd formation for cavalry. Into this mass Milhaud charged with four regiments of dragoons. The heaviness of their formation did not suffice to enable the Spaniards to stand. They broke when attacked, and went to the rear in disorder, leaving behind them eighty prisoners and some hundreds of killed and wounded. The French lost only a few scores, but among them was Paris, the not unworthy successor of the adventurous Lasalle in command of the light cavalry division attached to the 4th Corps.

Moving forward in pursuit of the routed squadrons, Sebastiani approached Ocaña, but halted on discovering that there was already Spanish infantry in the town. The head of Areizaga’s long column had reached it, while the cavalry combat was in progress: the rest was visible slowly moving up by cross-roads from the east. Soult was at once apprised that the enemy’s army was close in his front--so close that it could not get away without fighting, for its train and rearguard were still far behind, and would be cut off if the main body moved on without making a stand.

Areizaga, though he had shown such timidity when faced by Sebastiani’s 9,000 men at Aranjuez, and by Victor’s 20,000 on the Tajuna, now offered battle to the much more formidable force which Soult was bringing up. He was indeed compelled to fight, partly because his men were too weary to move forward that night, partly because he wished to give time for his train to arrive and get on to the _chaussée_.

On the morning of the nineteenth his army was discovered drawn up in two lines on each side of the town of Ocaña. There were still some 46,000 infantry and 5,500 cavalry under arms despite of the losses of the late week[98]. The oncoming French army was smaller; though it mustered 5,000 horse it had only 27,000 foot--the Germans and Poles of Sebastiani, Mortier with nearly the whole of the 5th Corps, a brigade of Dessolles’ division, the King’s guards, and the cavalry of Milhaud, Paris, and Beauregard[99]. Victor was too far off to be available; having found the flooded Tagus hard to cross, he was on this day barely in touch with the extreme rearguard of Areizaga’s army which was escorting the train. Being nearly twenty miles from Ocaña, he could not hope to arrive in time for the general action, if it was to be delivered next morn. If Areizaga stood firm for another day, Victor would be pressing him from the flank and rear while the main army was in his front: but it was highly probable that Areizaga would not stand, but would retreat at night; all his previous conduct argued a great disinclination to risk a battle. Wherefore Soult and the King, after a short discussion[100], agreed to attack at once, despite their great numerical inferiority. In the open plain of La Mancha a difference of 16,000 or 17,000 infantry was not enough to outweigh the superior quality and training of the French army.

[98] Colborne, in a letter dated December 5, says ‘we had 46,100 infantry and nearly 6,000 cavalry drawn out, in a very bad position.’ He was present all through the campaign, but wrote no full report.

[99] Viz. Mortier’s Infantry Divisions (Girard and Gazan), twenty-two batts. [one regiment deducted] about 12,000 men Sebastiani’s Polish Division and German Division (under Werlé and Leval) ” 8,000 ” Rey’s Brigade of Dessolles’ Division of the Central Reserve ” 3,500 ” The King’s Reserves, viz. four guard battalions and three others ” 3,500 ” Milhaud’s Dragoons, five regiments ” 1,800 ” Paris’s Light Cavalry, attached to 4th Corps, three regiments ” 1,000 ” Beauregard’s Cavalry of the 5th Corps, four regiments ” 1,500 ” The King’s Cavalry, one regiment of the Guards, one of Chasseurs ” 700 ” Artillery, Sappers, &c. ” 1,500 ” ------ Total about 33,500 men

[100] Joseph declared that he urged instant attack when Soult advised waiting for Victor. See his letter in vol. vii of Ducasse’s _Life and Correspondence of Joseph Napoleon_.

There is, so to speak, no position whatever at Ocaña: the little unwalled town lies in a level upland, where the only natural feature is a ravine which passes in front of the place; it is sufficiently deep and broad at its western end to constitute a military obstacle, but east of the town gradually grows slighter and becomes a mere dip in the ground. Areizaga had chosen this ravine to indicate the line of his left and centre; but on his right, where it had become so shallow as to afford no cover, he extended his troops across and beyond it. The town was barricaded and occupied, to form a central support to the line. There were olive-groves in the rear of Ocaña which might have served to hide a reserve, or to mark a position for a rally in case a retreat should become necessary. But Areizaga had made no preparation of this sort. His trains, with a small escort, had not arrived even on the morning of the nineteenth, but were still belated on the cross-roads from Noblejas and La Zarza.

The order of the Spanish army in line of battle is difficult to reconstruct, for Areizaga uses very vague language in the dispatch in which he explained his defeat, and the other documents available, though they give detailed accounts of some of the corps, say little or nothing of others. It seems, however, that Zayas, with the vanguard division, formed the extreme left, behind the deepest part of the ravine, with a cavalry brigade under Rivas on its flank and rear. He had the town of Ocaña on his right. Then followed in the line, going from left to right, the divisions of Vigodet, Giron, Castejon, and Lacy. Those of Copons, Jacomé, and Zerain appear to have formed a second line in support of the other four[101]. Vigodet’s left was in the town of Ocaña and strongly posted, but the other flank, where Lacy lay, was absolutely in the air, with no natural feature to cover it. For this reason Areizaga placed beyond it Freire, with the whole of the cavalry except the brigade on the extreme left under Rivas. Unfortunately the Spanish horse, much shaken by the combat of the preceding day, was a weak protection for the flank, despite its formidable numbers. The sixty guns of the artillery were drawn out in the intervals of the infantry divisions of the first and second line.

[101] This order seems the only one consistent with the sole sentence in Areizaga’s dispatch to the Junta in which he explains his battle-array: ‘Inmediatamente formé por mi mismo la primera linea en direccion de Ocaña, colocando por la izquierda la division de Vigodet, defendida por la frente de la gran zanja, y por su derecha las divisiones de Giron, Castejon y Lacy: la de Copons formaba martillo, junta á las tapias de la villa, inmediata á la de Giron, y las demás la secunda linea á distancia competente para proteger á la primera.’ The unnamed divisions which must have lain beyond Copons in the right of the second line are Jacomé and Zerain.

Soult’s plan of attack was soon formed. The ravine made the Spanish left--beyond Ocaña--inaccessible, but also prevented it from taking any offensive action. The Marshal therefore resolved to ignore it completely, and to concentrate all his efforts against the hostile centre and right, in the open ground. The scheme adopted was a simple one: Sebastiani’s Polish and German divisions were to attack the Spanish right wing, and when they were at close quarters with the enemy the main mass of the French cavalry was to fall upon Freire’s horse, drive it out of the field, and attack on the flank the divisions already engaged with the infantry. For this purpose Milhaud’s, Paris’s, and Beauregard’s regiments, more than 3,500 sabres, were massed behind the Poles and Germans. For a time their march would be masked by olive-groves and undulations of the ground, so that they might come in quite suddenly upon the enemy. Mortier with his first division--that of Girard--and a regiment of Gazan’s, followed in the rear of the Polish and German infantry, to support their frontal attack. Dessolles, with his own brigade and Gazan’s remaining one, took post opposite Ocaña, ready to fall upon the Spanish centre, when the attack to his left should have begun to make way. He had in his front the massed artillery of the 4th and 5th Corps, thirty guns under Senarmont, which took ground on a low knoll above the great ravine, from which they could both play upon the town of Ocaña and also enfilade part of the Spanish line to its immediate right--Vigodet’s division and half of Giron’s. Finally the King, with his guards and other troops, horse and foot, were placed to the right rear of Dessolles, to act as a general reserve, or to move against Zayas if he should attempt to cross the ravine and turn the French right.

The plan, despite of some checks at the commencement, worked in a satisfactory fashion. The German and Polish divisions of Leval and Werlé attacked Lacy’s and Castejon’s divisions, which gave back some little way, in order to align themselves with Vigodet who was sheltered by the slight eastern end of the ravine. The enemy followed and brought up six guns to the point to play upon the new position which the Spaniards had taken up. The forward movement was continuing, when suddenly to the surprise of the French, Lacy’s, Castejon’s, and Giron’s men, leaving their places in the line, made a furious counter-charge upon the Poles and Germans, drove them back for some distance, and threw them into disorder. This movement was no result of Areizaga’s generalship: he had betaken himself to the summit of the church-tower of Ocaña, an inconvenient place from which to issue orders, and practically left his subordinates to fight their own battle. Mortier was forced to bring forward Girard’s division to support his broken first line. It was hotly engaged with Lacy and Giron, when suddenly it felt the Spaniards slacken in their fire, waver, and break. This was the result of the intervention of a new force in the field. The great mass of French squadrons, which had been sent under Sebastiani to turn the Spanish right, had now come into action. Arriving close to Freire’s cavalry before it was discovered, it fell on that untrustworthy corps, and scattered it to the winds in a few minutes. Then, while three or four regiments followed the routed horsemen, the rest turned inwards upon the hostile infantry. The flanks of the first and second lines of Areizaga’s right were charged simultaneously, and hardly a regiment had time to get into square. Brigade after brigade was rolled up and dispersed or captured; the mass of fugitives, running in upon the troops that were frontally engaged with Girard, wrecked them completely. Of the five divisions of the Spanish left, a certain number of steady regiments got away, by closing their ranks and pushing ahead through the confusion, firing on friend and foe alike when they were hustled. But many corps were annihilated, and others captured wholesale. The last seems to have been the fate of nearly the whole of Jacomé’s division of the second line, as hardly a single unit from it is reported as rallied a month later, and the French accounts speak of a whole column of 6,000 men which laid down its arms in a mass before the light cavalry of the 4th Corps. Just as the Spanish right broke up, Dessolles with his two brigades, followed by the King’s reserve, crossed the ravine and attacked the town of Ocaña, and the two divisions--Vigodet and Copons--which lay in first and second line immediately to the east of it. These retired, and got away in better order than their comrades to the right. Of all the Spanish army only Zayas’s vanguard division, on the extreme left, now remained intact. Areizaga had sent it an order to cross the ravine and attack the French right, when he saw his army beginning to break up. Then, a few minutes later, he sent another order bidding it close to the right and cover the retreat. After this the Commander-in-Chief descended from his tower, mounted his horse, and fled. Zayas carried out the second order, moved to the right, and found himself encompassed by masses of fugitives from Giron’s, Castejon’s, and Lacy’s broken divisions, mixed with French cavalry. He sustained, with great credit to himself and his troops, a rearguard action for some miles, till near the village of Dos Barrios, where his line was broken and his men at last mixed with the rest of the fugitives[102].

[102] The only detailed accounts of the Spanish movements that I have discovered are the divisional reports of Lacy and Zayas, both in the Foreign Office archives at the Record Office. Areizaga’s dispatch is so vague as to be nearly useless.

The whole routed multitude now streamed wildly over the plain, with the French cavalry in hot pursuit. Thousands of prisoners were taken, and the chase only ended with nightfall. The fugitives headed straight for the Sierra Morena, and reached it with a rapidity even greater than that which they had used in their outward march a fortnight before. Victor’s cavalry arrived in time to take up the pursuit next morning: they had on their way to the field captured the whole of the trains of the Spanish army, on the road from Noblejas to Ocaña. The losses of Areizaga’s army were appalling; about 4,000 killed and wounded and 14,000 prisoners. Thirty flags and fifty out of the sixty guns had been captured. When the wrecks of the army had been rallied in the passes, three weeks after the battle, only some 21,000 infantry[103] and 3,000 horse were reported as present. The divisions of Lacy, Jacomé, and Zerain had practically disappeared, and the others had lost from a third to a half of their numbers. The condition of the cavalry was peculiarly disgraceful; as it had never stood to fight, its losses represent not prisoners, for the most part, but mere runaways who never returned to their standards. The French had lost about 90 officers and 1,900 men, nearly all in the divisions of Leval, Werlé, and Girard[104]. The cavalry, which had delivered the great stroke and won the battle, suffered very little. Mortier had been slightly wounded, Leval and Girard severely.

[103] Viz. Zayas, Vigodet, and Castejon, about 4,000 men each, Copons 3,000, Giron 2,500, remains of the other three divisions about 3,500. From the returns in the Madrid War Office.

[104] Martinien’s lists of officers killed and wounded show that the German division lost 19 officers, the Polish division 23, Girard’s division 28--in all 70 out of the total of 94 officers hit in the whole army.

Even allowing for the fact that Areizaga had been the victim of the Junta’s insensate resolve to make an offensive movement on Madrid, it is impossible to speak with patience of his generalship. For a combination of rashness and vacillation it excels that of any other Spanish general during the whole war. His only chance was to catch the enemy before they could concentrate: he succeeded in doing this by his rapid march from the passes to La Guardia. Then he waited three days in deplorable indecision, though there were only 10,000 men between him and Madrid. Next he resumed his advance, but by the circuitous route of Villamanrique, by taking which he lost three days more. Then he halted again, the moment that he found Victor with 20,000 men in his front, though he might still have fought at great advantage. Lastly he retreated, yet so slowly and unskilfully that he was finally brought to action at Ocaña by the 34,000 men of Mortier and Sebastiani. He was sent out to win a battle, since Madrid could not be delivered without one, and knew that he must fight sooner or later, but threw away his favourable opportunities, and then accepted an action when all the chances were against him. For he must have known by this time the miserable quality of his cavalry, yet gave battle in a vast plain, where everything depended on the mounted arm. In the actual moment of conflict he seems to have remained in a hypnotized condition on his church-tower, issuing hardly an order, and allowing the fight to go as it pleased. Yet he was, by all accounts, possessed of personal courage, as he had proved at Alcañiz and elsewhere. Apparently responsibility reduced him to a condition of vacillating idiocy. Perhaps the most surprising fact of the whole business is that the Junta retained him in command after his fiasco, thanked him for his services, and sent him an honorary present--as it had done to Cuesta after Medellin with somewhat better excuse. He was its own man, and it did not throw him over, even when he had proved his perfect incompetence.

To complete the narrative of the deplorable autumn campaign of 1809, it only remains to tell of the doings of Albuquerque and Del Parque. The former played his part with reasonable success; he was ordered to distract the attention of the enemy from the army of La Mancha, and did what he could. Having got some 10,000 men concentrated at Almaraz, he sent one column over the Tagus to demonstrate against the 2nd Corps from beyond the river, and with another threatened the bridge of Talavera from the near side. But Heudelet, now in command of the 2nd Corps, soon found that there was no reality in his demonstration, and that he was not supported by the English, though he had given out that Wellington was close in his rear. After skirmishing around Talavera from the 17th to the 22nd of November, the Duke hastily recrossed the river on hearing the news of Ocaña, and resumed his old positions.

Del Parque’s campaign was more vigorous and more unfortunate. While he lay in the passes above Bejar and Baños, he got early news of the withdrawal of Godinot’s and Marcognet’s troops toward Madrid, when Soult summoned them off to reinforce the main army. He reasoned that since he had now only the 6th Corps, shorn of one of its brigades, in his front, he might repeat the success of Tamames, for Marchand was weaker than he had been in October, while he himself was far stronger. Accordingly he disregarded an order from the Junta to extend his operations southward, and to join Albuquerque in the valley of the Tagus. Instead, he marched once more upon Salamanca on November 18, the day before the disaster of Ocaña. He drove in an outlying brigade of Marchand’s force from Alba de Tormes, and pressed it vigorously back towards the main body. Conscious that with his 10,000 men he could not hope to face 30,000, Marchand promptly evacuated Salamanca on December 19, and retired, just as he had done in October, behind the Douro, concentrating his whole corps at Toro. He sent urgent demands for help both to Kellermann at Valladolid, and to Soult at Madrid. By the time that they arrived Areizaga had been dealt with, and the army in New Castile could spare as many reinforcements as were required. Marcognet’s brigade, the one which had been borrowed from the 6th Corps, was first sent back from Segovia, the point which it had reached in its southward march, and Gazan’s division of the 5th Corps was ordered by Soult to follow.

Meanwhile Del Parque, still ignorant of the disaster in the south, had occupied Salamanca on November 20, and on the following day moved out towards Cantalapiedra and Medina del Campo, with the object of throwing himself between Marchand and Kellermann and the capital. This was an excellent move, and, but for what had happened at Ocaña, might have had considerable results, since the Army of the Left ought to have made an end of the small French force in Old Castile.

Kellermann, however, had seen the danger of Marchand’s retreat to Toro, and had directed him to close in towards the east, and to occupy Medina del Campo, as the strategical point that must be held in order to maintain touch with Madrid. Thus it chanced that on November 23 Labassée’s brigade and four regiments of cavalry, coming from Tordesillas, reached Medina del Campo just as Marcognet’s brigade, returning from Segovia, came into the town from the other side. They had hardly met when the approach of Del Parque’s army along the Salamanca road was reported. The two French brigadiers thought for a moment of fighting, and the cavalry was ordered to press back the Spanish advanced guard. They drove off with ease Anglona’s horsemen, who rode at the head of the long column, but were repulsed by Ballasteros’s infantry, which formed square in good style, and drove them off with a rolling fire of musketry. Seeing that the whole Spanish army was coming up, Marcognet and Labassée then evacuated Medina del Campo, and retired to Valdestillas. With one push more the Spaniards could have cut the line between Valladolid and Madrid.

On November 24 the whole 6th Corps and Kellermann’s dragoons, with a battalion or two from the garrisons of Old Castile, were concentrated at Puente de Duero, with their van at Valdestillas. If attacked, they must have gone behind the Douro and abandoned all touch with Madrid; for there were not more than 16,000 men in line, and they were forced to take the defensive. But, to their surprise, Del Parque made no advance. He had heard on that morning of the disaster of Ocaña, and guessed that reinforcements for Kellermann must already be on the march. Wherefore he resolved to regain the mountains without delay, and to give up Salamanca and his other conquests. With this prudent resolve he broke up from Medina del Campo, and marched hastily away in retreat, making, not for Salamanca, which was too much in the plains to please him, but for Alba de Tormes. He had gained a day’s start by his prompt action, but on the twenty-sixth Kellermann set off in pursuit, leaving orders for the troops that were expected from Madrid to follow him.

On the twenty-sixth and twenty-seventh the French cavalry failed even to get in touch with Del Parque’s rearguard, and found nothing but a few stragglers on the road. But on the afternoon of the twenty-eighth the leading squadrons reported that they had come upon the whole Spanish army encamped in a mass around the town of Alba de Tormes. The duke had flattered himself that he had shaken off his pursuers, and was surprised in a most unfortunate position. Two of his divisions (Ballasteros and Castrofuerte) were beyond the Tormes, preparing to bivouac on the upland above it. The other three were quartered in and about the town, while the cavalry was watching the road, but had fallen in so close to the main body that its vedettes gave very short notice of the approach of the enemy. Kellermann was riding with the leading brigade of his cavalry--Lorcet’s _chasseurs_ and hussars; the six regiments of dragoons were close behind him, so that he had over 3,000 sabres in hand; but the infantry was ten miles to the rear. If he waited for it, Del Parque would have time to cross the river and take up a defensive position behind it. The French general, therefore, resolved to risk a most hazardous experiment, an attack with unsupported cavalry upon a force of all arms, in the hope of detaining it till the infantry should come up. The Spaniards were getting into line of battle in a hurry, Losada’s division on the right, Belveder’s and La Carrera’s on the left, the cavalry--1,200 sabres at most--in their front. The divisions beyond the river were only beginning to assemble, and would take some time to recross the narrow bridge: but 18,000 men were on the right bank prepared to fight.

Without a moment’s delay Kellermann ordered Lorcet’s brigade to charge the Spanish right and centre: it was followed by the six regiments of dragoons in three successive lines, and the whole mass came down like a whirlwind upon Del Parque’s front, scattering his cavalry to the winds, and breaking the whole of Losada’s and the right of Belveder’s divisions. A battery of artillery, and nearly 2,000 prisoners were taken. The wrecks of the broken divisions fell back into Alba de Tormes, and jammed the bridge, thus preventing the divisions on the further side from recrossing it. Kellermann then rallied his squadrons, and led them against La Carrera’s division and the remaining battalions of that of Belveder. These troops, formed in brigade-squares upon a rising ground, held out gallantly and repulsed the charge. But they were cut off from the bridge, which they could only reach by a dangerous flank movement over rough ground. By continually threatening to repeat his attacks, Kellermann kept them from moving off, till, two hours and a half after the action had begun, the French infantry and guns commenced to come up. La Carrera saw that it would be fatal to await them, and bade his division retreat and reach the bridge as best it could. This was naturally done in disorder, and with some loss; but it was already growing dusk, and the bulk of the Spanish left got away.

While the Spaniards were defiling over the bridge, Marchand’s leading brigade attacked Alba, out of which it drove some rallied troops of Losada’s division, who held the town to cover La Carrera’s retreat. This was done with ease, for Del Parque had not brought over his two intact divisions, preferring to use them as a second line behind which the others could retire. Alba was stormed, and two guns, which had been placed behind a barricade at its main exit, were taken by the French.

Here the fighting stopped: the Spaniards had lost five flags, nine guns, most of their baggage, and about 3,000 killed or taken--no very ruinous deductions from an army of 32,000 men. The French casualties were less than 300 in all[105]. Del Parque was determined not to fight again next morning, and bade his army make off under cover of the night. The disorder that followed was frightful: the three divisions that had been in the battle dispersed, and went off in all directions, some towards Ciudad Rodrigo, others towards Tamames, others by the hill-road that leads towards Tala and the Pass of Baños. Many of the raw Leonese troops, though they had not been engaged, also left their colours in the dark[106]. It was a full month before Del Parque could collect his whole army, which, when it had been reorganized, was found to number 26,000 men, despite all its misfortunes. It would seem, therefore, that beside the losses in the battle some 3,000 men must have gone off to their homes. The duke fixed his head quarters at San Martin de Trebejos in the Sierra de Gata, and dispersed his infantry in cantonments about Bejar, Fuenteguinaldo, and Miranda de Castanar. Having only the ruined region around Coria and Plasencia, and the small district about Ciudad Rodrigo, to feed them, these troops suffered dreadful privations during the winter, living on half-rations eked out with edible acorns. By the middle of January they had lost 9,000 men from fever, dysentery, and starvation.

[105] Martinien’s lists show 4 officers killed and 14 wounded.

[106] There is a long report by Del Parque in the Record Office, in which he states that the panic was caused by a stray party of his own routed cavalry dashing in among the rearguard in the dark, and crying that the French were pursuing them. He afterwards court-martialled and shot some cavalrymen for cowardice.

Despite all this, it is fair to say that Del Parque’s campaign contrasts most favourably with that of Areizaga. He showed a laudable prudence when he twice evacuated Salamanca rather than fight a battle in the plain. His victory of Tamames was most creditable, showing that when prudently conducted, and ranged in a well-chosen hill-position, his army could give a good account of itself. But for the disaster of Alba de Tormes his record might be considered excellent. There, it is true, he committed a grave mistake, by separating his army into two halves by the river when his enemy was in pursuit. But in his defence it may be urged that his cavalry ought to have had vedettes out for ten or fifteen miles to the rear, and to have given him long warning of the approach of the French. And when the enemy’s horse did make its sudden appearance, it was contrary to the laws of probability that it would attack at once, without waiting for its infantry and guns. Kellermann’s headlong charge was a violation of all rules, a stroke of inspiration which could not have been foreseen. If the Spanish cavalry had been of any use whatever, and if Losada’s division had only known how to form square in a hurry, it ought to have been beaten off. But the resisting-power of a Spanish army was always a doubtful quantity. Kellermann resolved to take the risk of attacking, and was rewarded by a victory on which he was not entitled to reckon. He would probably have justified his tactics by urging that failure could have no severe penalty, for the Spaniards could not pursue him if he were repulsed, while success would bring splendid results. This was true: and if his infantry had been five miles more to the front, he might have captured the whole of La Carrera’s division.

SECTION XVIII

THE CONQUEST OF ANDALUSIA