CHAPTER I

THE INVASION OF VALENCIA. SIEGE OF SAGUNTUM. SEPTEMBER-OCTOBER 1811

In the last volume of this work the chronicle of all the campaigns of 1811 was completed, save in one corner of Spain, where, on the eastern coast, the fortunes of the French armies have only been pursued down to the recall of Marshal Macdonald to Paris on October 28th. Already, before the Duke of Tarentum had been added to the list of the generals who had been withdrawn and superseded for failure in Catalonia, another series of operations had been begun in the East, which was destined to lead directly to one more Spanish disaster, but indirectly to the ruin of the French cause in Spain. For, as has already been pointed out in the last pages of the last volume[1], it was to be the diversion by Napoleon’s orders of French divisions eastward, from the borders of Portugal to those of Valencia, that was to give Wellington his long-desired opportunity of opening a successful offensive campaign against his immediate opponents in the West. The fall of Valencia was to lead to the fall of Ciudad Rodrigo in January 1812.

[1] vol. iv. pp. 587-91.

It will be remembered that the Emperor’s ambitious schemes for the conquest of the kingdom of Valencia, the last district of eastern Spain where he had as yet secured no solid foothold, had been deferred perforce till Figueras fell, on August 19, 1811. As long as that great fortress, which lies only a few miles from the French frontier, and blocks the main road from Perpignan to Barcelona, had been maintained against Macdonald by the resolute Martinez, it was impossible to take up a new offensive campaign: all the disposable French troops in Catalonia were immobilized around the stubborn garrison. At length the remnant of the starving miqueletes had laid down their arms, and the troops which had been for so long blockading them became disposable for the assistance of Suchet, whose ‘Army of Aragon’ was to deliver the main blow against Valencia.

Six days after the surrender of Figueras the news that the obstacle to advance had been at last removed reached Paris, on August 25, and on the same evening Berthier wrote, by his master’s orders, to bid Suchet move forward: ‘Everything leads us to believe that Valencia is in a state of panic, and that, when Murviedro has been taken and a battle in the open field has been won, that city will surrender. If you judge otherwise, and think that you must wait to bring up your siege artillery for the attack on the place, or that you must wait for a better season [i. e. early autumn] to commence the operation, I must inform you that, in every case, it is the imperative order of the Emperor that your head-quarters are to be on Valencian territory on or about September 15th, and as far forward towards the city as possible.’

The orders were feasible, and (as we shall see) were duly executed: but Napoleon had committed his usual mistake of undervaluing the tenacity of the Spanish enemy, whom he so deeply despised. Suchet set his troops in motion on September 15th; he took Murviedro--but only after a desperate siege of two months--he beat the army of Valencia in a very decisive pitched battle, but the city by no means fulfilled the Emperor’s prophecy by a prompt surrender. Fighting round its walls went on for five weeks after Murviedro fell: and it was not till troops had been brought to aid Suchet from very remote provinces, that he at last compelled the capitulation of Valencia after the New Year of 1812 had passed. Before the city yielded Wellington was on the move, far away on the Portuguese frontier, and it was not many days after Suchet’s aide-de-camp brought the glorious news of the capitulation of Valencia, that Marmont’s aide-de-camp followed, with the wholly unexpected and unwelcome tidings that the British had stormed Ciudad Rodrigo, and that the hold of the French army on Leon and Castile had been shaken. The one piece of information was the complement and consequence of the other.

Suchet’s invasion of Valencia, in short, was a much harder and more venturesome enterprise than his master had calculated. It was true that the Spanish forces in front of him seemed in September wholly incapable of holding him back. The Army of Catalonia had been reduced by a series of disasters, culminating in the falls of Tarragona and Figueras, to a mere remnant of 8,000 men, lurking in the high hills of the interior. The Army of Valencia had made a miserable exhibition of itself during the last year: it had brought no effective help to the Catalans, and whenever any of its detachments came into contact with the French, they had invariably suffered discreditable defeats, even when their numbers were far greater than those of the invaders. Of all the armies of Spain this was undoubtedly the one with the worst fighting reputation. It was to small profit that the Captain-General was raising yet newer and rawer battalions than those which already existed, to swell the numbers, but not the efficiency, of his command. In July the nominal total of the Valencian army, including the irregulars of the ‘flying column’ of the Empecinado, had been just 30,000 men. By October there were 36,000 under arms, including the new ‘Reserve Division[2],’ whose six battalions of recruits had only 135 officers to 6,000 men--an allowance of one officer to 45 men, not much more than half of the proportion that is necessary even among good veteran troops. But in truth the only valuable fighting force that was present in the kingdom in September was the infantry of the two weak divisions of the old Albuera army, under Zayas and Lardizabal, whom Blake had brought round from Cadiz with him, when he assumed command of the Eastern provinces. They did not between them muster more than 6,000 bayonets, but were good old troops, who were to distinguish themselves in the oncoming campaign.

[2] ‘The Reserve Division’ consisted of a 3rd battalion from some of the old regiments of the Valencian army, viz. 1st of Savoya, Avila, Don Carlos, Volunteers of Castile, Cazadores de Valencia, Orihuela. They were each about 1,000 strong, but averaged only 22 officers per battalion.

In addition, it was possible that Valencia might be able to draw a few thousand men to her aid from the depleted army of Murcia, which had suffered so severely at Soult’s hands during the short campaign of the previous August[3]. But such assistance was purely problematical; if Soult should stir again from the side of Andalusia, it would be impossible for General Mahy to bring a single Murcian battalion to the succour of Blake. If, by good fortune, he should not, only a fraction of Mahy’s small army would be free, since the greater part of it would be required to watch the Andalusian frontier, and to protect the great naval arsenal and fortress of Cartagena.

[3] See vol. iv. pp. 475-83.

If the regular troops only in eastern Spain had to be counted, it was certain that Suchet could dispose of numbers superior to his adversaries. The gross total of the French Army of Catalonia, where General Decaen had now taken Macdonald’s place, was 30,000 men. That of Suchet’s own ‘Army of Aragon’ was nearly 50,000, if garrisons, sick, and drafts on the march are reckoned in it. With these deducted, it could still supply about 31,000 men of all arms for the field. But these were not the only resources available. On the upper Ebro, in Navarre and western Aragon, were the two newly arrived divisions of Reille and Severoli, which had entered Spain during the summer, and had hitherto had no occupation save a little hunting of Mina’s guerrilleros. These two divisions counted 15,000 fresh troops of good quality, and Suchet reckoned on their assistance to cover his rear, when he should begin his march on Valencia. Technically they belonged to Dorsenne’s ‘Army of the North,’ but Severoli’s Italians had been promised as a reinforcement for Aragon already, and when Suchet asked for the grant of Reille’s division also it was not denied him. There were 70,000 men in all to be taken into consideration when the attack on Valencia was planned out.

No such force, of course, could be set aside for the actual invasion. The reason why not half so many thousands could be utilized for the projected stroke was that the Spanish War, as we have already had to point out on many occasions, was not a normal struggle between regular armies. The French had not only to conquer but to occupy every province that they overrun. Wherever an adequate garrison was not left, the guerrilleros and miqueletes inundated the country-side, cut all communications, and blockaded such small detachments as had been left far apart from the main army. Suchet’s 70,000 men had to hold down Aragon and Catalonia, at the same time that they undertook the further extension of their master’s power on the Valencian side.

Decaen in Catalonia had 23,000 men fit for service, not including sick and drafts on the march. Lacy’s little army was not more than 8,000 strong in September: yet Suchet dared not take away a man from Catalonia. The large garrison of Barcelona, a whole division, and the smaller garrisons of Gerona, Rosas, and Mont Louis absorbed nearly half the effective total. The remainder were, as it turned out, not strong enough to keep the Catalans in check, much less to prosecute active offensive operations against them. It was in October, after Suchet had started against Valencia, that Lacy carried out the series of small successful raids against Igualada, Cervera, and Montserrat, which have been spoken of in an earlier chapter[4]. We need not wonder, then, that not a Frenchman was drawn from Catalonia: they were all wanted on the spot to keep a tight hold on the turbulent principality. The example of the surprise of Figueras in the last spring was sufficient to prove the necessity of keeping every point strongly garrisoned, on pain of possible disaster.

[4] See vol. iv. pp. 540-1.

As to the Army of Aragon, it was far stronger than the Army of Catalonia, but on the other hand it had even more fortresses to garrison. Saragossa, Tortosa, Tarragona, Lerida, were large places, each absorbing several battalions. In addition there were the smaller strongholds of Jaca, Mequinenza, Monzon, Morella, requiring care. All these were regular fortresses, but they did not exhaust the list of points that must be firmly held, if the communications of Suchet’s field-force with its distant base were to be kept free and unhampered. Southern Aragon and the mountain-ganglion where the borders of that kingdom and of Valencia and New Castile meet, in the roughest country of the whole Spanish peninsula, had to be guarded. For in this region lay the chosen hunting-ground of the guerrillero bands of the Empecinado, Duran, and many other lesser chiefs: and Mina himself, from his usual haunts in Navarre, not unfrequently led a raid far to the south of the Ebro. Suchet had therefore to place garrisons in Teruel, Daroca, Alcañiz, Calatayud, and Molina, none of which possessed modern fortifications. The detachments left to hold them had to utilize a large convent, a mediaeval castle, or some such post of defence, in case they were attacked by the roving hordes of the enemy. Able to protect themselves with ease against small parties, and to keep the roads open under ordinary circumstances, they were exposed to serious danger if the guerrilleros should mass themselves in force against any one garrison--more especially if the bands should have been lent a few cannon and gunners from the regular Spanish armies. For convents or old castles could not resist artillery fire.

To cover his rear Suchet was forced to set aside one whole division, that of Frère, thirteen battalions strong[5], and mustering over 7,000 men, and immense detachments of the three other French divisions of the Army of Aragon. The units told off for the field army left no less than 6,800 able-bodied men (besides sick and convalescents) behind them, while they took 22,000 to the front. Frère’s division remained on the side of Western Catalonia, holding Lerida and Tortosa in force, and the intermediate places with small posts. The detachments from Musnier’s, Harispe’s, and Habert’s French, and from Palombini’s Italian divisions, took charge of Southern Aragon, leaving a company here and a battalion there. But the Marshal selected with great care the men who were to march on the Valencian expedition: each regiment drafted its most effective soldiers into the marching units, and left the recruits and the old or sickly men in the garrisons. Thus the battalions used in the oncoming campaign were rather weak, averaging not much over 450 men, but were composed entirely of selected veterans. The only doubtful element taken forward was the so-called ‘Neapolitan Division’ of General Compère, which was only 1,500 strong--in reality a weak brigade--and had no great reputation. But what was left of this corps was its best part--the numerous men who wanted to desert had already done so, and its weaklings were dead by this time. Of his cavalry Suchet took forward almost the whole, leaving behind only two squadrons of the 4th Hussars for the service between the garrisons, and of the other regiments only the weakly men and horses[6]. Practically all his horse and field artillery also went forward with him.

[5] Composed at this time of the 14th and 42nd and 115th Line, and the 1st Léger, the first two and last each three battalions strong, the other (115th) with four.

[6] The 24th Dragoons left about 140 men behind, the 13th Cuirassiers 50 only, the Italian ‘Dragoons of Napoleon’ 124, but the 4th Hussars about 500, much more than half their force.

Of his own Army of Aragon, Suchet, as we have thus seen, left nearly 14,000 men ‘present under arms’ to cover his rear. But this was not enough to make matters wholly secure, so untameable were the Aragonese and Catalans with whom he had to deal. Indeed, if this force only had been left to discharge the appointed task, it is clear, from subsequent happenings, that he would have suffered a disaster during his absence in Valencia. He asked from the Emperor the loan of half Reille’s division from Navarre, as well as the prompt sending to the front of Severoli’s Italians, who had been promised him as a reinforcement when first they entered Spain. The petition was granted, and these troops entered northern Aragon, and took charge of the places along the Ebro, while the expeditionary army was on its way to Valencia. Most of them were ultimately brought forward to the siege of the great city, and without them neither could Aragon have been maintained nor Valencia captured. Practically we may say that Suchet, at his original start, took 26,000 men to beat the Valencians and capture their city, but that he left nearly 30,000 more behind him, to hold down the provinces already conquered and to deal with the guerrilleros.

Two main roads lead from the north to Valencia: the one, coming from Tortosa and Catalonia, hugs the coast of the Mediterranean, from which it is never more than a few miles distant. The other, far inland, and starting from Saragossa, follows the valley of the Xiloca among the hills of Southern Aragon, crosses the watershed beyond Teruel, and descends to the sea near Murviedro, where it joins the coast-road only a few miles north of Valencia. There is a third, and much inferior, route between these two, which starts from Mequinenza on the Lower Ebro, crosses the mountainous Valencian frontier near Morella, and comes down to the coast at Castellon de la Plana, twenty miles north of Murviedro. Of these roads the first was as good as any in Spain, and was suitable for all manner of traffic: but it had the disadvantage of being flanked at a distance of only two miles by the small but impregnable fortress of Peniscola, which lies on a rocky headland thirty miles beyond Tortosa, and of being absolutely blocked by the little town of Oropesa, twenty miles further south. Oropesa was no more than a ruinous mediaeval place, with two castles hastily repaired, without any modern works: but since the road passed through it, no heavy guns or wagons starting from Tortosa could get further south till its forts had been captured.

The second road, that from Aragon by Teruel and Murviedro, is marked on contemporary maps as a post-route fit for all vehicles: but it passed through a very mountainous country, and was much inferior as a line of advance to the coast-road. It was not blocked by any fortress in the hands of the Spaniards, but between Teruel and Segorbe it was crossed by many ridges and ravines highly suitable for defence. The third track, that by Morella, was unsuitable for wheeled traffic, and could only be used by infantry and cavalry. Its one advantage was that Morella, its central point, had been already for some time in French hands, and contained a garrison and stores, which made it a good starting-point for a marching column.

[Illustration: SUCHET’S CAMPAIGNS 1811-12 IN VALENCIA]

Suchet determined to use all three of these roads, though such a plan would have been most hazardous against a wary and vigorous enemy: for though they all converge in the end on the same point, Murviedro, they are separated from each other by long stretches of mountain, and have no cross-communications. In especial, the road by Teruel was very distant from the other two, and any isolated column taking it might find itself opposed by immensely superior forces, during the last days of its march; since Valencia, the enemy’s base and headquarters, where he would naturally concentrate, lies quite close to the concluding stages of the route Teruel-Murviedro. It must have been in sheer contempt for his opponent--a contempt which turned out to be justified--that the Marshal sent a detachment of eleven battalions by this road, for such a force of 5,000 men might have been beset by the whole Valencian army, 30,000 strong, and the other columns could not have helped it.

Suchet’s arrangements were governed by a single fact--his siege artillery and heavy stores were parked at Tortosa, and from thence, therefore, along the coast road, must be his main line of advance, though it would be necessary to mask Peniscola and to capture Oropesa, before he could get forward to his objective--the city of Valencia. It might have seemed rational to move the whole field army by this route: but some of the troops destined for it were coming from distant points, and to march them down the Ebro bank to Tortosa would have taken much time. Moreover if the whole force concentrated there, it would all have to be fed from the magazines at Tortosa, and those lying in Aragon would be of no use. The Marshal started himself from this point, on September 15, with the division of Habert, and an infantry reserve formed of Robert’s brigade of the division of Musnier, together with the whole of the cavalry and field artillery of the army. The siege-train guarded by the other brigade of Musnier’s division--that of Ficatier--followed: but Musnier himself did not accompany the expedition, having been left in general charge of the detachments placed in garrison on the Ebro and in Upper Aragon. The whole column made up about 11,000 combatants.

The second column, consisting of the two auxiliary divisions--Palombini’s eleven Italian battalions and Compère’s 1,500 Neapolitans--took (without any artillery to hamper them) the mountain road by Alcañiz and Morella: they were slightly over 7,000 strong, and, if all went well, were destined to unite with the main body somewhere near Oropesa or Castellon de la Plana. It was not likely that this column would meet with much opposition.

But the third detachment, Harispe’s 5,000 men from Upper Aragon, who were to take the inland and western road by Teruel, were essaying a very dangerous task, if the enemy should prove active and enterprising, more especially as they had no artillery and hardly any cavalry with them. Blake might have taken the offensive with 20,000 men against them, while still leaving something to contain--or at least to observe--Suchet’s main column.

The Spanish Commander-in-Chief, however, did nothing of the sort, and met the invasion with a tame and spiritless defensive on all its points. When Suchet’s advance was reported, Blake had his forces in a very scattered situation. Of the 36,000 men of whom he could nominally dispose, the Empecinado’s ‘flying column’ was as usual detached in the mountains of Molina and Guadalajara, harassing small French garrisons. Zayas’s division had been left far to the south at Villena, near Alicante, to work off the contagion of yellow fever which it had contracted while passing by Cartagena. For in that port the disease was raging terribly at the time. Obispo’s division was in the high hills on the borders of Aragon. In the neighbourhood of Valencia were only the troops of Lardizabal and Miranda, with the main body of the cavalry. The Army of Murcia, which was destined to send succour if it should not find itself beset by Soult on the other side, was lying cantoned at various points in that province. As the French were at this time making no demonstration from the side of Granada, it now became clear that it would be able to send certain succours to Blake. But they were not yet designated for marching, much less assembled, and it was clear that they would come up very late.

This dispersion of the available troops did not, in the end, make much difference to the fate of the campaign, for Blake had from the first made up his mind to accept the defensive, to draw in his outlying detachments, and to stand at bay in the neighbourhood of Valencia, without attempting to make any serious resistance on the frontier. Since his arrival he had been urging on the construction of a line of earthworks, forming fortified camps, around the provincial capital. The ancient walls of Valencia itself were incapable of any serious resistance to modern artillery, but outside them, all along the banks of the Guadalaviar river, for some miles inland to the West, and as far as the sea on the East, batteries, _têtes-de-pont_, trenches, and even closed works of considerable size had been constructed. It was by holding them in force and with great numbers that Blake intended to check the invasion. In front of his chosen position, at a distance of twenty miles, there was a great advanced work--a newly restored fortress of crucial importance--the fastness of Saguntum, or ‘San Fernando de Sagunto’ as it had just been re-christened. This was the acropolis of one of the most ancient towns of Spain, the Saguntum which had detained Hannibal so long before its walls at the opening of the Third Punic War. In the age of the Iberians, the Carthaginians, and the Romans, and even down to the days of the Ommeyad califs, there had been a large and flourishing city on this site. But in the later middle ages Saguntum had declined in prosperity and population, and the modern town--which had changed its name to Murviedro (_muri veteres_) had shrunk down to the foot of the hill. It was now a small open place of 6,000 souls, quite indefensible. But above it towered the steep line of rock which had formed the citadel in ancient days: its narrow summit was crowned with many ruins of various ages--from cyclopean foundations of walls, going back to the time of the ancient Iberians, to Moorish watch-towers and palaces. The empty space of steep slope, from the acropolis down to the modern town, was also sprinkled with decaying walls and substructures of all sorts, among which were cisterns and broken roadways, besides the remains of a large Roman theatre, partly hewn out of the live rock.

There had been no fortifications by Murviedro when Suchet last passed near Valencia, in his abortive raid of March 1810[7]. On that occasion he had scaled the citadel to enjoy the view and to take a casual survey of the picturesque ruins upon it[8]. But since then a great change had taken place. On the advice, as it is said, of the English general, Charles Doyle[9], Blake had determined to restore the citadel as a place of strength. This was when he last held command in Valencia, and before he joined the Cadiz Regency. But his idea had been carried out after his departure by the Valencian Junta and the successive Captains-General who had come after him. By means of more than a year’s work the citadel had been made a tenable fortress, though one of an irregular and unscientific sort. The old Iberian and Moorish walls had been repaired and run together in a new _enceinte_, with material taken from the other ruins all around. In especial the Roman theatre, hitherto one of the most perfect in Southern Europe, had been completely gutted, and its big blocks had proved most useful for building the foundations of weak points of the circuit of fortification. This was strong at some points, from the toughness and height of the old ramparts, but very sketchy at others. Where the slope was absolutely precipitous, a rough wall of dry stone without mortar alone had been carried along the edge of the cliff. The narrow summit of the rock formed a most irregular enclosure, varying much in height from one point to another. It was divided into four separate sections cut off from each other by cross-walls. The westernmost and lowest, facing the only point from which there is a comparatively gentle ascent to the summit, was crowned by a new battery called by the name of _Dos de Mayo_, to commemorate the Madrid Insurrection of 1808. Rising high in the centre of this work was an ancient bastion named the Tower of San Pedro. Much higher, on the extreme peak of the summit, was the citadel tower, called San Fernando, where the governor’s flag flew, and from whence the whole fortress could be best surveyed. From this point the rock descended rapidly, and its long irregular eastern crest was surrounded by weakly-repaired walls, ending in two batteries called by the names of Menacho, the gallant governor of Badajoz[10], and Doyle, the English general who had suggested the fortification of the place. But the greater part of this eastern end of the works lay above slopes so precipitous that it seemed unlikely that they would ever be attacked. The western end, by the Dos Mayo battery, was the obvious point of assault by an enemy who intended to use regular methods.

[7] See vol. iii. pp. 284-6.

[8] Suchet’s _Mémoires_, ii. p. 156.

[9] See Arteche, xi. p. 123.

[10] See vol. iv. p. 56.

The construction was by no means finished when Suchet’s expedition began: many parts of the new walls were only carried up to half their intended height, and no regular shelter for the garrison had been contrived. Instead of proper barracks and casemates there were only rough ‘leans-to,’ contrived against old walls, or cover made by roofing in with beams old broken towers, and bastions. The hospital was the only spacious and regular building in the whole _enceinte_: the powder magazine was placed deep down in the cellars of the fort San Fernando. The armament of the place was by no means complete: the guns were being sent up just as Suchet started. Only seventeen were ready, and of these no more than three were 12-pounders: the rest were only of the calibre of field artillery (4- and 8-pounders) or howitzers. A fortress which has only seventeen guns for an _enceinte_ of 3,000 yards, and possesses no heavy guns to reply to the 18- or 24-pounders of a siege-train, is in a state of desperate danger.

Blake had thrown into the place a brigade under the command of Colonel Luis Andriani, consisting of five battalions, two each of the regiments of Savoya and Don Carlos, one of the Cazadores de Orihuela. Of these two were new ‘third battalions[11]’ from the recently raised ‘Division of Reserve,’ incomplete in officers, only half drilled, and not yet fully provided with uniforms. The total force came to 2,663 officers and men, including about 150 artillerymen and sappers. It is probable that these troops would have made no better show in the open field than did the rest of the Valencian army, a few weeks later: but they showed behind walls the same capacity for unexpected resistance which had surprised the French on other occasions at Ciudad Rodrigo, Gerona, and Figueras. Andriani, the governor, seems to have made an honourable attempt to do his duty at the head of the doubtfully efficient garrison placed at his disposal.

[11] The battalions were the 2nd and 3rd of Savoya (the last a new levy) the 1st and 2nd of Don Carlos, and the 3rd of Orihuela, this last raw and newly raised like the 3rd of Savoya.

In addition to Saguntum Blake held two outlying posts in his front, Peniscola on its lofty headland, garrisoned by about 1,000 men under General Garcia Navarro, and the half-ruined Oropesa, which he had resolved to hold, because it blocked the sea-coast road so effectively. But its only tenable points were two mediaeval towers, one in the town commanding the high-road, the other by the shore of the Mediterranean. Their joint garrisons did not amount to 500 men, and it was obvious that they could not hold out many days against modern artillery. But the gain of a day or two might conceivably be very valuable in the campaign that was about to begin. It is clear, however, that his main hope of resistance lay in the line of entrenched camps and batteries along the Guadalaviar, in front of Valencia: here he intended to make his real stand, and he hoped that Saguntum, so little distant from this line, would prove a serious hindrance to the enemy when he came up against it.

Suchet’s three columns all started, as Napoleon had ordered, on September 15th. The Marshal’s own main body, coming from Tortosa, reached Benicarlo, the first town across the Valencian frontier, next day, and on the 17th came level with Peniscola, whose garrison kept quiet within the limits of its isthmus. The Marshal left a battalion and a few hussars to observe it, and to see that it did not make sallies against his line of communication. On the 19th the head of the marching column reached Torreblanca, quite close to Oropesa. A reconnaissance found that the place was held, and came into contact with some Spanish horse, who were easily driven off. This was the first touch with Blake’s field army that had been obtained. But the enemy was evidently not in force, and the garrison of Oropesa hastily retired into the two towers which formed its only tenable positions. On a close inspection it was found that the tower in the town completely commanded the high-road, wherefore the Marshal took a slight circuit by suburban lanes round the place, with his main body and guns, and continued his advance, after leaving a few companies to blockade the towers. On the same evening he was joined by Palombini’s column from Morella, consisting of the two Italian divisions. They had accomplished their march without meeting any resistance, though the road from Morella by San Matteo and Cabanes was rough and easily defensible. The united force, now 16,000 strong, proceeded on its march next day, and the Marshal was agreeably surprised when, on the morning of the 20th, the cavalry scouts on his right flank announced to him that they had come in touch with Harispe’s column from Teruel, which had appeared at the village of Villafanes a few miles from the main road. Thus the whole army of invasion was happily united.

Harispe, as it turned out, had left Teruel on the 15th, in obedience to his orders, by the post-road to Segorbe and the coast. But hearing on the second day that a large Valencian force was holding the defile of Las Barracas, where the road crosses the watershed, he had turned off by a bad side-path to Ruvielos in the upper valley of the Mijares, in the hope of joining his chief without being forced to storm a difficult position. Blake, as a matter of fact, much alarmed at the approach of a flanking column on the Teruel side, and ignorant of its strength, had sent the division of Obispo and some other detachments to hold the pass. But no enemy came this way--Harispe had diverged down the course of the Villahermosa river, by a country road only practicable for a force without guns or wheeled transport, and got down by rapid marches to the coast-plain beyond Alcora, without having seen any enemy save some scattered guerrillero bands. He had thoroughly distracted Blake’s attention and had run no danger, because he took an unexpected and difficult route, in a direction quite different from that by which the Spaniards expected him to appear[12].

[12] Vacani says that the Teruel column was intended by Suchet as a mere demonstration, and was never intended to follow the high-road Teruel-Segorbe, but to take a cross-route over the hills, such as was actually used by it. But Suchet, in his _Mémoires_, makes no such statement (ii. p. 152), and speaks as if Harispe had taken the Ruvielos route on his own responsibility.

The whole army was now concentrated near Villafanes on September 21, save the detachments left to block Peniscola and Oropesa, and the brigade of Ficatier, which, escorting the siege-train, had been left at Tortosa, to await orders for starting when there should be no enemy left in northern Valencia to molest it. The heavy guns were to come forward down the coast-road, first to breach the towers of Oropesa, and when the way past them was clear, to play their part, if necessary, in the more serious task of battering Saguntum.

On advancing from Castellon de la Plana on September 22 the French army found a very small Spanish rearguard--500 or 600 men--covering the bridge of Villareal over the Mijares. They gave way before the first attack, which was a very simple affair, since the river was nearly dry and everywhere fordable. No more was seen of the enemy next day, and on the 23rd Suchet found himself on the banks of the Palancia stream, which flows under the foot of the rock of Saguntum. The Spaniards had retired still further towards Valencia, leaving the fortress to its own resources. These were unknown to Suchet, who was aware that the ruinous citadel had been rebuilt, but could not tell without further reconnaissance what was its strength. In order to invest the place, and to make closer investigation possible, Harispe’s division crossed the Palancia to the right of Saguntum, Habert’s to the left. The latter sent six companies into the town of Murviedro, and drove up some Spanish pickets from it into the fortress which towered above. The two divisions then joined hands to the south of Saguntum, completing its investment, while Palombini’s Italians took post at Petres and Gillet on the road to Segorbe--to the north-west--in case Blake might have placed some of his troops on this side-route, with the object of troubling the siege by attacks from the rear. The cavalry went forward down the high-road to Valencia, and sent back news that they had explored as far as Albalete, only six miles from the capital, and had met no enemy. The division of Lardizabal and the cavalry of San Juan, which had been the observing force in front of Suchet, had retired beyond the Guadalaviar river, and had shut themselves up (along with the rest of Blake’s army) in the entrenchments behind that stream. The Spanish general was evidently acting on the strictest principles of passive defence.

The French marshal determined not to seek his enemy on his chosen ground, till he should have taken Saguntum and brought up his siege-train to the front. The former condition he thought would not prove difficult to accomplish. A survey of the fortress revealed its extremely irregular and incomplete state of defence. Though the cliffs were in all parts steep and in some places inaccessible, many sections of the works above them were obviously unfinished and very weak. After a close reconnaissance by his engineer officers had been made, Suchet determined that it would be worth while to try an attempt at escalade on some of the most defective points, without waiting for the arrival of the siege-train. He set his sappers and carpenters to work to make sixty ladders, which were ready in full number on the third day. The front chosen for the assault was in the _enceinte_ immediately overhanging the town of Murviedro, where two ancient gaps in the wall were clearly visible; the new work was not half finished, and a low structure, roughly completed with beams laid above the regular foundations, was all that blocked the openings. The masons of the garrison were heard at night, working hard to raise the height of the stone wall which was to replace the temporary wooden parapets. There being no artillery available, they could not be hindered in their building: but it did not seem to advance very rapidly.

Suchet set apart for the actual escalade two columns, each composed of 300 volunteers from Habert’s division: they were to be supported by a reserve of similar strength under Colonel Gudin, which was formed up, completely under cover, within the streets of Murviedro. At midnight on September 27th-28th the stormers pushed forward under cover of the darkness, and in small successive parties, into a large Roman cistern above the ruined theatre, which was ‘dead ground,’ and not exposed to fire from any part of the ramparts. Here they were only 120 yards from the two breaches. Meanwhile, as a diversion, six Italian companies from Palombini’s division were ordered to make a noisy demonstration against the distant part of the defences which lay under the tower of San Pedro[13]. General Habert was to have 2,000 men more under arms, ready to support the assailing column.

[13] The complete orders for the attack may be read in the first _Pièce justificative_ in Belmas’s history of the siege, pp. 115-17 of vol. iv of his elaborate work.

The stormers reached their appointed place apparently undiscovered, and the attack would have been delivered--according to Suchet’s dispatch--without any preliminary firing, but for an accident. The Marshal says that the Spaniards had pushed an exploring patrol down the hillside, which fell in with the French pickets and drew their fire. Thereupon the assaulting columns in the cistern, thinking themselves discovered, let off a few shots and charged uphill, a little ahead of the appointed time, and before the Italian demonstration had begun[14]. The governor, Andriani, in his dispatch, makes no mention of this, but merely says that about 2 a.m. his sentinels thought that they detected movements on the slopes, and that a short time afterwards a fierce attack was delivered. At any rate the garrison was not surprised as Suchet had hoped.

[14] Vacani (v. p. 381) contradicts Suchet, saying that there was no Spanish patrol, and that the French pickets fired from nervousness at an imaginary foe.

Owing to the lowness, however, of the walls blocking the two old breaches, the assailants had, in their first rush, a fair chance of breaking in. Many ladders were successfully planted, and repeatedly small parties of the French got a footing on the wooden parapets. If the garrison had flinched, the storm might have succeeded: but far from flinching, they offered a desperate resistance, overthrew the ladders, slew all who had gained the top of the _enceinte_, and kept up a furious musketry fire, which laid low many of the soldiers who kept pressing forward to the breaches. It was to no purpose that the demonstration by the Italians below San Pedro now began: the Spaniards fired hard and fast in this direction also, but did not withdraw any men from the real point of attack, where they maintained themselves very courageously. It was in vain that Colonel Gudin brought up his reserve: it could make no head, and the survivors threw themselves down among the rocks and ruins in front of the wall--unwilling to recede, but quite unable to advance. Seeing his attack a hopeless failure, Suchet ordered the stormers back just before daylight began to appear. They had lost 247 killed and wounded out of 900 men engaged: the garrison only 15 killed and less than 30 wounded[15].

[15] Vacani makes the losses 360 instead of 247, and it is possible that Suchet has given only the casualties at the main assault, and not those in the distant demonstrations. Vacani says that the Italians lost 52 men in their false attack.

The escalade having come to this disappointing conclusion, the Marshal saw that the siege of Saguntum would be anything but a quick business. It would be necessary to bring up the siege-train to the front: orders were sent back to Ficatier to start it at once from Tortosa; but it had to batter and take Oropesa before it could even reach Murviedro. There were some weeks of delay before him, and meanwhile Blake might at last begin to show some signs of life. Suchet therefore disposed his army so as to provide both a blockading force and a covering force, to see that the blockade was not interfered with from without. It being evident that many days would elapse before the siege artillery arrived, the French engineer officers got leave to employ many detachments in preparing roads fit to bear heavy guns up the western slopes of the hill of Saguntum, from which alone the regular attack on the fortress could be conducted. Several emplacements for batteries were also chosen, and work upon them was begun.

From September 23rd, the day of Suchet’s arrival before Saguntum, down to October 16, when the heavy guns at last arrived, the French army was practically ‘marking time’: the idea which the Emperor had conceived, and which his lieutenant had adopted, that Valencia could be conquered by a sudden rush, had been proved false. Apparently Suchet had gained no more by his rapid advance to the foot of the hill of Saguntum than he would have obtained by marching in more leisurely fashion, with his siege artillery in company, and taking Oropesa on the way. The reduction of that place indeed was (as it turned out) only a single day’s task for heavy guns: and if the Marshal had captured it on his march, he might have presented himself before Saguntum with his siege-train, and have begun an active attack on that fortress, some weeks before he was actually able to get to serious work. In fact he might have been battering Saguntum on October 1, instead of having to wait till October 16th. But this is ‘wisdom after the event’: Napoleon thought that Valencia could be ‘rushed,’ and Suchet was bound to make the experiment that his master ordered.

Blake meanwhile, finding, on September 23rd, that the enemy was not about to advance against his lines, and learning soon after that the French army had settled down before Saguntum, had to revise his plans, since it was clear that he was not to be attacked in his entrenchments as he had supposed. Three courses were now open to him: either he might collect every man for a decisive battle in the open, and try to raise the siege; or he might attempt to open up attacks on Suchet’s line of communications and on his base in Aragon, so as to force him to retire by indirect operations; or he might remain passive behind the lines of the Guadalaviar. The last was an almost unthinkable alternative--it would have ruined his reputation for ever to sit quiet and do nothing, as Wellington had done during the siege of Ciudad Rodrigo in 1810. Only a general with an established reputation for courage and ability could have dared to take such a course; and Blake’s record was a long series of disasters, while he was detested by the Valencians one and all--by the army, to whom he rightly preferred his own excellent troops, no less than by the Captain-General Palacios, and the Junta, whom he had sent out of the city to sit at Alcira, when they showed a tendency to hamper his operations. Practically he was forced by his situation to take some definite offensive move against Suchet.

He chose that of indirect operations, having a well-rooted distrust of the fighting powers of a great part of the troops that were at his disposition. The record of the Valencian army he knew: the state of the Murcian army, on which he could draw for reinforcements, was represented to him in the most gloomy colours by Mahy, who had recently replaced Freire in command. On September 12th Mahy had written to him, to warn him that the spirit of his troops was detestable: ‘the Army of Murcia was little better than a phantom: there were only four or five officers for whom the rank and file had any respect or esteem, the rest were regarded as timid or incapable: the men had no confidence in themselves or their chiefs. The best thing to do would be to break up the whole army, and incorporate it into the “Expeditionary Divisions,” whose commanders were known as good soldiers, and whose battalions were trustworthy[16].’

[16] See Mahy’s letter to Blake on pp. 109-12 of vol. xi of Arteche. The General is writing very carefully so as not to speak too ill of his army: but his views are clear.

In view of these facts Blake resolved to threaten Suchet’s flanks with demonstrations, which he had no intention of turning into attacks, but to endeavour to dislodge him from his forward position by turning loose the guerrilleros of Aragon on to his rear. With the former purpose he sent out two detachments from the Valencian lines, Obispo’s division to Segorbe,--where it cut the French communication with Teruel and southern Aragon,--Charles O’Donnell with Villacampa’s infantry and San Juan’s horse to Benaguacil, a point in the plains fifteen miles west of Saguntum, where his force formed a link between Obispo and the main body of the Valencian army, which still remained entrenched in the lines of the Guadalaviar[17]. These two detachments threatened Suchet’s flank, and even his rear, but there was no intention of turning the threat into a reality.

[17] Blake kept under his own hand in the lines the divisions of Zayas, Lardizabal, Miranda, and the Reserve.

The real movement on which Blake relied for the discomfiture of the invaders of Valencia was that of the guerrillero bands of Aragon and the neighbouring parts of Castile, to whom he had appealed for help the moment that Suchet commenced his march. He believed that the 6,000 or 7,000 men which Suchet had left scattered in small garrisons under General Musnier might be so beset and worried by the _partidas_, that the Marshal might be compelled to turn back to their aid. Even Mina from his distant haunts in Navarre had been asked to co-operate. This was an excellent move, and might have succeeded, if Musnier alone had remained to hold down Aragon. But Blake had forgotten in his calculations the 15,000 men of Reille and Severoli, cantoned in Navarre and along the Upper Ebro, who were available to strengthen the small force which lay in the garrisons under Musnier’s charge.

The diversion of the guerrilleros, however, was effected with considerable energy. On September 26th the Empecinado and Duran appeared in front of Calatayud, the most important of the French garrisons in the mountains of western Aragon. They had with them 5,000 foot and 500 horse--not their full strength, for a large band of the Empecinado’s men beset at the same time the remote castle of Molina, the most outlying and isolated of all Suchet’s posts. Calatayud was held by a few companies of French, to which an Italian flying column of a battalion had just joined itself. The guerrilleros, coming in with a rush, drove the garrison out of the town into their fortified post, the large convent of La Merced, taking many prisoners in the streets. Duran then beleaguered the main body in the convent, while the Empecinado took post at the defile of El Frasno on the Saragossa road, to hold off any succour that Musnier might send up from the Aragonese capital. This precaution was justified--a column of 1,000 men came out of Saragossa, but was far too weak to force the pass and had to retire, with the loss of its commander, Colonel Gillot, and many men. Meanwhile Duran pressed the besieged in the convent with mines, having no artillery of sufficient calibre to batter its walls. After blowing down a corner of its chapel with one mine, and killing many of the defenders, the guerrillero chief exploded a second on October 3, which made such a vast breach that the garrison surrendered, still 560 strong, on the following day[18].

[18] Vacani gives a long and interesting account of the siege (v. pp. 404-13) and attributes the weak defence to quarrels between the commander of the Italians and the French governor, Müller.

This success would have gone far to shake the hold of the French on Aragon, but for the intervention of Reille from Navarre. At the first news of the blockade of Calatayud, he had dispatched a column, consisting of the whole brigade of Bourke, 3,500 strong, which would have saved the garrison if it had had a less distance to march. But it arrived on the 5th to find the convent blown up, while the Spaniards had vanished with their prisoners. Bourke thereupon returned to Tudela, and the guerrilleros reoccupied Calatayud on his departure.

Meanwhile, however, the whole Italian division of Severoli, over 7,000 strong, marched down the Ebro to reinforce the small garrison of Saragossa. This large reinforcement restored the confidence of the French. Musnier himself took charge of it and marched at its head against Duran and the Empecinado. They wisely refused to fight, gave way, evacuated Calatayud, and took refuge in the hills (October 12). While the main field-force of the enemy was drawn off in this direction, Mina took up the game on the other side of the Ebro. Entering Aragon with 4,000 men he besieged the small garrison of Exea, which abandoned its post, and cut its way through the guerrilleros, till it met a column of 800 Italian infantry[19] sent out from Saragossa to bring it off. Colonel Ceccopieri, the leader of this small force, underrating the strength of his enemy, then marched to relieve the garrison of Ayerbe. He was surprised on the way by Mina’s whole force, and in a long running fight between Ayerbe and Huesca was surrounded and slain. The column was exterminated, two hundred Italians were killed, six hundred (including many wounded) were taken prisoners (October 16th).

[19] Belonging to the 7th Line of Severoli’s division.

Musnier returned in haste from Calatayud at the news of this disaster, but left the bulk of Severoli’s division to occupy western Aragon. He then set himself, with the help of Reille, to hunt down Mina. But the latter, marching with ease between the columns that pursued him, for the peasantry kept him informed day by day of every movement of the enemy, retreated westward. Easily eluding the French, he made an extraordinary excursion, right across Navarre, Alava, and Biscay, down to the sea coast at Motrico, where he handed over his prisoners to the captain of the British frigate _Isis_, and then returned unharmed to his familiar haunts. Of such a delusive nature was the hold of the French on Northern Spain, that a column of 5,000 men could march for 200 miles across it without being intercepted or destroyed.

All these exploits of the guerrilleros were daring and well planned, but though they had given Musnier much trouble, and cost the French many a weary hour of march and countermarch, they had not cleared Aragon of the enemy, nor shaken Suchet’s position. Indeed, on October 20, the general condition of affairs in Aragon was more favourable for the invaders than on September 20, for two fresh divisions had been drawn down into that province, and there were 20,000 French and Italian troops in it instead of 6,000. The petty disasters at Calatayud and Ayerbe were irritating rather than important. Suchet never for a moment felt inclined to relax his hold upon Valencia: that western Aragon was in an uproar affected him little, when his communication with his two main dépôts of stores at Tortosa and Morella was not interrupted.

Blake, it may be mentioned, did not content himself with setting the Empecinado and Duran in motion, he tried another division in another quarter with even less result. Rumours had reached him that King Joseph’s Army of the Centre was about to co-operate with Suchet, by sending a column across the mountains to Cuenca and Requeña. The news was false, for though Napoleon had ordered the King to do what he could to help in the invasion of Valencia, Joseph had replied that he had not even one brigade to spare for a serious demonstration, and had not moved--the guerrilleros gave sufficient occupation to his much-scattered army, of which a large portion was composed of untrustworthy Spanish _Juramentados_. But, listening to vain reports, Blake ordered Mahy to collect the best of his Murcian troops and to march on Cuenca to meet the supposed invaders. His subordinate, leaving Freire in command in Murcia, took seven selected battalions of foot under Creagh and the Marquis of Montijo, with 800 horse and one battery, and moved from his camp at Mula by Hellin and Chinchilla northward. The distance to be covered was great, the roads after Chinchilla very bad. Mahy arrived in front of Cuenca on October 15th, to find that there was only one battalion and two squadrons of Joseph’s army there. This little force evacuated the high-lying city in haste, and fled towards Madrid the moment that the Murcians showed themselves. No other French force could be heard of in any direction. At Cuenca Mahy received a dispatch from Blake (who had apparently discovered his mistake about the Army of the Centre), telling him to descend from the mountains by Moya and Liria, and to join the wing of the main army, which lay under Obispo at Segorbe. It was only on the 23rd October that he came in: his troops, the pick of the Murcian army, had been completely wasted for some twenty days in a circular march against a non-existent enemy. Meanwhile every man had been wanted in Valencia.

Suchet, when once he had settled down to the siege of Saguntum, had not failed to notice Blake’s weak demonstration against his flank by means of the divisions of Obispo and Charles O’Donnell. He did not intend to tolerate it, and on September 30 had sent Palombini with his own Italian division and Robert’s French brigade to beat up Obispo’s quarters at Segorbe. The Spanish division made a poor attempt to defend itself on a position in front of that town, but was easily beaten and retired into the mountains. It was then the turn of Charles O’Donnell; when Palombini had come back to the camp, Suchet took Harispe’s division, with Robert’s brigade, and two regiments of cavalry, to evict the Spanish division from Benaguacil. O’Donnell made a slightly better fight than Obispo had done, and deployed Villacampa’s infantry behind an irrigation canal, with San Juan’s cavalry on his flanks. But the French were superior in numbers as well as in confidence: one fierce charge broke O’Donnell’s line, and he had to retreat in haste to the hills behind him, losing 400 men, cut up in the pursuit by Suchet’s cavalry, while the French casualties barely reached three officers and sixty men (October 2nd). Blake, who had been quite close enough to succour O’Donnell if he had chosen, made no attempt to aid him, and kept quiet behind his lines on the Guadalaviar. There the routed troops joined him next day.

Suchet, having thus cleared his flanks, settled down to the siege of Saguntum, where his heavy artillery was now much needed. The besieging army had to content itself for another fortnight with making preparations for the expected train--levelling roads and constructing approaches on the ground which was destined for the front of attack, at the west end of the hill of Saguntum.

Meanwhile the siege-train was lumbering down from Tortosa by the coast-road. On October 6th Suchet started to meet it, taking with him the 1,500 Neapolitans of Compère. On the 8th he reached Oropesa, where he found the small Spanish garrison still holding the two towers which have before been mentioned. The first guns that came up were turned against the tower by the high-road; it was easily breached, and on the 10th surrendered: 215 men and four guns were captured. Next day came the turn of the other tower, that by the sea; but before the siege-battery had opened on it, the British 74 _Magnificent_ and a squadron of Spanish gunboats ran inshore, and took off the garrison of 150 men in their boats, under the ineffective fire of the French.

The moment that the tower which blocked the high-road had fallen, and before that on the shore had been evacuated, Suchet began to push the head of his precious convoy of heavy artillery southward. It made such a good pace that the first guns arrived at the camp before Saguntum as early as the night of October 12th. Meanwhile the Marshal himself returned thither, escorted by Compère’s Neapolitans: the brigade of Ficatier, which had escorted the train hitherto, was dispersed to cover the line of communications, placing its five battalions at Oropesa, Almenara, and Segorbe.

SECTION XXX: