Chapter XXXI: Section XXVIII: Chapter II (1)

THE SIEGE AND FALL OF TARRAGONA. MAY-JUNE 1811

Suchet had marched, as has been already mentioned, from Lerida, with Harispe’s division, on April 28th, Frère’s division following. On the 29th the head of the column reached Momblanch, where half a battalion was left behind in a fortified post, to keep open the Lerida road. On May 2nd the large manufacturing town of Reus, only ten miles from Tarragona, was occupied: on May 3rd the French advanced guard, Salme’s brigade, approached the city, and drove in the Catalan advanced posts as far as the river Francoli. But the siege could not begin till Habert’s force, escorting the battering-train, should come up from Tortosa; and this all-important column was much delayed. Its road ran along the seaside from the Col de Balaguer onward, and Codrington’s squadron of English frigates and gunboats accompanied it all the way, vexing and delaying it, by bombarding it whenever it was forced to come within gunshot of the beach. This was practically all the opposition that Suchet met with: a few miqueletes had shown themselves in the hills between Reus and Momblanch, but they were too weak to fight. Campoverde had carried off the best both of regulars and irregulars to the relief of Figueras, and Courten, who had barely 4,500 men in his division[638], had wisely shut himself up in Tarragona, where every man was wanted: for the enceinte was very long, and the sedentary garrison consisted of only five or six battalions. The troops inside the walls did not amount, when the siege began, to 7,000 men: hence came the weakness shown in the early days; it was not till Campoverde’s army came back from the north (May 10) that an adequate defensive force was in existence for such a large fortress.

[638] The composition of the divisions of the Army of Catalonia
was shifting, and hard to follow, but (as far as I can make out)
Courten’s division consisted of three battalions of Granada,
two each of Almanza, Almeria, and America, while the sedentary
garrison contained four or five battalions of the new Catalan
‘Sections’ or ‘local line,’ besides a battalion of Voluntarios de
Tarragona.

Tarragona, though some of its fortifications were not skilfully planned, was a very strong place. The nucleus of the works was the circuit of the old Celtiberian town of Tarraco, which afterwards became the capital of Roman Spain. This forms the upper city in modern times. It is built on an inclined plane, of which the eastern end (530 feet above sea-level), where the cathedral lies, is the higher side, and the slope goes downhill, and westward: the southern face, that towards the sea, is absolutely precipitous, the northern one hardly less so. Large fragments of the Cyclopean walls built by the Celtiberians, or perhaps by the Carthaginians, are visible along the crest on both of these sides. On the west, the lowest part of the old town, a line of modern fortifications divided the upper town from the lower; there was a sharp drop along this line: in most places it is very steep, and the road of to-day goes up the hillside in zigzags, to avoid the break-neck climb[639]. Below the fortifications of the upper city, and divided from them by a broad belt of ground free of houses[640], lay the port-town or lower city, clustering around the harbour, which is an excellent roadstead shut in by a mole 1,400 feet long, which runs out from the south-west corner of the place. The lower city was enclosed on its northern and western sides by a front of six bastions; its southern side, facing the port and the open sea, had not, and did not need, any great protection; it could only have been endangered by an enemy whose strength was on the water, and who could bring a fleet into action. There was a sort of citadel in the port-town, a work named the Fuerte Real, which lies on an isolated mound inside the north-west angle of the walls. About 400 yards west of the most projecting bastion of the place the river Francoli flows into the sea, at the western end of the harbour. In the angle between the river and the port was an outlying work, Fort Francoli, destined to keep besiegers away from the shipping, which they might easily bombard from this point, if it were not occupied. This fort was connected with the lower town by a covered way protected by a long entrenchment containing two lunettes.

[639] Apparently there was in 1811 no road of this sort, up
the steep slope above the railway station of to-day; the
main _chaussée_ from Valencia entered the upper city at its
north-western end, and there was no good road for carriages
up the south-western point, as there is now (the so-called
Despeñaperros).

[640] Long since built over. The line of the old fortifications
of the upper city is now marked by a broad promenade, the Rambla
de San Juan.

Notwithstanding the great strength of the high-lying upper city, it had been furnished with a second line of defence, outside its old Roman walls. Low down the hillside five forts, connected by a wall and covered way, protected its whole eastern front from the edge of the heights as far as the sea. The Barcelona road, crawling along the water’s edge, enters the place between two of these forts, and goes to the Lower, after sending a steep bypath up to a gate in the Upper, city[641].

[641] It is a stiff climb up a very steep ascent to enter the
upper city by its ‘Barcelona Gate.’

On the west and north-west the high-lying fortress commands all the surrounding country-side. But to the due north there is a lofty hill about 800 yards from the walls, called Monte Olivo. This dominates the lower town, since it is 200 feet high, or more, though it is itself dominated by the upper town. An enemy in possession of it has every advantage for attacking the north front of the lower town. Wherefore, during the course of the last two years, the summit of the hill had been entrenched, and a very large hornwork, the Fuerte Olivo, constructed upon it. This was a narrow fort, following the shape of the crest of the hill, with a length of 400 yards, and embrasures for forty-seven guns. Its outer front was protected by a ditch hewn in the solid rock: its rear was only slightly closed with a low wall crowned by palisades, so as to leave it exposed to the fire of the upper city, if by any chance the enemy should get possession of it. Such an extensive work required a garrison of over 1,000 men--a heavy proportion of the 6,500 which formed the total force of the Spaniards at the commencement of the siege.

When Suchet arrived in front of Tarragona, and had driven the Spaniards within their works (May 3rd-4th), his chief engineer and artillery officers, Rogniat and Vallée, had to conduct a long and careful survey of the fortifications opposed to them. They concluded that the northern front of the city was practically impregnable, from its precipitous contours, and that the eastern front, though a little less rocky, was equally ineligible, because of the trouble which would be required to transport guns first across the high ground to the north-east, and then down to the seashore. The south front, being all along the water’s edge, was inaccessible. There remained only the western front, that formed by the lower city, where the defences lay in the plain of the Francoli, and had no dominance over the ground in front of them. There was an additional advantage for the besieger here, in that the soil was partly river sand, partly the well-broken-up loam of suburban market gardens, and in all cases very easy to dig. But if they were to attack the west front, the engineers required the General-in-Chief to accomplish two preliminary operations for them. He must take Fort Olivo, which commanded with its flanking fires much of the ground on which they intended to work, and he must drive away from the northern side of the harbour the Anglo-Spanish squadron which lay there, since its heavy guns would enfilade all works started for the purpose of approaching the western front of Tarragona in the neighbourhood of the mouth of the Francoli.

This being the programme laid down, Suchet took up his positions round the fortress--Harispe’s division had charge of the main part of the northern front, its French brigade (Salme) occupying the ground in front of Fort Olivo, while its two Italian brigades stretched eastward along the distant heights, curving round so as to cut the Barcelona road along the sea-coast with their extreme detachment. Frère’s division had the central part of the lines, and lay on both sides of the course of the Francoli river, its main force, however, being on the left bank. Habert’s division, which had just come up from Tortosa, was placed near the mouth of the river, and facing towards the port; it formed the right wing of the army, and covered the siege-park, which was established at the village of Canonge, about a mile and a half from the walls of Tarragona. The magazines and hospitals were fixed at the large town of Reus, nine miles to the rear, under a considerable guard; for though the road from thence to the French lines ran over the gentle undulations of the coast plain, yet there was always danger that bands of miqueletes might descend from the hills for some daring enterprise. Several of the intermediate villages were fortified, to serve as half-way refuges for convoys and small parties on the move.

Some days were lost to the French in completing the survey of Tarragona, in settling down the troops into their permanent camps, and in bringing up from the rear, along the Tortosa road, the remainder of the battering-train and its munitions. It was not till May 8th that serious operations began. Suchet’s first object was to drive away from the northern end of the harbour the English and Spanish ships, whose fire swept the ground about the mouth of the Francoli, across which his siege-works were to be constructed. With this object a large fort was constructed on the shore, in which very heavy guns, fatal to shipping, were to be placed. Commodore Codrington, who was lying in the harbour with a small squadron of two 74’s and two frigates, assisted by several Spanish gunboats, bombarded the fort incessantly, but what he destroyed in the day the French rebuilt with additions every night, and on May 13th the fort was sufficiently completed to receive its armament of 24-pounders. The ship-guns were unable to cope with them, and the vessels of the Allies during the rest of the siege were compelled to keep to the south end of the port, and could only vex the besieger’s subsequent trench-building by a distant and ineffective fire. On the 16th a first parallel, directed against the most advanced Spanish work, Fort Francoli, was begun in the low ground beside the new fort.

Before this check to the squadron had been completed a great change in the situation was made by the arrival of Campoverde on May 10, with 4,000 regular troops brought by sea from Mataro--fractions of the divisions of Eroles and Sarsfield, though neither of these generals had come in person[642]. The garrison being strengthened up to 10,000 men, and raised in morale by the reinforcement, became very bold and enterprising. Sorties began almost at once: Harispe’s division having seized on the 13th May two slight outlying entrenchments below Fort Olivo, three battalions sallied out on the 14th and made a desperate attempt to retake them. It failed, but on the 18th an equally vigorous sortie was made against the fort beyond the Francoli, and the first parallel near it, by about 2,000 men, who drove in the trench-guards and destroyed a section of the works, but were finally thrust back into the lower city by the arrival of reinforcements led by General Habert. How hot the fighting had been here is shown by the fact that Suchet’s dispatch owns to a loss of over 150 men, with three officers killed and eleven wounded. The sallying force lost 218, a figure which Suchet enlarges in his report to 250 killed and 600 wounded. On the 20th the Spaniards made a third sally, on a different front, far to the north-east, across the high ground north of the Barcelona road, and tried to break through the line of blockade kept up by Harispe’s Italian brigades. This was on a smaller scale, and had no luck; it was apparently intended to open up communication with Sarsfield, who (marching by circuitous ways across central Catalonia) had reached Valls and Alcover, only ten miles from Tarragona, on the upper Francoli, with 1,200 men. This trifling force was to be the nucleus of an ‘army of relief’ which was to be collected from all quarters to threaten Suchet’s rear. Sarsfield made his appearance known to his chief in Tarragona by lighting beacons on the mountain tops. Learning that the Spanish force was insignificant, Suchet detached two battalions and some cuirassiers to drive Sarsfield further away from Alcover, and did so with small loss, forcing him to retire to the mountains above Valls.

[642] The regiments that landed with Campoverde seem to have been
the 2nd of Savoia (2 battalions), Voluntarios de Gerona, and two
of the Andalusian regiments which had formed the core of Reding’s
old Granadan division, which marched to Catalonia in 1808, viz.
Iliberia and Santa Fé, the first three battalions strong, the
other with two.

About this time the French artillery and engineer commanders reported to their chief that it would be at least ten days before they were in a position to begin a serious attack against the western front of the city, and Suchet resolved that the enforced delay should be utilized for an attack on Fort Olivo, whose capture would sooner or later be a necessity, if the main operations against the city were to prosper.

Accordingly, while the approaches against the west front went steadily on, a separate offensive advance against the Olivo was prepared. Between the 22nd and the 28th of May trenches were pushed towards the fort, and batteries containing thirteen guns erected to bear upon it. Their fire had effected serious damage on the parapets and the artillery of the fort by the 29th, yet the engineers reported that they could not fill the ditch, which was dug in the solid rock, and could not promise to make accessible breaches beyond it. But they reported that the rear face of the work, which the French artillery could not reach, was very weak, the low wall and palisade closing the gorge being no more than nine feet high. There was also a gap in the front protection caused by the entry, into the right end of the fort, of an aqueduct which carried water down into Tarragona. This structure made a sort of bridge across the ditch; it had not been cut, but only closed with palisades, which were being rapidly demolished by the French cannonade[643].

[643] This discovery was the work of the Italian engineer officer
Vacani, whose work on the campaign of the Italian troops of
Napoleon in Spain is one of the most valuable of our original
sources. See his vol. v. pp. 175-6. He was with the assaulting
column.

On the night of the 29th Suchet made the rather rash venture of trying to escalade Fort Olivo at the two weak points. One column was to turn the work under cover of the darkness, and to endeavour to break in at the gorge in its rear. The other was to try the imperfect breach in the right front, by crossing the aqueduct, though it was only seven feet broad, if it should be found that the ditch was impassable. Meanwhile a general demonstration was to be made by scattered _tirailleurs_ against the whole face of the Olivo, so as to distract the attention of the enemy, and the batteries down by the Francoli were to bombard the lower city with the same purpose. Both attacks were successful--more by luck than by their deserts, for the plan was most hazardous. The column which had gone round to the rear of the fort ran in upon a Spanish regiment[644] which was coming up the hill to relieve the garrison. The two forces hustled against each other in the dark, and became hopelessly mingled in a close combat just outside the postern gate of the gorge. The garrison was unable to fire upon their enemies, because they were intermixed among their friends, and, when the fight surged against the postern and the palisades, the French succeeded in entering the gorge, some by scrambling up the low and weak defences, others by bursting in at the gate along with the Spanish reinforcements with whom they were engaged. They might have been checked, for the defenders were fighting fiercely, if the other attack had not also succeeded. But at the right front of the fort, where the second assault was made, though many of the forlorn hope fell into the ditch, a desperate charge took the storming-party across the seven-foot gangway of the aqueduct, and over the shattered palisades that blocked it. The garrison could tell by the noise of the musketry that the enemy had entered both in front and in rear, and were stricken by despair[645]. But the greater part of them clubbed together and continued a desperate resistance, which was only subdued when Suchet sent in all his reserves and the trench-guards to back the stormers. They were then beset on all sides, and finally overwhelmed.

[644] Three battalions of Iliberia were holding the fort on
the 29th; they had been much tried, and two battalions of the
sister-regiment of Almeria were coming up to relieve them.

[645] Vacani is very positive that the stormers at the aqueduct
got into the fort before those at the gorge.

The losses of the garrison were terrible--of the five battalions of Iliberia and Almeria, and the two companies of artillery which had been engaged--some 3,000 men in all--very nearly one-third, as it would appear, were slain or captured[646]. The prisoners were about 970 in number, many wounded, including the commander of the fort, Colonel Gomez, who had received no less than ten bayonet stabs. Three or four hundred men had been killed--the French had given little quarter during the earlier part of the fighting. The remainder of the garrison had escaped into the city, by climbing over the low wall of the gorge and running down the slopes, at the moment of the final disaster. The French loss, according to Suchet, was only about 325 killed and wounded, and probably did not greatly exceed that figure. The assailants had, it must be confessed, extraordinary luck. If the turning column had not become mingled with the Spanish reinforcements it might never have been able to break into the gorge; while the other attack could not have succeeded if the governor had taken the proper precaution of cutting the aqueduct, which served the stormers as a bridge--for the ditch proved wholly impracticable, and the breach could not be approached.

[646] Suchet says that except 70 officers and 1,000 men taken
prisoners ‘the whole of the rest of the garrison had perished’
(ii. 60). Belmas (iii. 502) speaks of 970 prisoners and says that
1,200 were killed, but acknowledges that ‘some of the Spaniards’
got away. Vacani (v. 187) says that ‘many’ Spaniards escaped,
but that the bulk of six battalions were destroyed, 1,000 being
captured and 1,200 slain. The governor, Contreras, says that
there were 4,000 men in the fort, and that somewhere about 2,000
were killed or wounded (p. 248). But the figures must have been
lower: Iliberia was about 1,500 strong, Almeria about 1,200:
there were also 200 gunners in the fort: the total garrison
therefore was about 3,000. But at the end of the siege, a month
later, Iliberia surrendered 368 unwounded men, and Almeria 464.
They must have lost many hundreds during the last six weeks of
the leaguer, yet were still 832 strong. It is hard to see that
they can have lost more than 1,200 or 1,300 between them on May
29, and very probably Toreno and Arteche are right in putting the
total loss at only 1,100 and odd. If so, the killed, including
the gunners, must have been only between 300 and 400. The
ever-accurate Schepeler gives 1,200 for the total loss (p. 433),
and I suspect is nearest of all to the truth. Napier, as usual,
merely reproduces Suchet’s figures.

On the morning after the assault the spirit of the Spaniards was so little broken that a sortie was made with the purpose of retaking the Olivo, the survivors of the two regiments which had lost it volunteering to head the attack. Campoverde thought that the French might be caught before they had made new defences to protect the weak rear face of the fort, but they had built up the entry of the gorge with sandbags, and the assault--led by Colonel O’Ronan, a Spanish-Irish officer--was beaten off with loss, though a few daring men not only reached the gorge, but scrambled in through its broken palisades to die inside the work. All the guns of the upper city were then turned upon the Olivo, and reduced its rear to a shapeless mass of earth. But this did not seriously harm the French, who burrowed into its interior and made themselves strong there. They only wanted to be masters of the hill because it flanked their projected approaches in the low ground, and did not intend to use it as their base for any further active operations.

After the Olivo disaster Campoverde held a council of war (May 30th), and announced to his officers that the means by which Tarragona could be saved was the collecting of a great army of succour to fall upon Suchet’s rear. He was himself about to depart, in order to take command of it; Sarsfield’s and Eroles’s small detachments, all of which he would collect, must form its nucleus. The _somatenes_ of all central Catalonia should be called in, and the province of Valencia had promised to lend him a whole division of regulars. So saying, he departed by sea along with his staff and a number of the richer inhabitants of the city (May 31st). General Caro, who had hitherto acted as governor, was sent to hurry up the Valencians, and the command of the place was made over to an officer newly arrived from Cadiz, General Juan Senen Contreras[647], who by no means liked the task assigned to him. The garrison was still 8,000 strong, for just after the fall of the Olivo two battalions of regulars arrived from Valencia[648]--the first-fruits of the succours promised from that province--and a draft of 400 recruits landed from Majorca. It seems to have been a mistake of Campoverde to come to Tarragona at all--his presence would have been much more valuable in the interior, where a supreme commander was much wanted, and while he was shut up in the fortress (from May 10th to May 31st) little had been done outside. The Junta of Catalonia, now sitting at Montserrat, had been issuing many proclamations, but had not accomplished much in the way of gathering in the _somatenes_.

[647] See his pamphlet on the defence of Tarragona and his own
responsibilities, printed in the 3rd volume of _Mémoires sur la
Guerre d’Espagne_.

[648] Apparently 3rd battalion of Cazadores de Valencia, and 1st
battalion of the _First_ regiment of Savoia. This last must be
carefully distinguished from the 1st battalion of the _Second_
regiment of Savoia, which belonged to the Catalan army, and had
already been brought into Tarragona on May 10th by Campoverde.
See the history of the two in the Conde de Clonard’s colossal
work on the regimental histories of the Spanish line.

On June 3rd Campoverde reached Igualada, and established his head quarters there, but found only 3,000 men assembled under Sarsfield. He sent that general off to Tarragona, to act as second in command to Contreras, and took over charge of his few battalions; by calling in Eroles, and hunting up deserters and detachments, he had collected in a fortnight 5,280 regular infantry and 1,183 cavalry--all that there were of mounted men in Catalonia. The whole was much too small a force to justify him in attacking Suchet in his lines--even when the _somatenes_ should come in to join him. All depended on the expected succours from Valencia, and they were slow in arriving. Charles O’Donnell, the newly appointed Captain-General of that province, had made up a scheme for drawing off Suchet by attacking his garrisons in southern Aragon, and had gone off early in May with his main force against Teruel. This scheme had no effect whatever; Suchet had fixed his teeth into Tarragona, and was not to be distracted by any demonstrations against his more distant detachments. Campoverde grew so desperate that he offered to give over supreme command to O’Donnell, if the latter would come into Catalonia with his whole disposable force, and begin by attacking Mora, Suchet’s great dépôt on the Ebro. The Valencian Captain-General, though he refused to take this responsibility, finally agreed to send a division of regulars under General Miranda by sea to join the Catalans. This force, about 4,000 strong, appeared at Tarragona on June 14th, and came ashore, but was immediately afterwards reshipped by Campoverde’s orders, and transferred to Villanueva de Sitjes, where it landed, and marched inland to Igualada to join the ‘army of succour,’ which by its arrival was raised to nearly 11,000 regular troops.

While Campoverde was slowly beating up his reinforcements, Tarragona was already in grave danger. The formal attack on the lower city began on the night of June 1st, when Vallée and Rogniat, the commanders of the French artillery and engineers, declared that they had everything ready. The front selected for attack was the two south-westerly bastions, those called San Carlos and Orleans, but as a preliminary task it was necessary to drive the Spaniards out of the subsidiary and external defence formed by the outlying Fort Francoli, at the mouth of the river, and by the long entrenchment which joined it to the city, with the lunette of the Prince, a very small work, inserted in its midst. For Fort Francoli had a position which would enable it to enfilade the French trenches when they should draw near to the enceinte of the city.

On the night of June 1st the French threw up their first parallel at a distance of only a little over 300 yards from the bastion of Orleans: it was connected with the entrenchments beyond the Francoli by a zigzag trench. On the second night the parallel was completed for a length of 600 yards, and three batteries begun in it--one directed against the lunette of the Prince and the line joining it to Fort Francoli, the other two against the bastion of Orleans and the adjacent curtain. On June 3rd the besiegers began to work forward by a flying sap towards the fort, and by the 7th had pushed their front trench to within twenty yards of the work. On that same day the artillery began to play against it, not only from the new batteries, but from the old ones beyond the river, which had previously been directed against the fleet. The fort was weakly built, and a practicable breach was made in its left face before the bombardment had been twelve hours in progress. Serious damage had also been done to the long entrenchment connecting Fort Francoli with the lower city. Contreras, rightly regarding the work as untenable, ordered its serviceable guns to be removed the moment after dusk set in, and bade its commander, Colonel Roten, to draw off the garrison, two battalions of the regiment of Almanza. They withdrew at 8.30, and an hour and a half later three French columns charged out of the trenches and seized the fort[649]. They were surprised to meet with no resistance, not having detected the withdrawal of the Spaniards. Finding themselves unopposed, they tried to push along the entrenchment from the fort towards the town, but were stopped, with some loss, by the guns of the Prince lunette.

[649] Suchet in his memoirs conceals the fact that the fort was
already abandoned when his troops entered (ii. p. 72). But the
evidence of Contreras, Schepeler, and other Spanish authorities
is clear and unanimous.

The Spanish engineers had assured Contreras that the low-lying Fort Francoli would be untenable under the fire of the neighbouring bastion of San Carlos, the battery on the Mole on the other side of the harbour, and the heavy guns of the men-of-war. A fierce fire was opened against it from all these quarters, but proved insufficient to stop the French from burrowing into the ruins of the fort, connecting it with their trenches, and finally building in its right front a heavy battery, which bore along the line of the entrenchment and enfiladed the Prince lunette. This work faced northward, and exposed only a weak flank to the attack. Fort Francoli having ceased to be an obstacle, the besiegers could now throw out a second parallel from the first, which they had constructed in front of the bastion of Orleans. Five new batteries were placed in it, some bearing on Orleans, some on San Carlos, and one having the special task of beating down the Prince lunette. The Spanish guns in the lower city answered with a fierce fire which caused much damage and took many lives, but the work, nevertheless, went on unceasingly. On June 16th all the new batteries were ready to commence their work.

Contreras had been much chagrined by the complete failure of the best efforts of his artillery to hold back the advance of the enemy, and reports that the morale of the troops was disagreeably affected by the arrival of the Valencian division of Miranda on June 14th and its prompt departure, after staying less than two days in the place. The garrison had looked upon it as a seasonable reinforcement, and were dashed in spirits when it made no stay with them. It seems to have been a complete mistake to have brought these 4,000 men to Tarragona at all: they should have been landed at once in Villanueva de Sitjes to join the army of succour.

Nevertheless the governor did his best to delay the progress of the French attack, and when his artillery proved ineffective, sent out two strong sorties on the 11th and 14th, which did some damage to the trenches[650] but were driven back in the end, as was inevitable. On the 16th all the new French batteries were ready, as well as that in Fort Francoli, and the bombardment began. The advanced batteries were within 120 yards of the bastions which they were attacking, and had a tremendous effect. By evening there was the commencement of a breach in the left face of the Orleans bastion, and several other parts of the enceinte were badly damaged, as was also the Prince lunette. This had not been effected without grave loss: one French battery had been silenced, a reserve magazine had been blown up, and the loss in men among the artillery had been very heavy. Nevertheless the assailants had the superiority in the cannonade, and were well satisfied. After dark the columns of stormers carried the Prince lunette by assault, one of the parties having slipped round its flank by descending on to the beach, where a few yards at the water’s edge had been left unfortified. The battalion of the regiment of Almanza which held the work was practically exterminated. Thus the Spaniards lost the last of the outer protections of Tarragona, and the captured lunette became the emplacement of one more battery destined to play upon the bastion of San Carlos (June 17).

[650] The first was a big affair, 3,000 men under Sarsfield
taking part in it, but oddly enough it is not mentioned by Suchet
or Belmas.

It was clear that the crisis was now at hand: the French were now lodged close under the walls of the city, and had already damaged its enceinte. But to storm it would be a costly business, and Contreras showed no signs of slackening in his energy, though his letters to Campoverde and his narrative of the siege both show that he thought very badly of his position. It was clear that both sides must now utilize their last resources: Suchet had already ordered up from beyond the Ebro the brigade of Abbé, which had hitherto been observing the Valencians, in order that its 3,600 men might compensate him for the heavy losses that he had suffered[651]. Contreras began to call on the Captain-General very hotly for help; at his departure Campoverde had promised that the army of succour should be pressing Suchet’s rear within seven days, and now seventeen had elapsed and no signs of its approach were to be seen. He complained bitterly that many of his officers were failing him; even colonels had gone off by sea to Villanueva de Sitjes pretending sickness, or absconding without even that excuse[652]. Naturally the spirit of the rank and file had suffered from this desertion. There is good contemporary Spanish authority for the notion that Contreras himself contributed somewhat to the discouragement, by exhibiting too openly his failing hope, and stating that Tarragona must fall in a fortnight if the field army did not save it[653]. But so far as practical precautions went he did his duty, strengthened the damaged places in the walls as best he could, and devoted much energy to seeing that the troops were properly paid and fed, and that the breaches were mined, and protected to the rear by cuttings and traverses. Very different was the conduct of Campoverde, who showed that he was absolutely unfit for command by his miserable conduct during the critical weeks. After having been joined by Miranda’s Valencian division on June 16th he had 11,000 regular troops under his hand, a force insufficient to meet Suchet in the open field, but quite large enough to give the French grave trouble--indeed to make the continuance of the siege impossible if it were properly handled. But to bring effective pressure upon the enemy it was necessary to come up close to him, and Campoverde for many days tried a policy which was bound to fail. He kept far away, cut Suchet’s communication by placing himself at Momblanch, and sent Eroles and other officers to molest the French detachments on the Lower Ebro and to cut off the convoys coming up from Tortosa and Lerida. Apparently he hoped that these distant diversions would cause Suchet to draw off great part of his army from the siege, in order to succour his outlying posts. But the French general did nothing of the kind, and took no notice of the loss of convoys, or the danger to remote dépôts; he stuck tight to the siege, and at this very time, by calling up Abbé’s brigade from the south to Tarragona, he had deliberately risked even more than before on the side of Valencia. Between the 16th and the 24th June, the critical days in the siege, Campoverde and his 11,000 men had no effect whatever on the course of operations. Yet he kept sending messages to Contreras promising him prompt assistance, and on the 20th bade him dispatch Sarsfield out of the city, to assume command of his old division in the fighting which was just about to begin. That fighting never took place--to the Captain-General’s eternal disgrace--for at the last moment he flinched from placing himself within engaging distance of Suchet. It seems clear that his true policy was to push much closer to the enemy’s lines, so as to force the French to come out against him, and then either to let them attack him in some strong position in the hills to the north-east of Tarragona, where their cavalry would have been useless, or else to avoid an engagement by a timely retreat when Suchet should have been drawn well away from the fortress. In either case he would have compelled his adversary to draw so many men away from the siege that it could not have proceeded. For it would have been useless for Suchet to march against him with less than 7,000 or 8,000 men, and the total of the besieging army had dwindled down to 16,000 by this time, while Abbé’s reserve brigade had not yet come up. Probably, as Napier suggests, a blow at the French magazines and hospitals at Reus, only ten miles from Tarragona, would have forced Suchet to draw off two divisions for a fight, and Campoverde need not have accepted it, unless he had found himself some practically impregnable position. But to skulk in the hills many miles away and send detachments against outlying French posts could have no effect.

[651] It seems to have arrived about June 20, in time for the
storm on the 21st.

[652] See his memoirs, p. 273, and Arteche, x. 274.

[653] See especially Schepeler (i. 438). Arteche (x. 317) grants
that he had not the iron resolution of Alvarez, the governor of
Gerona, but thinks that he did his honest best, and this I think
can hardly be denied. His own narrative is simple and modest, but
does not conceal the fact that he was from the first downhearted,
and feared the worst. Napier (iii. 240) calls him vacillating and
deceitful. There is some foundation for the former charge (see p.
513 below), but the latter seems unjustifiable. The only evidence
brought to justify the accusation can be explained away (see
Codrington’s letters in the appendix to Napier, vol. iii).

While Campoverde hesitated, Suchet took the lower city of Tarragona by a vigorous effort. At seven o’clock on the evening of the 21st the assault was delivered by five storming-columns, composed of the massed grenadier and _voltigeur_ companies of all his French regiments, 1,500 strong, and supported by a brigade under General Montmarie, There were now two good breaches in the bastions of San Carlos and Orleans, the curtain between them had also been much injured, and even the Fuerte Real, the inner stronghold behind the Orleans bastion, had been damaged by shot and shell which passed over the outer works. Contreras had sent down into the lower city 6,000 out of the 8,000 men who were still at his disposition, and had handed over the charge of them to Sarsfield, the officer who had the best reputation in the whole of Catalonia. But by an ill-chance there was actually no one in command when the assault was delivered. Campoverde’s dispatch recalling Sarsfield to the field army had come to hand that morning, and Contreras, thinking himself bound to obey it, sent Sarsfield a passport to leave the city, and designated General Velasco to take his place. Sarsfield, whose courage cannot be impeached, but whose judgement was evidently at fault on this afternoon, left at once, embarking on board a boat in the harbour at three o’clock, without going to see Contreras or waiting for the arrival of the officer who was to supersede him. He merely sent for the senior colonel in the lower city, handed over the command to him, and put out to sea at a moment’s notice. Four hours later, when the storm took place, Contreras was not aware that Sarsfield had yet departed, and Velasco, coming down to take charge of the troops, found himself in the middle of the fighting before he had reached the walls, or discovered the manner in which their garrison was distributed. There was clearly something wrong here--apparently Sarsfield and Contreras were not on good terms, and the former acted with small regard for the welfare of the service[654].

[654] Toreno (ii. p. 544) makes out that the fault was with
Contreras, and Napier, who used far more violent language, says
that he ‘acted a shameful part.... The assault was momentarily
expected, yet he ordered Sarsfield to embark immediately,
averring that he had Campoverde’s peremptory commands. Sarsfield
remonstrated vainly, saying that the troops would be left to
an inefficient subordinate, but was compelled to embark, and
Velasco, coming a few hours later, found only the dead bodies of
his garrison. Contreras then assured Codrington and the Junta
that Sarsfield had gone without orders, and betrayed his post!’
There is much misrepresentation in this: (1) Contreras _had_
Campoverde’s command to send him Sarsfield without delay. (2)
He passed the order on to Sarsfield, but did not know that the
latter had gone off, without waiting a moment, after sending him
a note to say that Don José Carlos, the senior colonel in the
lower city, was incompetent to take command. Naturally Contreras
thought that Sarsfield would have waited to see Velasco and hand
over the troops to him. (3) Therefore Contreras was correct in
saying to the Junta that Sarsfield departed ‘sin conocimiento
mío,’ though he had sent him a passport to leave the town. See
Arteche, ix. 288-9.

At seven o’clock Suchet let loose the stormers, who were led by the Italian General Palombini, while, to distract the attention of the garrison to other points, he ordered a general bombardment of the northern front, and showed a column on the side of Fort Olivo. The assault was immediately successful at both the critical points: the forlorn hope on the side of the Orleans bastion, starting from its ditch, went up the breach like a whirlwind, losing somewhat from the musketry fire of the defenders but not from their cannon, for all had been silenced. The Spaniards were cleared out so quickly from the bastion that they had not time to fire two mines, which would have blown up the breach and the storming-column if they had worked. They rallied for a few minutes at the gorge, but were driven from it by the French reserve, who poured into the town. At the San Carlos breach matters went almost as rapidly: the first attacking column was checked, but when the second supported it, the united mass carried the breach and burst into the town; a retrenchment and a row of palisades erected behind the breach were crossed in face of a half-hearted defence. All the columns having penetrated within the walls, those who turned to the left attacked the Fuerte Real, the weak and somewhat damaged work which served as citadel to the lower city: it was carried by assault without any great difficulty, partly because its earthen ramp had been somewhat damaged by the bombardment, and could be climbed at some points, but more because the garrison defended themselves very badly, and gave way when they saw that the streets on their flanks and behind them were inundated by the enemy. The other section of the stormers, inclining to their right, moved towards the mole and the large magazines at its base, where they met General Velasco, who had only arrived at the moment that the assault began, with the Spanish reserve. There was fierce fighting here for a moment, but, turned by a column which had passed around their flank by the quay at the water’s edge, Velasco’s men broke like the rest. The whole of the garrison rushed up the slope, towards the one gate which leads into the upper city, and finally entered it under cover of a heavy fire kept up from the neighbouring ramparts. Contreras reports that some of the pursuers came on so fiercely that they were shot down while actually battering at the closed gate.

The losses of the two parties were about equal in numbers--Contreras reports that he found no more than 500 and odd men missing when the battalions from the lower city were reassembled; Suchet gives 120 killed and 362 wounded as his total loss. The casualties on both sides would have been heavier if the garrison had fought better--but it is clear that, when the breaches were once gained, no serious attempt was made to defend the Fuerte Real, the retrenchments, or the barricaded houses of the lower city. Only Velasco’s reserve battalions made any fight in the streets; the rest fled early. The French might perhaps have made more captures--they only took 200 wounded prisoners--if they had not turned at once to plundering the houses and magazines. But they fell into great disorder; many of the unfortunate inhabitants of the quarter about the port were not only stripped of their goods but murdered, and a great number of dwellings were wantonly set on fire. Eighty guns were captured on the walls of the lower city, and a great quantity of food and stores in the dépôts along the quay. But the soldiers destroyed more than was saved--especially in the wine stores.

Not the least disastrous result of this unhappy affair was that the harbour was now closed to the Spaniards. The English men-of-war and the native merchant vessels which had hitherto sheltered under the west end of the mole had to put out to sea. The traders went off to Villanueva, Minorca, or Valencia, but Codrington’s squadron sought the bare roadstead off Milagro Point, under the precipitous southern face of the upper city. Here there were no quays, and when the sea was rough it was impossible to land. But in ordinary weather boats could communicate freely with the shore, and Tarragona was not yet deprived of its access to the water, though that access had become difficult and dangerous. Suchet proceeded to make it more so on the 23rd, when he erected a battery, near the base of the mole, to play on the roadstead. Landing, however, was made rather exciting than dangerous by these guns, which never did much harm.

On the morning after the storm of the lower city the French engineers began to make surveys for the attack on the inner line of defence of Tarragona. Its strength lay in its commanding position, and in the fact that along many parts of its short front the ground just below it was too steep and too rocky to allow of approaches being constructed on it. Its weakness was that the wall was weak and old--a seventeenth-century work built only to resist the cannon of that day. There was no ditch or other outer defence, unless a hedge of prickly aloes counted as such. The front was composed of four bastions; counting from north to south they were named San Pablo, San Juan (at whose left side lay the only gate), Jesus, and Cervantes. The last named overhung the precipitous cliff looking down to the sea above Milagro Point. The French engineers reported in favour of making the attack on the curtain wall between San Juan and San Pablo, the ground here being less steep than elsewhere, and showing soil which could be dug into; there was also some cover to be found, in half-ruined houses along the road up to the gate. Moreover the other, or southern section of the wall, was not only on a steeper ascent, but might be exposed to high-trajectory fire from the ships in the roadstead below. The first parallel, therefore, was thrown up opposite San Juan and San Pablo, with a communication to the rear covered by the buildings and gardens along the road, and three batteries were planned in it, and commenced on June 24th. A fourth battery, down in the plain outside the city, was to co-operate by a flanking fire uphill.

This day saw Campoverde’s first and last demonstration in favour of the garrison. It was a miserable affair. Driven to do something by Contreras’s appeals, and by the openly displayed discontent of his own army, he at last drew in close to the French lines. On the 23rd his army marched from Momblanch to Villarodona, fifteen miles north-east of Tarragona. On the next day it was divided into two columns; the first (composed of Miranda’s Valencian division) marched over the hills, with orders to fall upon the encampments of Harispe’s Italian brigades on the north-east side of the French lines. The second, or Catalan division, under Sarsfield, with which went the Captain-General himself, marched by another road more to the east, and was to come into line on Miranda’s left. Meanwhile Contreras was to make a sally out of the eastern side of Tarragona with 4,000 men of the garrison, and to attack Harispe’s rear when he saw his front engaged with the ‘army of succour.’ Both Campoverde’s columns reached the points designated for them, Miranda getting unopposed to Pallaresos, and the second division to Cattlar, three miles further east. Suchet, warned by his outlying cavalry, concentrated Harispe’s and part of Frère’s divisions in the rear of his lines to the French left of Fort Olivo, leaving Habert, Abbé, and the rest of Frère’s troops to hold the lower city and the trenches. His line, composed of some 8,000 men including all his cavalry, was plainly visible both to the Spaniards outside and those within the city, and Contreras formed his sallying column ready to rush down when the first cannon-shot should be heard. But Miranda, on finding himself in touch with the enemy, sent back messages to the effect that he was not sure of his route, that the French seemed very strong, and that he dared not advance. Instead of depriving him of his command, and then bringing up the second column to the help of the first, Campoverde, after some hesitation, gave him leave to draw back, and both divisions retired that night to Vendrils, ten miles to the rear, in the eastern hills. Not a shot had been fired, and Contreras, whose men had been waiting under arms the whole afternoon, had to draw them back into the city without having seen a single man of the relieving army, which, though only four miles away, was hidden from him by the intervening hills. So ended a day of great peril for Suchet, who with 11,000 men in front of him, and 4,000 more ready to attack his rear, might well have suffered a disaster, or at least have proved unable to prevent the junction of the two hostile forces. For the Spaniards were not bound to descend and attack him in the plain, but might have manœuvred along the hills and forced him to take the offensive in unfavourable ground, under pain of seeing them break his blockading line. Codrington summed up the situation by writing to his chief, Pellew, that ‘the Marquis [Campoverde] blamed Generals Miranda and Caro, while the latter retorted the accusation, and I am inclined to think by giving full credit to what each says of the other, neither will suffer ignominy beyond that to which his conduct has entitled him.’ For Tarragona, as Contreras was truly repeating in every dispatch that he sent out, was in imminent danger, and if the army of succour did not give it _immediate_ help might fall at any moment. The city, as a matter of fact, was taken only four days later.

Campoverde, however, had now formed the conclusion that he was still too weak to attack Suchet. He wrote orders to Contreras to send him out of the city his two best regiments, Iliberia and Almeria, and General Velasco to command them. He made a desperate appeal to the _somatenes_ to rally to his colours, which had little effect, for his reputation was now gone, and he was suspected of timidity or even of treason. Finally he got news that there was a small British expeditionary force in Catalan waters, and sent his lieutenant Eroles to sea, to look for it, and to invite it to land at Villanueva and join him. In a week or so he would have 20,000 men, as he supposed, and would then try something desperate. Meanwhile, unjustly suspecting Contreras of cowardice, he sent secret letters into Tarragona to the brigadier-generals of the garrison, bidding them depose and confine the governor if he showed signs of capitulating. Disgusted at this move, the generals showed the epistles to Contreras, who was driven still deeper into despair by seeing that the Captain-General distrusted him, regarded his views as to the danger of the city as exaggerated, and was evidently deferring succour for an indefinite period. Nevertheless he concealed his knowledge of the plan for his deposition, and prepared, under protest, to send the regiment of Almeria off by sea; to dispatch Iliberia also he refused, saying that his garrison was already insufficient. But rough weather on the 27th prevented the regiment from embarking from the dangerous Milagro roadstead, which was unapproachable by boats during an east wind.