SECTION XXXVIII
THE BATTLES OF THE PYRENEES
CHAPTER I
THE SIEGE OF ST. SEBASTIAN. FIRST PERIOD
Having pushed his whole front line up to the French border, from Giron at Irun, hard by the ocean, to Byng, far inland on the watershed of the historic Pass of Roncesvalles, Wellington turned his attention to the siege of St. Sebastian, on which he intended to press hard and rapidly, while Pampeluna was to be left for the present to the slower process of blockade and starvation. He had no intention of pushing forward into France till the result of the negotiations in Germany should be known. He was kept informed of the doings at Dresden from time to time, but occasionally bad weather in the Bay of Biscay delayed the courier from Downing Street, and there was an interval of many days, during which the absence of political information began to cause him anxiety. On the whole, as he was informed, it was probable that the war would be renewed; but the intentions of Austria were still suspect, and the danger of a separate peace in the North never ceased to disturb him until hostilities actually recommenced in the middle of August. On July 10th he had still many weeks to wait, before the news arrived that Metternich had broken with Napoleon—after a certain famous and stormy interview—and that Austria had come into the Grand Alliance. Meanwhile he intended to make his defensive position on the Pyrenean frontier sure, by the capture of the fortresses. It is curious to note that Napoleon understood and approved from the military point of view of this policy. On Aug. 5 he wrote to his minister Maret from Dresden, ‘it is certain that Lord Wellington had a very sensible scheme: he wished to take St. Sebastian and Pampeluna before the French Army could be reorganized. He supposed that it had suffered more heavily than was actually the case, and thought that more than a month would be required to refurnish it with artillery[772].’ This was a guess that came very near the truth, though the Emperor could not know that the cause of the British general’s delay was mainly political and not military.
It was fortunate for Wellington that the astounding triumph of Vittoria had put him in a position to dictate his own policy to the Ministry at home. For Lord Bathurst and Lord Liverpool had been sending him suggestions which filled him with grave apprehensions. Before the news of Vittoria reached London, the War Minister had ‘thrown out for consideration’ two most inept projects. If the French in Spain had been driven beyond the Ebro, would it not be possible to ‘contain’ them for the future by Spanish and Portuguese armies, and to bring round Wellington himself and the bulk of his British troops to Germany, leaving only a small nucleus to serve as a backbone for the native levies of the Peninsula? The appearance of Wellington with 40,000 of his veterans on the Elbe would probably induce the Allies to appoint him Generalissimo against Napoleon, and if Great Britain had the chief command in Germany she could dictate the policy of the Coalition. But supposing that the war was not resumed in Central Europe, and that a general peace was negotiated, would it be possible or expedient to allow the Emperor to retain the boundary of the Ebro for the French Empire, on condition that he yielded to the demands of the Allies in other parts of the Continent? Or would Wellington guarantee that in the event of a Peace from which Great Britain stood out, he would be able to maintain himself in Spain and Portugal, as in 1810 and 1811, even though Napoleon had no other enemies left, and could send what reinforcements he pleased across the Pyrenees?[773]
Even after the great event of June 21st had become known in London, we find the Prime Minister repeating some of Lord Bathurst’s most absurd suggestions. He wrote in a tentative and deferential style, to set forth the hypothesis that it might be possible to so fortify the Spanish frontier that it might be held by a small British force aided by Spanish levies, ‘applying the principle upon which the Lines before Lisbon were formed to the passes of the Pyrenees’; and he inquired whether the warlike Navarrese and Biscayans could not defend this new Torres Vedras, if backed by a small force—20,000 or at most 30,000—veteran Anglo-Portuguese troops[774]. If this could be done Lord Liverpool obviously hoped great things from the effect of the appearance of Wellington, with his new prestige and the bulk of his victorious battalions, in Central Europe.
The recipient of these ill-advised suggestions would have nothing to do with them. When on July 12th he got Bathurst’s dispatch of June 23rd, he replied that, being the servant of the State, he must obey any orders given him by the Prince Regent and the Ministry, but that he saw no profit in going to Germany, which he did not know, and where he was not known, whereas in Spain he had a unique advantage in ‘the confidence that everybody feels that what I do is right’. ‘Nobody could enjoy that same advantage here—while I should be no better than another in Germany. If any British army should be left in the Peninsula, therefore, it is best that I should remain with it.’ As to making any general peace which left the French the line of the Ebro as a frontier, that problem had been settled by the battle of Vittoria. ‘I recommend you not to give up an inch of Spanish territory; I think I can hold the Pyrenees as easily as I can Portugal.’ It would really be better to have Joseph Bonaparte as King of Spain (considering how ready all Bonapartes are to break away from their brother), than to have Ferdinand back with the Ebro as his frontier. ‘In the latter case Spain would inevitably belong to the French[775].’
As to the ridiculous idea of fortifying the Pyrenean passes in the style of the Lines of Torres Vedras, Wellington speaks with no uncertain sound. ‘I do not think we could successfully apply to the frontier of Spain the system on which we fortified the country between Lisbon and the sea. That line was a short one, and the communications easiest and shortest on our side. The Pyrenees are a very long line; there are not fewer than seventy passes through these mountains, and the better communications, so far as I have been able to learn, are on the side of the enemy. We may facilitate defence by fortifying some of the passes: but we can never make in the Pyrenees what we made between the Tagus and the sea[776].’
It looks as if Lord Liverpool, in serene ignorance of local geography, imagined the Pyrenees to be a simple line of precipices, pierced by a limited number of passes, which could be sealed up with walls, in the style in which Alexander the Great in the mediaeval romance dealt with the ‘Caucasian Gates’. For we cannot suppose that even the most unmilitary of politicians could have conceived it possible to build something like the Great Wall of China for several hundred miles, along the whole frontier from Fuenterrabia to Figueras. Yet this is what his proposal, if taken literally, would have meant; and Wellington, to show its absurdity, gave the total number of passes available for mules or pedestrians between the Bay of Biscay and the Mediterranean. Those by which guns or wheeled transport could pass were (of course) no more than five or six—but there would be no finality in sealing up these few: for the enemy could turn the flanks of any of them with infantry scrambling up accessible slopes—as Soult indeed did during the battles of the Pyrenees, only a month after Lord Liverpool had made his egregious suggestion—or as Napoleon had done at the Somosierra in the winter of 1808.
Lastly, as to the Prime Minister’s proposal to hand over the defence of the Pyrenees to Spanish armies backed by a mere 20,000 Anglo-Portuguese, Wellington replied that the Spanish Government had shown itself most consistently unable to feed, pay, and clothe its armies, or to provide the transport which would make them mobile. In June 1813 there were 160,000 men on the Spanish muster rolls, but only a third of that number at the front and actively engaged with the French, even when all troops in the Pyrenees, Aragon, and Catalonia were reckoned up. To trust the defence of the frontier to a government which could not move or feed more than a third of its own troops, would be to invite disaster. The best way to help the Spaniards was not to ask for more men, but to improve their finances, so that they should be able to employ a greater proportion of their already existing troops. As things stood at present, it would be insane to request them to take over the main burden of the war. ‘It is my opinion that you ought not to have less than 60,000 British troops in the field, let the Spaniards have what numbers they may[777].’
Wellington had by this time established himself in such a commanding position that the ministers had to accept his decision—however much they would have liked to move him round to Germany, and to press for his appointment to the unenviable position which Prince Schwarzenberg occupied during the second campaign of 1813. The best proof of the wisdom of his determination to remain in the Peninsula was that the French, far from adopting a defensive policy, resumed the offensive under Soult within a few weeks. It would be an unprofitable, as also a dismal, task to consider what would have happened during the battles of the Pyrenees had Soult been opposed, not by Wellington and his old army, but by a Spanish host under O’Donnell or Freire, backed by two or three Anglo-Portuguese divisions.
Controversy with Lord Bathurst on general principles was only one of Wellington’s distractions during the middle weeks of July: he was at the same time carrying on an acrimonious correspondence with the Spanish Minister of War, concerning the removal of Castaños and Giron from their posts, and enduring many controversial letters from Sir John Murray regarding the Tarragona fiasco. Murray had realized the disgust which his wretched policy had roused, alike in the army and in the British public, and was endeavouring to justify himself to his commander-in-chief by long argumentative epistles. Wellington refused to commit himself to any judgement, and agreed with the ministers at home that a court martial would be required[778]. He suggested to the authorities at the Horse Guards that Murray should be charged firstly with making no proper arrangements to raise the siege of Tarragona or to bring off his guns, secondly with having disobeyed the clause of his instructions which bade him return at once to Alicante if he failed in Catalonia, and thirdly with having betrayed General Copons, by bringing him down to the coast and then absconding without giving any proper warning of his departure[779]. After mature reflection, however, he advised that the third charge should be dropped, not because it was unjustified, but because there would be great inconvenience in bringing Copons or his representatives to bear witness at a British court martial, when they were wanted in Catalonia.
But all this correspondence, and much more on less important subjects, did not prevent Wellington from paying his promised visit to General Graham, to inspect in person the allied positions on the lower Bidassoa, and to supervise the arrangements for the siege of St. Sebastian, which indeed required supervision, for already it was evident that things were not going so well as might have been hoped in this direction (July 12).
The first impression made on the observer by the fortress of St. Sebastian is that it is an extremely small place. This is true—though it is also true that its effect is somewhat dwarfed by its immense surroundings—the limitless expanse of the Ocean on one side, and the high and fantastic peaks of the Jaizquibel, the Peña de la Haya, and the Four Crowns, which have just been passed by the traveller coming from the side of Bayonne. The fortress consists of a lofty sandstone rock, as steep as Edinburgh Castle and 400 ft. high, beaten on three sides by the sea, and united to the land by a low sandy isthmus about 900 yards long. There was an old castle called La Mota, on the summit of the rock, which is named Monte Urgull, with three modern batteries on its southern or landward face[780], which is very steep and only accessible at two points by winding roads. But these were not the main defences of the place, which lay lower down. The town of St. Sebastian was built on the lower ground below the rock, extending from sea to sea and occupying about a third of the isthmus, the rest of which was broken sandy ground about 400 yards broad. At the southern neck of the isthmus was a hill crowned by the large monastery of San Bartolomé, which the French held as an outwork at the beginning of the siege. The space between this hill and the wall of the town, now thickly packed with the hotels and avenues of a fashionable seaside resort, was in 1813 waste ground dotted with a few isolated houses and gardens, which the governor was busily engaged in levelling when the blockading force arrived. At two points there was a sufficient accumulation of buildings to form a small suburb. One group was at the head of the bridge across the river Urumea, which joined the peninsula to the eastern mainland; this was named Santa Catalina, from a chapel of that saint which lay in it. The other, called (for a similar reason) San Martin, was immediately at the foot of the hill of San Bartolomé. The bridge had been blown up by the French when the Spaniards first appeared, and the suburbs set on fire, but their roofless and partly fallen houses still gave a good deal of cover.
The land-front of St. Sebastian was very formidable; it was only 400 yards broad and was composed of a very high curtain from which projected one large bastion in the centre and two demi-bastions overhanging the water on each flank, that of San Juan, above the estuary of the Urumea, that of Santiago on the western side nearest the harbour. In front of the bastions was a very broad hornwork, having outside it the usual counterscarp, covered-way, and glacis. Specialists, wise after the event, criticized the whole land-front: they found its trace defective, considered that the scarp of the hornwork ought to have been more than 23 feet high, and pointed out that in the rear line the two demi-bastions did not well protect the flanks, since they were not themselves covered by the hornwork, and had no ditch or glacis, because the salt water came up to their foot at high tide[781].
But as the land-front was never attacked, these objections were of comparatively little importance. The weak side of St. Sebastian was really its eastern water-front. The western water-front on the open bay was inaccessible; but on the opposite side, where the town wall ran along the estuary of the river Urumea, the conditions were peculiar. The Urumea is a tidal river—at high water it washes right up to the foot of the walls, but at low tide it recedes into a narrow channel and leaves exposed an expanse of rock and sand, ranging in breadth from 50 to 150 yards, all along the water-wall for a length of some 400 yards—half the east side of the town. The rampart here was a plain curtain, not very high and only 8 feet thick: of course it could have no ditch or other outer fittings, since the salt water reached right up to it for half the day. Its only salients, from which flank fire could be used, were one small bastion (St. Elmo) and two ancient round towers called Los Hornos and Amezqueta, which projected slightly from the straight curtain.
At high tide this water-front was unapproachable—at low tide it was weak and accessible, for the Urumea was fordable in many places, and shrank into a channel only 50 yards broad, meandering through a waste of shingle, rocks, and mud. On its eastern bank were rolling sandhills of some height, which commanded the isthmus and the southern half of the town, from a distance of no more than 700 or 800 yards. To the right of the sandhills, called Los Chofres, was a steep hill, the Monte Olia, facing the castle of St. Sebastian across the broader part of the estuary, and by no means out of gunshot of it, since not more than 1,000 or 1,100 yards of water lies between them.
Since it had been rebuilt as a fortress of the Vauban-Cohorn style in the last years of the seventeenth century, St. Sebastian had only once been besieged: for during the Revolutionary War of 1792-5 it surrendered without any resistance to the Jacobin armies, while in 1808 Napoleon had seized it by treachery and not by force. The one formal attack made on it was by the Duke of Berwick in 1719, when the armies of the Regent Orleans crossed the Bidassoa, in the short war provoked by the ambitions of Elizabeth Farnese and Cardinal Alberoni. Berwick had established his siege-batteries not on the isthmus, to batter the land-front, but on the Chofres sandhills, from whence he pounded the eastern sea-front to pieces. Several breaches having been made in it, and trenches on the isthmus having been pushed as far as the glacis of the hornwork, the governor very tamely surrendered, without waiting for an assault[782]. Unwarned by this revelation of the weakness of the defences along the Urumea, the engineers of Philip V made no improvements in them when they were reconstructed after the war, and the fortress was in 1813 little different from what it had been in 1719.
The easy success of Marshal Berwick turned out a very unhappy thing for the British army; for the history of the last siege being well remembered, Major Charles Smith, the senior engineer with Graham’s column, reported that Berwick’s plan was the right one[783], when the first survey of the place was made. And when the head-quarters staff came up, his superior officer, Sir Richard Fletcher, agreed with him, as did Colonel Dickson commanding the artillery. They all forgot that to make a breach with ease is not necessarily the same thing as to capture a fortress, and that the eighteenth-century slackness of the governor, who capitulated when his walls were once breached, gave no indication of what might be done by a very resourceful and resolute French officer, who had in his mind Napoleon’s edict of 1811 that every commandant who hauled down his flag before standing at least one assault should be sent before a court martial. Nor is there any doubt that Wellington himself must take his share of the responsibility, as he went round the place on July 12th and had a good look at it, in company with Charles Smith and Dickson, from Monte Olia and the Chofres sandhills. Dickson summed up the results of their discussion as follows:—‘The project is to effect a breach in the uncovered sea-wall forming an angle on the left of the land-front, between two towers [Los Hornos and Amezqueta], being the same spot the Duke of Berwick breached in 1719, when he took St. Sebastian. The convent of San Bartolomé, which the French have fortified and occupied in force, is absolutely necessary to take first, in order to be able to advance on the isthmus in support of our breaching operations from the Chofres. The plan of attack was the proposition of Major Smith[784].’
This, of course, was not the regular and orthodox way to attack such a fortress—as it involved the making of breaches only accessible at low tide, in a part of the wall to which no trenches could draw near, since its lower courses were six feet under water for half the hours of the day. There would have to be a long advance from the nearest dry ground on which approaches could be dug, across the tidal waste, in order to reach the spot selected for breaching. And this would undoubtedly be a very bloody business for the exposed storming troops. The advantages to be gained were rapid action and a quick end, without the necessity which an orthodox scheme would have involved of sapping up to the hornwork, storming it, and then tackling the lofty bastion behind it. The British engineer who wrote the history of the siege considered that ‘the operations against San Sebastian afford a most impressive lesson on the advantage of due attention to science and rule in the attack of fortified places: the effort then made to overcome and trample on such restrictions caused an easy and certain operation of eighteen or twenty days to extend over sixty days, and to cost the besiegers 3,500 men killed, wounded, or taken, bearing strong testimony to the truth of the maxim laid down by Marshal Vauban, that hurry in sieges does not lead to an early success, often delays it, and always makes it very bloody[785].’
St. Sebastian had been abandoned to its own resources on June 28, when, after the departure of Foy’s troops from Oyarzun, the Spaniards had closed in on the place, and established a blockade. The force employed was four battalions of Biscayan volunteers under Mendizabal, the remains of the bands which Foy had beaten in May, and which had taken part on June 26 along with the British 1st Division in the combat of Tolosa[786]. They were very irregular troops, and their indiscipline and incompetence shocked General Graham when he came to visit their lines on July 6th[787]. They had no notion of guarding themselves, and had suffered severely, first from an attempt to storm San Bartolomé on June 28th, and later from a French sortie on July 3, when Rey sent out two columns of 700 men from the Convent, which surprised their camp, took many prisoners, and drove them back for nearly two miles. Graham recommended that they should be replaced at once by solid troops, and at Wellington’s suggestion told off the British 5th Division and the unattached Portuguese brigade of Bradford for the siege[788]. Mendizabal’s men were sent off to join in the blockade of Santoña[789]. On July 7th the siege-troops arrived; the independent brigade established itself on the Chofres, the 5th Division—still under Oswald as at Vittoria—took post on the heights of Ayete, facing the hill and convent of San Bartolomé. From this moment the serious operations may be considered to have begun. A blockade of the seaside had been established, in a rather intermittent fashion, four days earlier, when Sir George Collier appeared in the bay with a frigate, the Surveillante, a corvette, and two brigs, all the force that he could collect for the moment in response to Wellington’s appeals to the Admiralty. The amount was wholly inadequate, even when supplemented by some local fishing craft and pinnaces, which were manned by the battalion of marines from Giron’s Galician army. All through the siege French trincadores and luggers from Bayonne and St. Jean de Luz ran the blockade at night, bringing in food, munitions, and reinforcements. For Rey having asked for more gunners, he was sent several detachments of them, who all arrived in safety, while he got rid of many of his wounded by the returning vessels. Wellington was full of justifiable wrath at the miserable help given him—his objurgations brought from Lord Melville an unsatisfactory reply, to the effect that the operations in the Baltic and the American War absorbed many of the British light craft which would have been available in earlier years, and that the Admiralty had not been warned in the spring that any large fighting force would be wanted on the Biscay coast in July. ‘Neither from you, nor from any other person at your suggestion, did we ever receive the slightest intimation that more was expected than the protection of your convoys, till the actual arrival of Sir Thomas Graham on the coast after the battle of Vittoria.’ Melville then proceeded to write lengthy observations on the dangerous nature of the coast of Northern Spain, ‘where ships cannot anchor without extreme risk, and are exposed to almost certain destruction in a gale, when from its direction they can neither haul off shore nor run for shelter into a port. If you will ensure the ships a continuance of east winds, they could remain with you, but not otherwise. All the small craft in the British navy could not prevent the occasional entry of small boats by night into San Sebastian’s, though it may be rendered more difficult and uncertain.’ Melville then proceeded to tax Wellington with grave professional irregularity, for having written to Sir George Collier, and other naval officers, letters taxing the Admiralty with neglect and incompetence. ‘Appeals to subordinate officers against their superiors are not customary in any branch of the service, and must be injurious to the public interest[790].’
It was not so much the arguments used by Melville as his offensive tone which Wellington resented. The victor of Salamanca was naturally irritated at finding himself treated de haut en bas by a personage of such complete insignificance, who had not even the equivocal reputation of his father (the devisor of so many unlucky expeditions, the impeached minister of 1806) to lend weight to his lectures. He replied in a very short letter that he was not desirous of getting into discussions, that his demands for more naval assistance had begun in the winter of 1812-13, and that after reading over again his letters to naval officers he could see nothing to find fault with—save that he had once sent to Sir George Collier an extract of a dispatch to Lord Bathurst, instead of a separate communication on the same topic addressed to Collier personally—a trifling subject on which to start a controversy[791]. Meanwhile he considered, as a dozen letters show, that he was being poorly served by the Admiralty, and ‘formed private opinions on the subject—which private opinions may not perhaps deserve much attention,’ as he sardonically observed.
One thing, however, the Admiralty had contrived to carry out according to Wellington’s request—the stores and battering train, which he had accumulated at Corunna in the spring, had come round to the Biscay coast as soon as could have been expected; they had been sent to Santander, as he directed, and arrived in that harbour on the very day (June 29) on which Major Frazer, sent off four days after Vittoria, came from Head-Quarters to inquire for them[792]. The orders were to land them at Deba in Guipuzcoa, fifteen miles west of St. Sebastian; but just as the disembarkation was commencing (July 3), a change of destination was made—everything was to be put ashore at Passages, a much more convenient place, only a couple of miles from the blockaded fortress. This is a most astonishing little harbour, the safest port on the coast, for it is absolutely land-locked; its entrance is so narrow that the circular basin looks like a lake without an issue, till the eye has with some difficulty discovered its exit. It pays for its security, however, by being hard to enter in time of storm, when very careful navigation is required. But a besieging army has seldom been granted a base so near to its field of operations—it was only two and a half miles from the water’s edge at Passages to the Chofres sandhills, where the chief batteries were to be constructed, though four miles to the head of the isthmus where the left attack was to be made. And the fording of the Urumea, whose bridge the French had burned, was sometimes dangerous—more than one gun was lost in the quicksands[793].
By the time that Wellington had worked out the scheme of assault with Charles Smith, the guns were already arriving, not only the battering train from Corunna, which had begun to come ashore on July 7, but the heavy artillery reserve of the army, which had traversed all the weary miles of road from Ciudad Rodrigo. The total of pieces available was twenty-eight from the convoy, six from the heavy battery of the army reserve, and six lent by Sir George Collier from the main deck of his ship, the Surveillante—or forty in all[794]—a very different train from the miserable four guns which had battered Burgos in the preceding autumn.
Before any regular siege work could begin, it was obviously necessary to clear the French out of the monastery of San Bartolomé, on its hill at the neck of the isthmus: till this should be taken, it was impossible to get forward on the shore from which the assault was to be delivered. Under-estimating the strength of the post Graham and the engineers, contrary to the advice of Major Hartmann of the K.G.L., the senior artillery officer present before Dickson’s arrival[795], had tried to reduce it on July 7th by the fire of a Portuguese field battery belonging to the artillery reserve; but the walls proved too solid for 8-pounders to damage, and some firing with red hot shot failed to set fire to the roof. It was clear that it would be necessary to wait for the heavy guns to come up. Meanwhile the French continued to strengthen San Bartolomé, throwing up an earthwork in its cemetery and loopholing and barricading the ruined houses near it, as well as those in the burnt-out suburb of San Martin, which lay at its foot. They also commenced a redoubt on the isthmus, half-way between the outlying convent and the hornwork in front of the city wall.
While the siege train was being got ashore, the engineers made ready the emplacements for the guns—they started with throwing up two batteries on the Ayete heights to play on the convent of San Bartolomé[796], three on the Chofres sandhills[797] to act against the sea-wall where the main breach was to be formed, and one on the lofty Monte Olia, for long-distance fire against all the fortifications, including those of the rock of Monte Urgull[798]. But on the 14th, when the battering began, the only guns used were those on the left attack opposite the convent, as it was considered unwise to molest the city fortifications before there was any way of getting near them: approach was impossible as long as the isthmus was all in the enemy’s hands.
Two days’ battering (July 14-15) brought down the roof and part of the walls of San Bartolomé, and damaged the subsidiary works around it severely. But General Rey was determined to hold on to his outer defences as long as possible: the best part of two battalions was told off to hold the ruins and the suburb of San Martin behind them. On the afternoon of the 15th General Oswald, judging from the dilapidation of the buildings that they were untenable, sent in the 8th Caçadores from his Portuguese brigade to storm them; but the French had dug themselves well in, a furious fire broke out against the assailants, and the attack was turned back with the loss of 65 men[799]. It was obvious that more battering was required to evict the gallant garrison, so the fire was resumed on the 16th, with the aid of a field battery playing across the Urumea to enfilade the defences, and to pound the ruined houses of San Martin, where the French reserves were sheltering.
The fire was very effective: the inner woodwork of the monastery blazed up, its porch was levelled to the ground, great gaps appeared in the walls, and the earthwork in the cemetery was much damaged. At 10 a.m. on the 17th a second storm was tried, with forces much larger than those used at the first—not one battalion, but the equivalent of three: Oswald drew upon Bradford’s independent brigade from across the Urumea for 700 Portuguese volunteers. Two columns were launched against the French position, each consisting of a screen of Caçadores, a support of Portuguese Line troops, and a reserve of British companies from the first brigade of the 5th Division[800]. The right column aimed at the cemetery earthwork and the fortified houses—the left at the main buildings of the convent, which were still smouldering from the fire. Both achieved their purpose without any very serious loss, and the garrison was driven down hill into the suburb of San Martin, pursued by the stormers. Here the French reserves intervened, and for a moment the attack was turned back; but the companies of the 1/9th which formed the British supports re-established the fight, and the enemy retired in disorder towards the town. Unfortunately the victors pursued recklessly, and came under the fire of the French guns in the hornwork and the bastion behind it, whence many unnecessary casualties. For it had never been intended that the troops should advance on to the isthmus, where of course they could do nothing against an intact front of fortifications. The total losses of the assailants were 207 killed and wounded—the 1/9th which had headed the rush on to the bare ground suffering most, with 70 casualties. The French had 40 killed and 200 wounded.
It was now possible to commence the subsidiary attack on the side of the isthmus, as the French held nothing outside the walls save the recently constructed redoubt in front of the hornwork[801]. Two batteries were thrown up on the San Bartolomé hill, and on the 18th the guns with which the convent had been breached were moved down into them[802]. The main object of these batteries was to enfilade the stretch of sea-wall, on which the greater weight of metal on the Chofres downs was to play from the front. Two additional batteries were thrown up, more to the right, which, while aiding the enfilading fire, were also intended to shell the land-front, and keep down the fire from the bastion and hornwork[803]. Finally, the Chofres attack was strengthened with two more batteries[804]. On the night of the 19th sixteen guns in all were in place on the isthmus front, or left attack, twenty-three on the right attack to the east of the Urumea, including six on the summit of the lofty Monte Olia, on the extreme flank[805]. Meanwhile the approaches, by which the troops would have to move out to assault the breaches when made, were not nearly so far forward as could have been desired: little more had been done than to push a zig-zag down the slope of San Bartolomé into the ruins of San Martin—further advance was made difficult by the fact that the French were still holding the forward position of the redoubt in the centre of the isthmus. It was only on the night between the 19th and 20th that they evacuated this post[806]. Moreover the fire of the defence was as yet intact, since the British heavy batteries had not begun their work: hence the sap-head was a decidedly unhealthy place.
At 8 a.m. on the morning of July 20 nine of the eleven siege batteries opened—all but the two new ones on San Bartolomé, where the work was a little belated. The effect of the first day’s work was fairly satisfactory—the parapet of the sea-wall began to crumble, the enemy’s guns en barbette on the land-front were partly silenced, and the high trajectory fire from the lofty Monte Olia battery searched the streets, the back of the bastion, and the interior of the hornwork. The enemy concentrated all his attention on the largest British battery upon the Chofres dunes, the eleven guns which were farthest forward (No. 3) and were doing much damage to the sea-wall. The counter-fire here was severe—one gun was split by a ball striking its muzzle, others had wheels broken, three had to stop firing because their embrasures, badly built with sand, fell in and could not be kept clear[807]. By evening only six of the eleven guns were still at work.
An afternoon of wind and showers, which had made accurate aim difficult, was followed on the night of the 20th-21st by a torrential rain, which much impeded the special work which had been set aside for the dark hours. Seven hundred men from Spry’s Portuguese brigades had been told off to open a parallel right across the isthmus, starting from the lodgment already established in the ruined houses of San Martin. ‘In the perfect deluge the working parties sneaked away by degrees into houses and holes and corners. After numerous difficulties and exertions only about 150 could be collected and set to work at 11 p.m. The party was not discovered, and the soil being light soon got under cover.... But the parallel finished was for only about one-third of the extent required[808].’
On the next morning Sir Thomas Graham sent out a parlementaire to summon the governor to surrender—rather prematurely as it would seem, for no real breach had yet been formed, nor had the fire of the defence been subdued. General Rey having made the answer that might be expected from a resolute officer[809], the battering recommenced, and was continued on the 21st-22nd-23rd-24th July. It was quite effective: on the 23rd fifty yards of the curtain, between the towers of Los Hornos and Amezqueta, fell outward on to the strand that was exposed at low water—appearing to present a practicable breach. But partly to make things sure, and partly because the trenches on the isthmus were still remote from the point which it was intended to reach, the battering was continued for two days longer. Not only was more of the wall thrown down near the tower of Los Hornos, but at Graham’s desire a second and minor breach was made, some distance farther north in the sea-wall than the first[810]. The temptation to make it was that Spanish refugees brought intelligence that the ramparts here were very thin—which proved to be the case, as one day’s battering brought down a broad patch of stones. On the other hand the specialists are said to have warned Graham that the second breach would be rather useless, because the space of shore below it exposed at low tide was very narrow—under 50 feet wide—and could only be reached by skirting along for 300 yards below the walls and past the foot of the first breach. However, the creation of a second point of attack would certainly distract part of the attention of the garrison from the real place of danger, and if the small breach could be carried, the flank of the defenders of the great breach would be effectually turned.
It was a thousand pities that the formation of the first breach on July 23rd was not followed by an immediate attempt to storm, as the enemy was given two whole days to perfect his inner defences. But an assault was impossible till the approaches should have got close to the walls, and owing to the initial delays they were not ready till two days later. On the night of the 22nd-23rd the parallel reaching right across the isthmus was begun and half completed, with the French ‘Cask Redoubt’ at its left extremity. Digging close to this spot, the working parties found a large drain or channel 4 feet high and 3 feet wide, which belonged to the aqueduct which supplied the town with water. It was empty, as the Spaniards had cut it off at its source when they first blockaded the town. Lieutenant Reid of the Engineers volunteered to explore it, and crawling 230 yards forward, found that he was under the west corner of the counterscarp of the hornwork, where the French had cut off the channel and blocked it with a door. This discovery suggested to the engineers that the channel might be used as a mine; if its farther end were filled with powder, and then built up with sandbags, the explosion might bring down the counterscarp and fill the ditch, so that the hornwork might be stormed across the débris. Unfortunately there was complete uncertainty as to whether the plan would work: quite conceivably it might result in the door being blown open, and the force of the powder might be spent in sending a local blast along the ditch, without much damaging the superincumbent earth. However, the experiment was tried; layers of sandbags were placed against the door, then thirty barrels of powder were inserted, standing on end but not tamped in, for foul air greatly troubled the miners, so that there was much empty space above and between the barrels. The near end was then built up with more sandbags and a train laid to the mouth of the channel in the trenches[811].
During the day of the 23rd and the following night the parallel was completed, and good communications made from it to the base, the hill of San Bartolomé. The greater part of the enemy’s visible guns were silenced, and the enfilading fire from the two last-erected batteries on the isthmus set ablaze the quarter of the town immediately behind the two breaches. This last achievement turned out to be anything but an advantage.
For arrangements for the assault having been made for daybreak on the 24th, and the storming party and supports having been brought down to the trenches after midnight, to wait till half low-tide, it was found that the whole of the streets adjacent to the great breach were blazing so fiercely that it was thought that any entry into the town would be impossible. Graham therefore countermanded the storm—this was probably necessary; but Rey thereby gained another day for perfecting his internal defences.
These were most elaborate and excellent, surpassing by far even the very competent arrangements made by Philippon at Badajoz. The peculiar part of them was that Rey had to depend almost entirely on engineering work—his artillery having been completely crushed during the preliminary battering. Such guns as he contrived to use on this day were either very remote—on the rock and its slopes—or else pieces that had been withdrawn and kept under cover till the critical moment should arrive. The British artillery officers, whose telescopes were always busy, discovered some of them, but informed Graham that they undertook to crush them when they should be brought out, provided that light and atmospheric conditions were satisfactory[812].
But the main defence which Rey had prepared depended on fortification, not on artillery fire. His chief advantage was that at both breaches the stones had fallen outward, and the back of the ramparts was intact. All along the sea-wall there was a drop of 15 to 20 feet from the rampart-walk into the street behind: this still existed, as the core of the wall was still standing, though the facing had given way. To descend from the lip of the breach into the town, therefore, was more than a feasible leap: it could not be done without ladders. Rey had demolished some houses which had been built with their back to the ramparts, and had cut away all stairs leading up to them. He had blocked with stone barricades all streets opening towards the breaches, and had loopholed all buildings commanding a view of them. When the fire in the quarter that had been ablaze on the 24th died down, he reoccupied the ruined houses, making shelters in the débris. He had cut off the rampart-walk on each side of the breaches by building a succession of stone traverses across it, so that troops who had mounted to the top could not push out sideways. As it was clear that the storming columns would have to advance for 300 yards parallel with the eastern faussebraye of the hornwork, and close to it, because of the water on their right, he told off a large number of picked marksmen to line this flank, and ranged live shells along it, which could be rolled down on to the strand while the enemy were passing.
The preparation for a storm on the 24th, and its countermand because of the conflagration, had been noted by the governor, who understood the cause of the delay. Obviously the attack would come on the following morning, so the reserved guns were got into position after midnight—two in the casemates of the high curtain, two others in the ditch below it, one in the eastern face of the hornwork, two in the south front of the bastion of St. Elmo: into the towers of Los Hornos and Amezqueta, on each side of the main breach, three pieces were hoisted not without difficulty, since the towers, though still standing, were in bad order. Even in the distant and lofty Mirador battery, on top of the rock, some guns were trained on the point of danger[813]. Having made all possible preparations Rey waited for the assault, which came just as he had expected.
The exact share of responsibility for the details of the storm of July 25th, which fell respectively on the engineers who had settled the scheme of attack, on Graham who designated the hour, on Oswald who arranged the order of the troops, and on the artillery officers who had undertaken to keep down the enemy’s fire, is not easy to determine. All apparently have to take their portion; but that of Charles Smith, who devised, and Sir Richard Fletcher, who approved, the choice of the sea-wall for the breach-spot would seem to be the heaviest, and that of the gunners the least.
The troops were to sally out from the eastern end of the long parallel across the isthmus, when there should be a sufficient breadth of strand exposed for them to rush straight for the breaches, passing between the hornwork on their left and the receding tidal water on their right. Unfortunately the hour of low tide and the hour of daybreak were not conveniently correlated: the ideal combination would have been that the four hours during which there was good access across the exposed flats should have been from 5.30 a.m. till 9.30 a.m.: daybreak falling at about 5.20 a.m. But as it chanced the extreme of low tide that day was at 6 a.m., and the practicable hours were from 4 a.m. onward. In order to lose none of the limited time available, Graham determined that the rush should be made at 5 o’clock sharp, before it was full daylight; and this choice had one unlucky effect. The engineers settled that the signal for assault should be the blowing up of the mine in the aqueduct[814]; but they had so little confidence in its effect, that the only arrangements made to utilize it were that supposing it should do much damage and blow down scarp and counterscarp (which was hardly expected), the Portuguese companies which held the extreme left of the parallel should make a rush at the west flank of the hornwork, and enter over the débris. This was a very half-hearted scheme. For the real assault General Oswald seems to have made a very faulty disposition of his troops. The opening made in the parallel was so narrow that not more than two or three men could issue abreast. The troops, therefore, had to start in a sort of narrow file or procession, and took an unconscionable time in trickling out from the trench. The head of the column was composed of the right wing of the 3/1st (Royal Scots); then came a ladder party from all three regiments of Hay’s brigade; then the left wing of the Royal Scots: all these were to strike for the main breach. Next were the 1/38th, who were directed to pass between the leading troops and the water’s edge, and to make for the lesser breach, along the ever-narrowing strip of exposed beach. Last came the 1/9th, the third battalion of the brigade, with orders to support wherever it could make itself most useful. Some picked shots from the 8th Caçadores of Spry’s brigade were placed along the front of the parallel, and in a ditch which had been scraped during the night a few yards in advance of it, with orders to try to keep down the enemy’s musketry from the eastern flank of the hornwork.
At 5 o’clock, or perhaps a little earlier[815], the mine was fired, and did much more damage than was expected, blowing down the counterscarp of the western flank of the hornwork, injuring the scarp, and filling the ditch with earth. The Portuguese troops in the opposite trench made an assault, according to their orders, swept over the covered way and into the ditch, and tried to enter the work. They failed with loss, because no proper preparations had been made to utilize the opportunity[816], which indeed had hardly been taken into consideration. They might have got in, for the French garrison flinched at first, terrified by the explosion.
On the side of the main attack the Royal Scots led out of the parallel immediately that the explosion was heard, and pushed forward in almost complete darkness on to the strand. It was found to consist in many places of hard rock overgrown with slippery seaweed, and interspersed with deep pools. The men stumbled over, and into, many traps which might have been avoided by daylight, and all order was lost. But for the first few minutes there was little fire bearing on them, and the platoons at the head of the column reached the main breach. The conducting engineer, Lieutenant Harry Jones, and Major Fraser, commanding the wing of the Royal Scots, had actually got to the lip of the breach, with the leading men of the storming party, before the enemy really opened. That the crowning of the breach was not followed by a rush into the town was due to Rey’s precautions—there was a sudden drop of twenty feet between the top of the wall and the street below, and no means to descend, as all stairs and ramps had been cut away. There were some ladders with the column, but not at its head—they were being carried in the rear of the right wing of the Royal Scots. As the leading company came to a halt the French began to shoot hard from behind traverses on the rampart-walk, barricades, and loopholed houses, and the guns in the two flanking towers played on the breach with grape. The head of the column began to wither away; Fraser and Jones were both wounded—the former mortally—and so was every man who had reached the summit. The survivors on the breach threw themselves down among the stones and began to return the enemy’s fire—the first impetus being lost, and any chance of success with it.
But it was only about the equivalent of a company which had advanced so far; the rear of the right wing of the Royal Scots were still passing along the flank of the hornwork, in a straggling file, when the enemy’s fire began to be serious. It came most fiercely from an entrenchment which Rey had thrown up across the main ditch between the hornwork and the high front of the demi-bastion behind it. It seems, from the narrative of an officer who took a prominent part in the storm—Colin Campbell of the 1/9th, who was with the ladder party—that many of the Royal Scots, arriving in the darkness at this opening into the main ditch, mistook it for a passage by which they might force their way into the place, or even for the great breach itself. At any rate they turned in toward it, and on meeting with strenuous resistance began to fire upon the enemy. The men following gathered in upon the first comers, and a crowd accumulated at this point, which checked the ladder party and the left wing of the Royal Scots as they came up. Hardly any one pressed on to join the head of the column on the slope of the breach. Meanwhile the tail of the column was blocked, as it tried to press on past the hornwork. The French, standing above, rolled down the shells which had been laid ready upon the British below, and kept up a vigorous discharge of musketry. With great exertions individual company officers succeeded in collecting parties of their own men, and leading them out of the crowd, so as to pass on to the great breach. But the attack there had already come to nothing: the stones all up the slope were strewn with dead and wounded—a few survivors were keeping up an ineffectual return fire upon the well-concealed enemy. Three or perhaps four attempts to mount again were made by small parties of the rear companies of the Royal Scots, but there were never more than 80 or 90 men acting together, and the officers and leading men were always shot down on the crest. At last the senior captain surviving, seeing the impossibility of getting forward, ordered the stormers to retire, which they did—only half an hour or less after the assault commenced—though the time had seemed much longer, as was natural, to those involved in the bloody business. The larger body of men engaged farther back, opposite the main ditch and under the demi-bastion of St. Juan, gave way also when the head of the column fell back among them: they had themselves suffered heavily, and of course had made no progress. Just as the whole body rolled back along the beach they came into collision with the front of the 1/38th, who had only just finished filing out of the parallel, so slow was the process of emerging from its narrow exit. Their commanding officer, Colonel Greville, had halted them for a short time to let their rear close up, and to prevent them from dribbling forward in a thin string of small parties, as the Royal Scots had done. Just as they came parallel with the north end of the hornwork the broken mass of men from the front ran in upon them; all order was lost at once, and after some vain attempts to get forward the 38th fell back along with the rest over the slippery shore[817]. The disordered crowd suffered heavily from the French grape and musketry—they were a mark impossible to miss, and strewed the rocks and pools with dead. The 1/9th, who were just beginning to file out of the parallel, were of course ordered back at once—‘but had lost almost as many heads as they showed[818].’
So quickly was the whole affair over that the artillerymen, standing by their guns on the opposite side of the Urumea, ready to co-operate as soon as dawn should come, believed at first that there had been a feint or a false attack. It was only when the growing light showed them the breach strewn from lip to foot with red coats, and the strand below thickly dotted with them also, that they realized that the assault had been made and had failed[819], without their being given the chance to intervene as they had promised.
The loss had, of course, been very heavy—out of 571 casualties of all ranks[820] more than 330 belonged to the unfortunate Royal Scots, whose right wing companies were almost exterminated. Six officers and 118 men, almost all wounded, fell into the hands of the enemy. It should be remembered to the credit of the garrison that, the moment that the storm was over, they collected from the foot of the breach and the neighbouring strand many officers and men who would otherwise have been drowned by the returning tide[821]. When it was seen that this was the task of the French who were seen busy on the beach, a flag of truce was hoisted, and all firing ceased for an hour. The prisoners, one of whom, the engineer Harry Jones, has left an interesting account of his experiences during the next six weeks[822], were very well treated. The extremely moderate loss of the garrison was 18 killed and 49 wounded—all by musketry fire, since by the abominable misarrangement of the assault the British artillery had no chance of acting.
Two comments by eye-witnesses on this woeful business are worth giving. Colin Campbell wrote:
‘One main cause of failure was the narrow front and consequent length and thinness of the column in which we advanced. This necessarily became more loosened and disjointed by the difficult nature of the ground it had to pass over in the dark. It reached the breach in driblets and never in such body or number as to give the mind of the soldier anything like confidence in success. If some means had been devised of starting the Royals in one big honest lump, which might have been contrived without much difficulty or danger, so that they could have started in some dense form, with the 38th well packed up in a front of fours in readiness to start immediately behind after them, the stoppage at the demi-bastion would never have occurred, and some 200 men at least of the Royals would have reached the breach in a compact body. Such a number would have forced bodily through all opposition. Even under all the disadvantages of bad arrangements, I firmly believe that if we had moved forward by daylight, when an officer could have seen, and been seen by, his men, when the example of the former would have animated the exertions of the latter, the Royals would have gone over the breach on July 25th[823].’
Campbell’s blame, therefore, would fall mainly on Oswald, who, though he had protested against the points chosen for attack, had actually arranged the troops, and failed to remember that men should go ‘over the top’ on a broad front, and on Graham, who fixed the ‘Zero’ point half an hour before dawn. Gomm, of the 1/9th, also a most distinguished and capable officer, goes for other game—the officers of the scientific services:
‘I am afraid our success at Ciudad Rodrigo and Badajoz, owing to the almost miraculous efforts of the troops, has stopped the progress of science among our engineers, and perhaps done more; for it seems to have inspired them with a contempt for so much of it as they had attained before. Our soldiers have on all occasions stood fire so well that our artillery have become as summary in their processes as our engineers. Provided that they have made a hole in the wall, by which we can claw up, they care not about destroying defences. In fact, we have been called upon hitherto to ensure the success of our sieges by the sacrifice of lives. Our Chief Engineers and Commandants of Artillery remind me of Burke’s “Revolutionary Philosophers” and their “dispositions which make them indifferent to the cause of humanity; they think no more of men than of mice in an air pump”. We came before the place well equipped with all the means necessary for attacking it en règle, and I saw no reason for attacking it otherwise. I may dwell longer than I ought to do on this subject. But it is, at least, pardonable in us, who are most nearly concerned, to become tedious in passing our censure upon the methods of those whom we cannot but consider as the authors of our calamity; which, as it was foreseen by others, and might have been by them, could have been avoided.’
The narrative of the assault seems to show that both Campbell and Gomm had reason for their complaints, though they chose different points to criticize. But of the various errors made, undoubtedly the most fatal of all was attempting to storm breaches that could only be approached by a long defile along the flank of the hornwork, which was intact and well garrisoned. And for this the engineers have to take the responsibility. Camp rumour very cruelly put the blame on the troops[824], alleging that there had been a panic: this was a monstrous injustice. Everything that mismanagement could accomplish had been done to discourage them; but it was not poor spirit, but physical incapacity to finish a task impossible under the conditions set, that caused them to retire.