CHAPTER II

THE BRIDGE OF ALMARAZ. MAY 1812

On April 24th Wellington halted his pursuing army at Fuente Guinaldo and Sabugal, on hearing that Marmont had escaped him by a margin of twenty-four hours. The French were in full march for Salamanca, and it was impossible to pursue them any further, firstly because the allied army needed a few days of rest after the forced march from Badajoz, and secondly because its train had dropped behind, food was nearly out, and convoys had to be brought up from Lamego and São João de Pesqueira. There was, of course, nothing to be got out of the unhappy region in which Marmont’s locusts had just been spread abroad. The only fortunate thing was that the Duke of Ragusa had turned his raid against the Beira Baixa, and left the great dépôts on the Douro unmolested. From them ample sustenance could be got up, in a week, to the positions behind the Agueda and Coa where the army had halted.

Wellington, as it will be remembered, had contemplated an attack on Andalusia after Badajoz fell. But the necessity for seeing to the relief and revictualling of Almeida and Ciudad Rodrigo had brought him up to the frontiers of Leon with the main body of his host. In the position where he now lay, he was well placed for an advance on Salamanca, and an attack on the Army of Portugal. To return to Estremadura would involve a long and weary countermarch. Moreover there was no doubt that operations in Leon would be more decisive than operations in Andalusia. As Marmont was to write to Berthier a few days later, a victory of the allies in the North would involve the evacuation of the South by Soult, while a victory in Andalusia would leave the French power in the valleys of the Douro and Tagus unshaken[350]. Advancing from the line of the Agueda against Salamanca and Valladolid, Wellington would have his base and his main line of communications in his direct rear, safe against any flank attack. A raid against Andalusia, even if successful, would separate him from Lisbon, and compel him to take up a new base at Cadiz--a doubtful expedient. But what seems, in the end, to have been the main cause for Wellington’s choosing Leon rather than Andalusia as his next sphere of operations, was that Marmont (as he judged) had the larger available army for field movements outside his own ground. Soult was more pinned down to his viceroyalty by local needs: he would not raise the siege of Cadiz or evacuate Granada and Cordova. Therefore he could not collect (as his movement at the time of the fall of Badajoz had shown) more than 24,000 men for an offensive operation. This was the absolute limit of his power to aid Marmont. But the latter, if he chose to evacuate Asturias and other outlying regions, could bring a much larger force to help Soult. Therefore an attack on Andalusia would enable the enemy to concentrate a more numerous defensive force than an attack on Leon. ‘Of the two armies opposed to us that of Portugal can produce the larger number of men for a distant operation. Marmont has nothing to attend to but the British army, as he has been repeatedly told in [intercepted] time he may lose some plunder and contributions, but he loses nothing that can permanently affect his situation, or which he could not regain as soon as he has a superiority, particularly of cavalry, in the open plains of Castille. Marmont’s, then, being what may be called of the two the _operating_ army, the movement which I might make into Andalusia would enable the enemy to bring the largest body of men to act together on one point. It would be a false movement, and this must by all means be avoided[351].’

[350] See above, p. 311.

[351] _Dispatches_, ix. p. 173.

This decision was not made immediately on Marmont’s retreat of April 24th: for some days after the British headquarters settled down at Fuente Guinaldo, Wellington had not quite made up his mind between the two operations: his letters to Lord Liverpool, to Hill, and Graham, are full of the needs of the moment, and do not lay down any general strategical plan. The staff, in their discussions with each other, canvassed the situation. ‘While Marmont remains in Old Castile he [Wellington] must leave a certain force near the frontier of the Beira. But leaving the 3rd, 4th, 5th Divisions, and Pack’s and Bradford’s Portuguese (perhaps 18,000 men) for that purpose, he can move upon Andalusia, if he wishes, with the 1st, 6th, 7th, and Light Divisions, afterwards picking up Power’s Portuguese brigade and all General Hill’s _corps d’armée_--perhaps 36,000 infantry. This would do.’ So wrote D’Urban the chief of the Portuguese staff in his private diary, on May 5, evidently after discussion with Beresford, and others of those who were nearest the centre of decision. Wellington, however, was pondering over alternatives: he could not move for a week or two at the best, for he had to replenish his stores at the front, and to see that the repairs and revictualling of Almeida and Rodrigo were completed, before he could start on any offensive movement. In that time, too, he would be able to learn how Marmont was disposing of his army, and whether Soult was showing any tendency to reinforce Drouet’s force in Estremadura.

It seems that an insight into his enemies’ purposes was made specially easy for Wellington at this moment by the successive capture of a great deal of French correspondence. When Marmont was in Portugal, between the 1st and 23rd of April, three of the duplicates of his dispatches were captured, one by Portuguese Ordenança, the others by Julian Sanchez between Rodrigo and Salamanca[352]. They were all in cipher, but the ingenuity of Captain Scovell, the cipher-secretary at head-quarters, was capable of dealing with them, and from them could be made out a great deal about the strength of the Marshal’s army, and his general views on the campaign. If they had been taken and sent in a little earlier, they might have enabled Wellington to complete that surprise and dispersion of the French expeditionary force which had been in his mind.

[352] The cipher-originals are all in the Scovell papers, worked out into their interpretation by that ingenious officer: Wellington only kept the fair copies for himself. The dispatches are dated Sabugal, 11 April (to Brennier about the Agueda bridge); Sabugal, April 16 (to Berthier); Fuente Guinaldo, April 22 (to Berthier). The last two are full of the most acrimonious criticism of Napoleon’s orders for the invasion of Beira. Scovell made out much, but not all, of the contents of these letters.

But though they arrived too late for this purpose, they were valuable, as showing Marmont’s dislike of the imperial orders that he had been sent to carry out, and his preference for his own schemes. They were also full of bitter complaints of the neglect in which the Army of Portugal was left as to pay, stores, and transport. Wellington might reasonably deduce from them that any reconcentration of that army would be slow, and that if it had to march to reinforce Soult in the South, the effort would be a severe one.

But shortly after Marmont’s return to Salamanca, his adversary got an even more valuable insight into his plans. The guerrilleros carried off, between Salamanca and Valladolid, an officer bearing five dispatches, dated April 28 and April 30th. One was directed to Dorsenne, two to Berthier, one to Jourdan, the fifth contained the parole to Bayonne of the great scout, Colquhoun Grant[353]. The first, couched in very peremptory terms, asked for food--the Army of Portugal must absolutely receive 8,000 quintals of wheat, once promised, without delay--it was in a state of danger and penury, and could not keep concentrated to face the British. Of the letters to Berthier one announced that Bonnet’s division was duly in march for the Asturias, and that without it the Marshal thought his own strength dangerously low. The other asked for 4,000,000 francs owing to the Army of Portugal for pay and sustenance, and declared that, unless money came to hand at once, it was impossible to see how the troops were to be kept alive in the two months still remaining before harvest. A postscript asked for a siege-train to be sent on at all costs--the Marshal had heard that one was on the way from Bayonne: but nothing was known about it at Burgos. The letter to Jourdan was the most important of all[354]: it was the document, already quoted in the previous chapter, in which the Marshal detailed his intentions as to the dispersion of his army, protested against being obliged to send too many men into the valley of the Tagus, and explained the importance of the bridge-forts and magazines at Almaraz, by which his troops at Avila, &c., would debouch southward whenever they were ordered to concentrate for a junction with Soult. ‘On ne peut agir que par Lugar Nuevo [the name by which Marmont always designates the Almaraz forts] ... il faut bien se garder de jeter trop de troupes sur le Tage, et se contenter de bien assurer une défense de huit jours pour les forts de Lugar Nuevo et Mirabete, temps suffisant pour que les troupes rassemblées à Avila débouchent.... Un dépôt de 400 à 500 mille fanegas (qui n’est pas au delà de ce que Madrid et La Manche peuvent fournir) donnerait les moyens d’agir sans compromettre la subsistance des troupes.’

[353] All the originals are in the Scovell Papers.

[354] It is the one printed in Ducasse’s _Correspondence of King Joseph_, viii. pp. 413-17.

Undoubtedly it was the deciphering of the greater part of this letter, which set forth so clearly the importance of the Almaraz bridge, and showed at the same time that only one French division [Foy’s at Talavera] was anywhere near it, that determined Wellington to make the sudden stroke at that central strategical point which he had thought of in February[355]. At that time he had refused to try it, because there were three French divisions on the Tagus. Now there was only one at Talavera, two marches from Almaraz, and the nearest reinforcements at Avila were two very long marches from Talavera. The possibility presented itself that a column might strike at Almaraz from somewhere on the Portuguese frontier, and take the place by a _coup-de-main_, with or without first beating Foy, whose strength of 5,000 men was perfectly known to Wellington.

[355] See above, p. 202.

Hill could count on two or three days of undisturbed operations before the nearest reinforcing division, that of Foy, could reach Almaraz: on four or five more, before troops from Avila could come up. It must be noted that everything would depend on the absolute secrecy that could be preserved as to the start of the expedition: but on this Wellington thought that he could count. The Spanish peasantry seldom or never betrayed him: the French had no outlying posts beyond Almaraz which might give them warning. The garrison was in a normal state of blockade by guerrillero bands haunting the Sierra de Guadalupe.

It may be added that a blow at Almaraz was just as useful as a means for keeping Soult from joining Marmont as Marmont from joining Soult. It would be profitable if Wellington’s final decision should be given in favour of an Andalusian expedition. But his mind was by now leaning towards an attack on Leon rather than on the South. The final inclination may have been given by the receipt of another intercepted dispatch--Soult’s to Jourdan of April 17[356], sent in by guerrilleros who had probably captured the bearer in the Sierra Morena about April 20th. This document, which we have already had occasion to quote for another purpose[357], was full of angry denunciations of Marmont for letting Badajoz fall unaided, and served to show that, if Soult had to help the Army of Portugal, he would do so with no good will to its commander. Moreover it was largely occupied by proposals for the circumventing of Ballasteros and the siege of Tarifa--movements which would disperse the Army of the South even more than it was already dispersed, and would clearly prevent it from succouring Marmont within any reasonable space of time.

[356] Original in the Scovell Papers. Place of capture uncertain, but clearly taken by guerrilleros between Seville and Madrid.

[357] See above, pp. 269-70.

The decision that Hill should make his long-deferred _coup-de-main_ upon Almaraz first appears in Wellington’s dispatches on May 4th[358], but Hill had been warned that the operation was likely to be sanctioned some days earlier, on April 24, and again more definitely on April 30th[359]. That the final judgement of Wellington was now leaning in favour of the advance on Salamanca rather than the Andalusian raid appears to emerge from a note of D’Urban dated May 6th--’The retirement of Marmont within a given distance--the slow progress of the Spaniards at Rodrigo, which renders it unsafe to leave that place and this frontier--the retiring altogether of Soult, and the state of his army not making him dangerous now--these and other combining reasons determine Lord Wellington to make his offensive operation _north_ of the Tagus, and to move upon Marmont. All necessary preparations making, but secretly: it will be very feasible to keep the movement unforeseen till it begins. Meanwhile General Hill is to move upon and destroy everything at Almaraz[360].’

[358] Wellington to Graham, Fuente Guinaldo, May 4, _Dispatches_, ix. p. 114.

[359] Ibid., p. 101.

[360] D’Urban’s unpublished diary, under May 6.

The orders for Hill’s move were given out on May 7th. He was to march from his head-quarters at Almendralejo with two British brigades (Howard’s and Wilson’s) of the 2nd Division, and the Portuguese brigade attached to the division (Ashworth’s), one British cavalry regiment (13th Light Dragoons), and to cross the Guadiana at Merida. Beyond the Guadiana he would pick up Campbell’s Portuguese cavalry brigade, which was lying at Arroyo dos Molinos. The march was then to be as rapid as possible, via Jaraicejo and Miravete. The expeditionary force made up 7,000 men in all.

There were left in Estremadura to ‘contain’ Drouet the two English cavalry brigades of Hill’s force (Slade’s and Long’s)[361], one British infantry brigade (Byng’s) of the 2nd Division, Hamilton’s Portuguese division, and Power’s unattached Portuguese brigade (late the garrison of Elvas, and more recently acting as that of Badajoz). The whole would make up 11,000 men. Power, or at least some of his regiments, was now disposable, because the Spaniards destined to hold Badajoz had begun to arrive, and more were daily expected[362].

[361] Minus, of course, the 13th Light Dragoons.

[362] Erskine was the senior officer left with the corps--a dangerous experiment. One marvels that Wellington risked it after previous experience.

But this was not the only precaution taken against Drouet, who had recently been reported as a little inclined to move northward from Fuente Ovejuna--detachments of his cavalry had been seen as far north as Zalamea[363]. Wellington determined to move down towards the Guadiana the southern or right wing of his main army--the 1st and 6th Divisions under Graham. First one and then the other were filed across the bridge of Villa Velha and sent to Portalegre. Here they would be in a position to support the force left in front of Drouet, if Soult should unexpectedly reinforce his Estremaduran corps. Wellington acknowledged that he disliked this wide extension of his army, but justified himself by observing that, if he had now his left wing almost touching the Douro, and his right wing almost touching the Sierra Morena, he might risk the situation, because he was fully informed as to Marmont’s similar dispersion. The Army of Portugal was scattered from the Asturias to Talavera, and from its want of magazines and transport, which Marmont’s intercepted dispatches made evident, would be unable to concentrate as quickly as he himself could.

[363] Wellington to Graham, May 7, _Dispatches_, ix. p. 128.

The movement of Graham’s two divisions from the Castello Branco region to south of the Tagus had an additional advantage. If reported to the French it would tend to make them believe that the next offensive operation of the allied army would be in the direction of Andalusia, not towards the Tormes. If Soult heard of it, he would begin to prepare to defend his own borders, and would not dream that Marmont was really the enemy at whom Wellington was about to strike; while Marmont, on the other hand, thinking that Soult was to be the object of Wellington’s attentions, might be less careful of his own front. The expedition to Almaraz would not undeceive either of them, since it was well suited for a preliminary move in an attack on Andalusia, no less than for one directed against Leon.

Hill’s column reached Merida on May 12th, but was delayed there for some hours, because the bridge, broken in April, had not yet been repaired, as had been expected, the officers sent there having contented themselves with organizing a service of boats for the passage. The bridge was hastily finished, but the troops only passed late in the day; they picked up in the town the artillery and engineers told off for the expedition, Glubb’s British and Arriaga’s Portuguese companies of artillery, who brought with them six 24-pounder howitzers, a pontoon train, and wagons carrying some 30-foot ladders for escalading work. The importance attached to the raid by Wellington is shown by the fact that he placed Alexander Dickson, his most trusted artillery officer, in charge of this trifling detachment, which came up by the road north of the Guadiana by Badajoz and Montijo to join the main column.

Once over the Guadiana, Hill reached Truxillo in three rapid marches [May 15], and there left all his baggage-train, save one mule for each company with the camp-kettles. The most difficult part of the route had now been reached, three successive mountain ranges separating Truxillo from the Tagus. On the 16th, having crossed the first of them, the column reached Jaraicejo: at dawn on the 17th, having made a night march, it was nearing the Pass of Miravete, the last defile above the river. Here, as Hill was aware, the French had outlying works, an old castle and two small forts, on very commanding ground, overlooking the whole defile in such a way that guns and wagons could not possibly pass them. The British general’s original intention was to storm the Miravete works at dawn, on the 17th, and at the same time to attack with a separate column the forts at the bridge. With this purpose he divided his troops into three detachments. Ashworth’s Portuguese and the artillery were to keep to the _chaussée_, and make a demonstration of frontal attack on the Castle: General Tilson-Chowne [interim commander of the 2nd Division at the moment[364]] was, with Wilson’s brigade and the 6th Caçadores, to make a détour in the hills to the left and to endeavour to storm the Castle from its rear side. General Howard, with the other British brigade, was to follow a similar bridle path to the right, and to descend on to the river and attack the forts by the bridge.

[364] This was the Tilson of 1809: he had lengthened his name.

A miscalculation had been made--the by-paths which the flanking columns were to take proved so far more steep and difficult than had been expected, that by dawn neither of them had got anywhere near its destination. Hill ordered them to halt, and put off the assault. This was fortunate, for by a long and close reconnaissance in daylight it was recognized that the Castle of Miravete and its dependent outworks, Forts Colbert and Senarmont, were so placed on a precipitous conical hill that they appeared impregnable save by regular siege operations, for which the expeditionary force had no time to spare. The most vexatious thing was that the garrison had discovered the main column on the _chaussée_, and it could not be doubted that intelligence must have been sent down to the lower forts, and most certainly to Foy at Talavera also. After a thorough inspection of the ground, Hill concluded that he could not hope to master Miravete, and, while it was held against him, his guns could not get through the pass which it so effectively commanded. It remained to be seen what could be done with the forts at the bridge.

The Almaraz forts crowned two hills on each side of the Tagus. The stronger, Fort Napoleon, occupied the end of a long rising ground, about 100 yards from the water’s edge; below it, and connecting with it, was a masonry _tête-de-pont_ covering the end of the pontoon-bridge. The weaker work, Fort Ragusa, was on an isolated knoll on the north bank, supporting the other end of the bridge. Fort Napoleon mounted nine guns, had a good but unpalisaded ditch around its bastioned front, and a second retrenchment, well palisaded, with a loopholed stone tower within. Fort Ragusa was an oblong earthwork mounting six guns, and also provided with a central tower. It had as outwork a _flèche_ or lunette, commanding the north end of the bridge. The small _tête-de-pont_ mounted three guns more. Half a mile up-stream was the ruined masonry bridge which had formed the old crossing, with the village of Almaraz on the north bank behind it. Between the _tête-de-pont_ and the old bridge were the magazines and storehouses in the village of Lugar Nuevo.

The garrison of the works consisted of a depleted foreign corps, the _régiment de Prusse_ or 4th Étranger, mustering under 400 bayonets, of a battalion of the French 39th of the Line, and of two companies of the 6th Léger, from Foy’s division, with a company of artillery and another of sappers. The whole may have amounted to 1,000 men, of whom 300 were isolated in the high-lying Castle of Miravete, five miles from the bridge-head. The governor, a Piedmontese officer named Aubert, had manned Fort Napoleon with two companies of the 6th and 39th. The foreign corps and one company of the 6th were in Fort Ragusa and the bridge-head; Miravete was held by the centre companies of the 39th.

Though delay after the French had got the alarm was dangerous, Hill spent the whole of the 17th in making fruitless explorations for vantage-ground, from which Miravete might be attacked. None was found, and on the 18th he made up his mind to adopt a scheme hazardous beyond his original intention. It would be possible to mask the Castle by a false attack, in which all his artillery should join, and to lead part of his infantry over the hills to the right, by a gorge called the Pass of La Cueva, for a direct attack by escalade, without the help of guns, upon the Almaraz forts.

The detachment selected for this purpose was Howard’s brigade (1/50th, 1/71st, 1/92nd), strengthened by the 6th Portuguese Line from Ashworth’s brigade, and accompanied by 20 artillerymen in charge of the ladders. So rough was the ground to be covered, that the long 30-foot ladders had to be sawn in two, being unwieldy on slopes and angles, as was soon discovered when they were taken off the carts for carriage by hand. The route that had to be followed was very circuitous, and though the forts were only five miles, as the crow flies, from the place where the column left the road, it took the whole night to reach them. An eye-witness[365] describes it as a mazy sheep-walk among high brushwood, which could not have been used without the help of the experienced peasant-guide who led the march. The men had to pass in Indian file over many of its stretches, and it resulted from long walking in the darkness that the rear dropped far behind the van, and nearly lost touch with it. Just before dawn the column reached the hamlet of Romangordo, a mile from the forts, and rested there for some time before resuming its march.

[365] Captain MacCarthy of the 50th.

The sun was well up when, at 6 o’clock, the leading company, coming to the edge of a thicket, suddenly saw Fort Napoleon only 300 yards in their front. The French had been warned that a column had crossed the hills, and had caught some glimpse of it, but had lost sight of its latest move: many of the garrison could be seen standing on the ramparts, and watching the puffs of smoke round the Castle of Miravete, which showed that the false attack on that high-lying stronghold had begun. General Tilson-Chowne was making a noisy demonstration before it, using his artillery with much ostentation, and pushing up skirmishers among the boulders on the sides of the castle-hill[366].

[366] The statement in Jones’s _Sieges_, i. p. 259, that the enemy were unaware of the turning column is disproved by the official reports of the surviving French officers Sêve and Teppe.

Hill was anxious to assault at once, before the sun should rise higher, or the garrison of the forts catch sight of him. But some time had to be spent to allow a sufficient force to accumulate in the cover where the head of the column was hiding. So slowly did the companies straggle in, that the General at last resolved to escalade at once with the 50th and the right wing of the 71st, all that had yet come up. Orders were left behind that the left wing of the 71st and the 92nd should attack the bridge-head entrenchment when they arrived, and the 6th Portuguese support where they were needed.

At a little after 6 o’clock the 900 men available, in three columns of a half-battalion each, headed by ladder parties, started up out of the brake on the crest of the hillside nearest Fort Napoleon, and raced for three separate points of its _enceinte_. The French, though taken by surprise, had all their preparations ready, and a furious fire broke out upon the stormers both from cannon and musketry. Nevertheless all three parties reached the goal without any very overwhelming losses, jumped into the ditch, and began to apply their ladders to such points of the rampart as lay nearest to them. The assault was a very daring one--the work was intact, the garrison adequate in numbers, the assailants had no advantage from darkness, for the sun was well up and every man was visible. All that was in their favour was the suddenness of their onslaught, the number of separate points at which it was launched, and their own splendid dash and decision. Many men fell in the first few minutes, and there was a check when it was discovered that the ladders were over-short, owing to their having been sawn up before the start. But the rampart had a rather broad berm[367], a fault of construction, and the stormers, discovering this, climbed up on it, and dragging some of the ladders with them, relaid them against the upper section of the defences, which they easily overtopped. By this unexpected device a footing was established on the ramparts at several points simultaneously--Captain Candler of the 50th is said to have been the first man over the parapet: he was pierced by several balls as he sprang down, and fell dead inside. The garrison had kept up a furious fire till the moment when they saw the assailants swarm over the parapet--then, however, there can be no doubt that most of them flinched[368]: the governor tried to lead a counter-charge, but found few to follow him; he was surrounded, and, refusing to surrender and striking at those who bade him yield, was piked by a sergeant of the 50th and mortally wounded. So closely were the British and French mixed that the latter got no chance of manning the inner work, or the loopholed tower which should have served as their rallying-point. Many of the garrison threw down their arms, but the majority rushed out of the rear gate of the fort towards the neighbouring redoubt at the bridge-head. They were so closely followed that pursuers and pursued went in a mixed mass into that work, whose gunners were unable to fire because their balls would have gone straight into their own flying friends. The foreign garrison of the _tête-de-pont_ made little attempt to resist, and fled over the bridge[369]. It is probable that the British would have reached the other side along with them if the centre pontoons had not been sunk: some say that they were struck by a round-shot from Fort Ragusa, which had opened a fire upon the lost works; others declare that some of the fugitives broke them, whether by design or by mischance of overcrowding[370].

[367] The berm is the line where the scarp of the ditch meets the slope of the rampart: the scarp should be perpendicular, the rampart-slope tends backward, hence there is a change on this line from the vertical to the obtuse in the profile of the work. The berm should have been only a foot or so wide and was three.

[368] The official report of the French captain, Sêve of the 6th Léger, accuses the grenadiers of the 39th of giving way and bolting at the critical moment, and this is confirmed by the report of the _chef de bataillon_ Teppe of the 39th, an unwilling witness.

[369] According to Teppe’s narrative they left the walls, and many hid in the bakehouses, while most of the officers headed the rush for the bridge.

[370] Foy says that the centre link of the bridge was not a regular pontoon but a river boat, which could be drawn out when the garrison wanted to open the bridge for any purpose, and being light it collapsed under the feet of the flying crowd (p. 163).

This ought to have been the end of Hill’s sudden success, since passage across the Tagus was now denied him. But the enemy were panic-stricken; and when the guns of Fort Napoleon were trained upon Fort Ragusa by Lieutenant Love and the twenty gunners who had accompanied Hill’s column, the garrison evacuated it, and went off with the rest of the fugitives in a disorderly flight towards Naval Moral. The formidable works of Almaraz had fallen before the assault of 900 men--for the tail of Hill’s column arrived on the scene to find all over[371]. Four grenadiers of the 92nd, wishing to do something if they had been disappointed of the expected day’s work, stripped, swam the river, and brought back several boats which had been left moored under Fort Ragusa. By means of these communication between the two banks was re-established, and the fort beyond the river was occupied[372].

[371] The 92nd and the right wing of the 71st reached the _tête-de-pont_ just as the fugitives from Fort Napoleon entered it, and swept away the garrison. They only lost two wounded.

[372] Gardyne’s history of the 92nd gives the names of two of these gallant men, Gauld and Somerville.

The loss of the victors was very moderate--it fell mostly on the 50th and 71st, for Chowne’s demonstration against Miravete had been almost bloodless--only one ensign and one private of the 6th Caçadores were wounded. But the 50th lost one captain and 26 men killed, and seven officers and 93 men wounded, while the half-battalion of the 71st had five killed and five officers and 47 men wounded[373]. The 92nd had two wounded. Thus the total of casualties was 189.

[373] Hill’s total of casualties is 2 officers and 31 men killed: 13 officers and 143 wounded. The second officer killed was Lieutenant Thiele of the Artillery of the K.G.L., accidentally blown up by a mine on the day of the evacuation. But two of the wounded officers died.

Of the garrison the 4th Étranger was pretty well destroyed--those who were neither killed nor taken mostly deserted, and its numbers had gone down from 366 in the return of May 15 to 88 in that of July 1. The companies of the 39th and 6th Léger also suffered heavily, since they had furnished the whole of the unlucky garrison of Fort Napoleon. Hill reports 17 officers and 262 men taken prisoners, including the mortally wounded governor and a _chef de bataillon_ of the 39th[374]. It is probable that the whole loss of the French was at least 400.

[374] Teppe by name, whose narrative, written in captivity, is our best source for the French side. It is a frank confession of misbehaviour by the troops--particularly the 4th Étranger.

The trophies taken consisted of a colour of the 4th Étranger, 18 guns mounted in the works, an immense store of powder and round-shot, 120,000 musket cartridges, the 20 large pontoons forming the bridge, with a store of rope, timbers, anchors, carriages, &c., kept for its repair, some well-furnished workshops, and a large miscellaneous magazine of food and other stores. All this was destroyed, the pontoons, &c., being burnt, while the powder was used to lay many mines in the forts and bridgehead, which were blown up very successfully on the morning of the 20th, so that hardly a trace of them remained. Thiele of the German artillery, the officer charged with carrying out the explosions, was unfortunately killed by accident: a mine had apparently failed; he went back to see to its match, but it blew up just as he was inspecting it.

[Illustration: ALMARAZ]

Having accomplished his purpose with complete success, Hill moved off without delay, and by two forced marches reached Truxillo and his baggage on the 21st. Here he was quite safe: Foy, being too weak to pursue him to any effect, followed cautiously, and only reached Miravete (whose garrison he relieved) on the 23rd and Truxillo on the 25th, from whence he turned back, being altogether too late. He had received news of Hill’s movement rather late on the 17th, had been misinformed as to his strength, which report made 15,000 men instead of the real 7,000, and so had been disposed to act cautiously. He had ordered a battalion of the 6th Léger from Naval Moral to join the garrison of Almaraz, but it arrived on the afternoon of the 19th, only in time to hear from fugitives of the disaster[375]. He himself was confident that the forts could hold out eight days even against artillery, which was also Marmont’s calculation. Hence their fall within 48 hours of Hill’s appearance was a distressing surprise: Foy had calculated on being helped not only by D’Armagnac from Talavera but by the division of Clausel from Avila, before moving to fight Hill and relieve them.

[375] D’Armagnac also sent the battalion of Frankfort for the same purpose, which arrived late with less excuse. See Foy, p. 375.

Wellington appears to have been under the impression that this expedition, which Hill had executed with such admirable celerity and dispatch, might have been made even more decisive, by the capture of the castle of Miravete, if untoward circumstances had not intervened. In a letter to Lord Liverpool, written on May 28[376], he expresses the opinion that Tilson-Chowne might have taken it on the night of the 16th--which must appear a hazardous decision to those who look at the precipitous position of the place and the strength of its defences. He also says that Hill might have stopped at Almaraz for a few days more, and have bombarded Miravete with Dickson’s heavy howitzers, if he had not received false news from Sir William Erskine as to Drouet’s movements in Estremadura. There can be no doubt, as we shall see, about the false intelligence: but whether the bombardment would have been successful is another thing. Probably Wellington considered that the garrison would have been demoralized after what had happened at Almaraz.

[376] _Dispatches_, ix. p. 189.

As to Drouet’s movements, having received rather tardy notice of Hill’s northward march from Merida, he had resolved to make a push to ascertain what was left in his front. Lallemand’s dragoons, therefore, pressed out in the direction of Zafra, where they came into contact with Slade’s outposts and drove them in. At the same time Drouet himself, with an infantry division and some light cavalry, advanced as far as Don Benito, near Medellin, on the 17th May, from whence he pushed patrols across the Guadiana as far as Miajadas. This movement, made to ascertain whether Hill had departed with his whole corps, or whether a large force had been left in Estremadura, was reported to Sir William Erskine, the commander of the 2nd cavalry division, along with rumours that Soult was across the Sierra Morena and closely supporting Drouet. Erskine sent on the news to Graham at Portalegre, and to Hill, who was then before Miravete, with assertions that Soult was certainly approaching. This, as Wellington knew, was unlikely, for the Marshal had been before Cadiz on the 11th, and could not possibly have crossed the Sierra Morena by the 17th. As a matter of fact he only learnt on the 19th, at Chiclana, that Hill had started, and Drouet’s move was made purely to gain information and on his own responsibility. But Graham, naturally unaware of this, brought up his two divisions to Badajoz, as he had been directed to do if Estremadura were attacked during Hill’s absence. And Hill himself was certainly induced to return promptly from Almaraz by Erskine’s letter, though it is doubtful whether he would have lingered to besiege Miravete even if he had not received it. For Foy might have been reinforced by D’Armagnac and the Avila division up to a strength which would have made Hill’s longer stay on the Tagus undesirable.

Drouet did no more; indeed, with his own force he was quite helpless against Hill, since when he discovered that there was a large body of allied troops left in Estremadura, and that more were coming up, it would have been mad for him to move on Merida, or take any other method of molesting the return of the expedition from Almaraz. Though Soult spoke of coming with a division to his aid, the succours must be many days on the way, while he himself could only act effectively by marching northward at once. But if he had taken his own division he would have been helpless against Hill, who could have beaten such a force; while if he had crossed the Guadiana with his whole 12,000 men, he would have been cut off from Soult by the ‘uncontained’ allied force left in Estremadura, which he knew to be considerable.

But to move upon Almaraz on his own responsibility, and without Soult’s orders, would have been beyond Drouet’s power: he was a man under authority, who dared not take such a step. And when Soult’s dispatches reached him, they directed him not to lose touch with Andalusia, but to demonstrate enough to bring Hill back. The Marshal did not intend to let Drouet get out of touch with him, by bidding him march toward the Tagus.

Hill’s column, then, was never in any danger. But Wellington, who had for a moment some anxiety in his behalf, was deeply vexed by Erskine’s false intelligence, which had given rise to that feeling, and wrote in wrath to Henry Wellesley and Graham[377] concerning the mischief that this very incapable officer had done. He was particularly chagrined that Graham had been drawn down to Badajoz by the needless alarm, as he was intending to bring him back to join the main army within a short time, and the movement to Badajoz had removed him three marches from Portalegre, so that six days in all would be wasted in bringing him back to his original starting-point. It is curious that Wellington did not harden his heart to get rid of Erskine after this mishap: but though he wrote bitterly about his subordinate’s incapacity, he did not remove him. ‘Influence’ at home was apparently the key to his long endurance: it will be remembered that this was by no means the first of Erskine’s mistakes[378].

[377] To both on June 1. _Dispatches_, ix. p. 197. Erskine’s name is the blank to be filled up.

[378] See vol. iv. pp. 133 and 191.

The fall of the Almaraz forts, as might have been expected, was interpreted by Marmont and Soult each from his own point of view. The former, rightly as it turned out, wrote to Foy that he must be prepared to return to Leon at short notice, and that the Army of the Centre and Drouet must guard the valley of the Tagus on his departure[379]. Soult, on the other hand, having heard of Graham’s arrival at Badajoz and Hill’s return to Merida, argued that the allies were massing on the Guadiana for an advance into Andalusia. He made bitter complaints to Jourdan that he had violated the rules of military subordination by sending a letter to Drouet warning him that he might be called up to the Tagus. It was unheard of, he said, to communicate directly with a subordinate, who ought to be written to only through the channel of his immediate superior. He even threatened to resign the command of the Army of the South[380]--but when Joseph showed no signs of being terrified by this menace, no more was heard of it. The viceroyalty of Andalusia was not a thing to be lightly given up.

[379] Marmont to Foy, June 1.

[380] See Jourdan’s _Mémoires_, pp. 399-400.

It soon became evident to Wellington that the surprise of Almaraz was not to be resented by the enemy in any practical form. Foy was not reinforced, nor was Drouet brought up to the Tagus: it was clear that the French were too weak to take the offensive either in the North or the South, even under such provocation. They could not even rebuild the lost bridge: the transport from Madrid of a new pontoon train as a substitute for the lost boats was beyond King Joseph’s power. One or two boats were finally got to Almaraz--but nothing that could serve as a bridge. Nor were the lost magazines ever replaced.

It was at this same time that Wellington took in hand a scheme for facilitating his communications north and south, which was to have a high strategical importance. As long as Ciudad Rodrigo and Badajoz were in the enemy’s hands, the most eastern crossing of the Tagus practicable for the Anglo-Portuguese army was the boat-bridge of Villa Velha. But when these two fortresses were regained, it was possible to open up a line farther east, which had not been available for two years. Since Mayne blew up the ancient Roman bridge of Alcantara in June 1809[381], the Middle Tagus had been impassable for both sides. The allies had usually been in possession of both banks of the Tagus in this direction, but so intermittently that it had never been worth their while to restore the passage, which would have been lost to them whenever the French (as not unfrequently happened) extended their operations into the Coria-Zarza Mayor country on the north bank, or the Caçeres-Albuquerque country on the other. But when the enemy had lost both Badajoz and Rodrigo, and had no posts nearer to Alcantara than the Upper Tormes, the forts of Miravete, and Zalamea, when, moreover, he had adopted a distinctly defensive attitude for many months, Wellington thought it worth while to recover possession of a passage which would shorten the route from Estremadura to the frontiers of Leon by a hundred miles, and would therefore give him an advantage of six marches over the enemy in transferring troops from north to south. Whether Almaraz were again seized and reoccupied by the French mattered little: the restoration of Alcantara would be safe and profitable.

[381] See vol. ii. p. 444.

Accordingly, on May 24th, Colonel Sturgeon[382] and Major Todd of the Royal Staff Corps were sent to Alcantara to report on the practicability of restoring the broken arch, which, owing to the immense depth of the cañon of the Tagus, overhung the river by no less than 140 feet. It was intended that if the engineering problem should prove too hard, a flying bridge of rafts, boats, or pontoons should be established at the water level[383]. But Sturgeon and Todd did more than Wellington had expected, and succeeded in a very few days in establishing a sort of suspension-bridge of ropes between the two shattered piers of Trajan’s great structure. The system adopted was that of placing at each end of the broken roadway a very large and solid beam, clamped to the Roman stones, by being sunk in channels cut in them. These beams being made absolutely adhesive to the original work, served as solid bases from which a series of eighteen cables were stretched over the gap. Eight more beams, with notches cut in them to receive the cables, were laid at right angles across the parallel ropes, and lashed tight to them. The long cables were strained taut with winches: a network of rope yarn for a flooring was laid between the eight beams, and on this planks were placed, while a screen of tarpaulins supported on guide-ropes acted as parapets. The structure was sound enough to carry not only infantry and horses, but heavy artillery, yet could always be broken up in a short time if an enemy had ever appeared in the neighbourhood[384]. Several times it was rolled up, and then replaced.

[382] An officer probably better remembered by the general reader as the husband of Sarah Curran, Robert Emmet’s sometime fiancée, than as the executor of some of Wellington’s most important engineering works. He fell before Bayonne in 1814.

[383] See Wellington to Graham, 23rd and 24th May. _Dispatches_, ix. pp. 163-5.

[384] The best and most elaborate account of this is in Leith Hay, i. pp. 300-1.

When the completion of the repairs of Alcantara and the destruction of the French bridge of Almaraz are taken together, it must be concluded that Wellington’s work in May gave him an advantage over the French of at least ten or twelve marches in moving troops from north to south or vice versa. For the route from Ciudad Rodrigo to Merida, now open to him, had at least that superiority over the only itinerary of the enemy, which would be that by Avila, Talavera, Toledo, and the eastern passes of the Sierra Morena. Though the narrow bridge of Arzobispo on the Middle Tagus still remained in French hands, it did not lead on to any good road to Estremadura or Andalusia, but on to the defiles of the Mesa d’Ibor and the ravines of the Sierra de Guadalupe. No large force could march or feed in those solitudes.

All was now ready for the advance upon the Tormes, which Wellington had made up his mind to execute.

SECTION XXXIII: