CHAPTER IV

THE SALAMANCA FORTS. TEN DAYS OF MANŒUVRES, JUNE 20TH-30TH, 1812

Wellington’s conduct on reaching Salamanca was not that which might have been expected. When a general has, by a careful and well-arranged concentration, collected all his own troops into one solid mass, and then by a rapid advance has thrown himself into the midst of the scattered cantonments of an enemy who has no superiority to him in numbers, it is natural for him to press his pursuit vigorously. Far the most effective way of opening the campaign would have been to cut up the two divisions which Marmont had just led out of Salamanca, or at least to follow them so closely that they could be brought to action before all the outlying divisions had come in. This would certainly have been Napoleon’s method.

Wellington, however, wanted to fight a battle in one of his favourite defensive positions, and he thought that he had a means of compelling Marmont to attack him, by laying siege to the Salamanca forts. After Ciudad Rodrigo and Badajoz, no French marshal would like to see a third important post captured ‘under his nose.’ The British general judged that Marmont would fight him, in order to save his prestige and his garrison. And since he believed that Bonnet would not evacuate the Asturias, and that Caffarelli would send help late, if at all, he thought that he could count upon a superiority of numbers which rendered victory certain.

This seems to be the only rational way of explaining Wellington’s conduct on June 17th. On arriving in front of Salamanca his army made a majestic encircling movement, Picton’s column crossing the Tormes by the fords of El Canto below the city, Beresford’s and Graham’s by those of Santa Marta above it. The use of the unbroken town-bridge was made impossible by Marmont’s forts. The heads of the two columns met on the north side, and they then moved three miles on, and took up a long position below the heights of San Cristobal, which lie outside Salamanca on its northern and eastern front. These formed the chosen defensive fighting-ground which Wellington had already in his mind.

Only the 14th Light Dragoons and Clinton’s infantry of the 6th Division turned into Salamanca by the Toro gate, and acted as Wellington’s escort, while he was received by the municipality and made his arrangements for the attack on the forts, which, though they commanded the bridge, had no outlook on the spacious arcaded Plaza Mayor, where the reception took place. It was a lively scene. ‘We were received with shouts and _vivas_,’ writes an eye-witness. ‘The inhabitants were out of their senses at having got rid of the French, and nearly pulled Lord Wellington off his horse. The ladies were the most violent, many coming up to him and embracing him. He kept writing orders upon his sabretash, and was interrupted three or four times by them. What with the joy of the people, and the feeling accompanying troops about to attack a fortress, it was a half-hour of suspense and anxiety, and a scene of such interest as I never before witnessed[422].’

[422] Tomkinson’s _Diary_, p. 162.

Head-quarters were established that night in the city, and Clinton’s division invested the forts, which looked formidable enough to require close study before they were attacked. The rest of the army took up its bivouacs, with the cavalry out in front, and remained practically without movement on the ground now selected, for the next two days, till Marmont came to pay his expected visit.

The three Salamanca forts were built on high ground in the south-west corner of the city, which overlooks the long Roman bridge. To make them Marmont had destroyed a great part of the old University quarter of the place, levelling the majority of the colleges--for Salamanca, till 1808, had been a university of the English rather than the usual continental type, and had owned a score of such institutions. Nearly all the buildings on the slopes had been pulled down, leaving a wide open glacis round three massive convents, which had been transformed into places of strength. San Vincente occupied the crest of the knoll overlooking the river, and lay in the extreme angle of the old city wall, which enclosed it on two sides. The smaller strongholds, San Cayetano and La Merced, were separated from San Vincente by a narrow but steep ravine, and lay close together on another rising-ground of about the same height. The three formed a triangle with crossing fires, each to a large extent commanding the ground over which the others would have to be approached. The south and west sides of San Vincente and La Merced overhung precipitous slopes above the river, and were almost inaccessible. The north sides of San Cayetano and San Vincente were the only fronts that looked promising for attack, and in each elaborate preparations had been made in view of that fact. Marmont had originally intended to enclose all three forts and many buildings more--such as the Town Hospital, the convent of San Francisco, and the colleges of Ireland and Cuenca, in an outer _enceinte_, to serve as a large citadel which would contain several thousand men and all his magazines. But money and time had failed, and on the slopes below the forts, several convents and colleges, half pulled to pieces, were still standing, and offered cover for besiegers at a distance of some 250 yards from the works. The garrison consisted of six flank-companies from the 15th, 65th, 82nd, and 86th of the line and the 17th Léger, and of a company of artillery, under the _chef de bataillon_ Duchemin of the 65th. They made up a total of 800 men, and had thirty-six guns in position, of which, however, the greater part were only light field-pieces: two guns (commanding the bridge) were in La Merced, four in San Cayetano, the remaining thirty in San Vincente, the most formidable of the three.

Wellington had come prepared to besiege ‘three fortified convents,’ and had been sent a confused sketch of them drawn by an amateur’s hand[423]. They turned out much stronger than he had been led to expect, owing to the immense amount of hewn stone from the demolished colleges and other buildings that was available to build them up. The walls had been doubled in thickness, the windows stopped, and scarps and counterscarps with solid masonry had been thrown around them. The roofs of the two minor forts had been taken off, and the upper stories casemated, by massive oak beams with a thick coating of earth laid upon them. This surface was so strong that guns, protected by sandbag embrasures, had been mounted on it at some points. There was also an ample provision of palisades, made from strong oak and chestnut beams. Altogether it was clear that the works would require a systematic battering, and were not mere patched-up mediaeval monasteries, as had been expected.

[423] Jones, _Sieges_, i. p. 269.

It was, therefore, most vexatious to find that the very small battering-train which Wellington had brought with him from Ciudad Rodrigo was obviously insufficient for the task before it; there were no more than four iron 18-pounder guns, with only 100 rounds of shot each, at the front; though six 24-pound howitzers, from the train that had taken Badajoz, were on their way from Elvas to join, and were due on the 20th. It was not, however, howitzers so much as more heavy 18-or 24-pounders that were required for battering, and the lack of them at the moment was made all the more irksome by the known fact that there were plenty of both sorts at Rodrigo and Almeida, five or six marches away. The mistake was precisely the same that was to be made again at Burgos in the autumn--undervaluation of the means required to deal with works of third-class importance. Whether Wellington himself or his artillery and engineer advisers were primarily responsible is not clear[424].

[424] At any rate Dickson was not, as he was with the howitzers that were coming up from Elvas, and had not started from Rodrigo with the army.

The responsibility for the working out of the little siege with inadequate means fell on Lieut.-Colonel Burgoyne, as senior engineer (he had with him only two other officers of that corps and nine military artificers!), and Lieut.-Colonel May, R.A., who was in charge of the four 18-pounders. The latter borrowed three howitzers from field-batteries to supplement his miserable means, and afterwards two 6-pounder field-guns, which, of course, were only for annoying the garrison, not for battering.

It looked at first as if the only practicable scheme was to build a battery for the 18-pounders on the nearest available ground, 250 yards from San Vincente to the north, and lower down the knoll on which that fort stood. There was good cover from ruined buildings up to this distance from the French works. On the night of the occupation of Salamanca 400 workmen of the 6th Division commenced a battery on the selected spot and approaches leading to it from the cover in the ruins. The work done was not satisfactory: it was nearly full moon, the night was short, and the enemy (who knew well enough where the attack must begin), kept up a lively fire of artillery and musketry all night. Unfortunately the 6th Division workmen had no experience of sieges--they had never used pick or shovel before, and there were only two engineer officers and nine artificers to instruct them. ‘Great difficulty was found in keeping the men to work under the fire: the Portuguese in particular absolutely went on hands and knees, dragging their baskets along the ground[425].’ By daylight the projected line of the battery was only knee-high, and gave no cover, so that the men had to be withdrawn till dusk. An attempt had been made during the night to ascertain whether it were possible to creep forward to the ditch, and lay mines there, to blow in the counterscarp. But the party who tried to reach the ditch were detected by the barking of a dog, who alarmed the French out-picket, and the explorers had to retire with several men wounded.

[425] Burgoyne’s diary in his _Life_, i. p. 192.

Seeing that the fire of the garrison was so effective, the officers in charge of the siege asked for, and obtained from Wellington, three hundred marksmen to keep down the _tiraillade_. They were taken from the Light Brigade of the King’s German Legion, and spread among the ruins to fire at the embrasures and loopholes of the French. They also hoisted, with some difficulty, two field-guns on to the first floor of the convent of San Bernardo, which lies north-west of San Vincente, and kept up a lively discharge ‘out of the drawing-room window, so to speak. We fired for some hours at each other, during which time an unlucky shot went as completely through my captain’s (Eligé’s) heart as possible. But considering how near we were, I am much surprised that our loss was so slight--one killed and one wounded at my own gun[426].’ But the fire of the San Vincente artillery was by no means silenced.

[426] Letter of F. Monro, R.A., lent me by his representative. See _Fortnightly Review_ for July 1912.

On the night of June 18th-19th the working party of the 6th Division succeeded in finishing the battery which was to breach the main fort, and also commenced two smaller batteries, to right and left, in places among the ruins, one by the College of Cuenca, the other below San Bernardo[427]. On the morning of the 19th the four 18-pounders and three howitzers opened, and brought down the upper courses of the masonry of that part of San Vincente on which they were trained. But they could not move its lower part, or reach the counterscarp. Wherefore two howitzers were put into the second battery, near the College of Cuenca, which could command the counterscarp. The play of these guns proved insufficient, however, to shake it, and the garrison concentrated such a fire upon them, mainly from musketry at loopholes, that twenty gunners were killed or hurt while working the two howitzers.

[427] Nos. 2 and 3 in the map respectively.

Next morning Dickson’s six howitzers from Elvas came up, and served to replace those borrowed from the field companies, wherefore there was only an addition of three pieces net to the battering-train. Two of the 18-pounders were moved round to the battery (No. 2) which had been so hard hit on the preceding day: their fire proved much more effective than that of the howitzers, and brought down an angle of the upper wall of San Vincente and part of its roof, which fell on and crushed many of the French.

But on the 21st it was impossible to continue the battering, for the ignominious reason that there were hardly any more shot left to fire. Only sixty balls remained in store for the 18-pounders, and a little over one hundred for the howitzers[428]. The calculations of the besiegers had been so erroneous that they had used up their stock just as the critical moment had arrived. On the previous day Wellington, seeing what was coming, had sent a hurried message to Almeida for more shot and powder--but the convoy, though urged on with all possible speed, did not arrive at Salamanca till the 26th.

[428] Of course a few rounds more for the howitzers could have been borrowed from the field-batteries with the divisions. For the 18-pounders, the really important guns, there was no such resource for borrowing.

Meanwhile the general engagement for which Wellington had prepared himself seemed likely to come off. Marmont had all his army, save Bonnet alone, collected by the 19th, at Fuente Sauco. On the following day he came boldly forward and drove in the British cavalry vedettes. He showed three columns moving on a parallel front, which observers estimated at 18,000 foot and 2,000 horse--but there were more behind, still invisible. At four in the afternoon he was drawing so close that Wellington assumed his battle position. Five divisions and the two independent Portuguese brigades formed the fighting-line, from San Cristobal southward to Cabrerizos on the bank of the Tormes: the order was (from right to left) 1st-7th-4th-Light-3rd-Pack and Bradford. The reserve was composed of the 5th Division, of Hulse’s brigade of the 6th (of which the remainder was left to blockade the Salamanca forts), and of Carlos de España’s 3,000 Spaniards. Alten’s cavalry covered the British right, Ponsonby’s[429] the left, Bock’s and Le Marchant’s heavy squadrons were in reserve.

[429] Acting vice G. Anson, absent.

It looked at first as though Marmont intended to force on the battle that Wellington desired. Moving with great order and decision, his three columns deployed opposite the heights, and advanced to within a very moderate distance of them--not more than 800 yards at one point. They were extremely visible, as the whole country-side below the British position was a fine plain covered with ripening wheat. The only breaks in the surface were the infrequent villages--in this part of Spain they are all large and far apart--and a few dry watercourses, whose line could be detected winding amid the interminable cornfields. Warning to keep off the position was given to the French by long-range fire from several of the British batteries on salient points of the line. The enemy replied noisily and with many guns: Wellington’s officers judged that he was doing his best to make his approach audible to the garrison of the besieged forts.

At dusk the French occupied the village of Castellanos de Morisco, in front of the right centre of the heights, and then advanced a regiment to attack Morisco, which was absolutely at the foot of them, and had been occupied by Wellington as an advanced post. It was held by the 68th regiment from the 7th Division, a battalion which had come out from England in the preceding autumn, but had, by chance, never been engaged before. It made a fine defence, and beat off three attacks upon the village: but after dark Wellington called it back uphill to the line of the position, abandoning Morisco[430]. Apparently he was glad to see the French pressing in close, and looked for an attack upon his position next morning. Standing on the sky-line above Castellanos at dusk, with a map in his hand, he demonstrated to all the assembled generals commanding divisions the exact part which they were to play, till several French round-shot compelled him to shift his position a little farther back[431]. The whole army slept that night in order of battle, with strong pickets pushed down to the foot of the slopes.

[430] The 68th lost four officers and 46 men killed and wounded, and one officer taken prisoner. For a good account of the fight see the Memoirs of Green of the 68th, pp. 89-90.

[431] See Tomkinson’s _Diary_, p. 165.

There was, however, no attack at dawn. Marmont’s two rear divisions (those of Foy and Thomières) and a brigade of dragoons were not yet on the ground, and only got up in the course of the afternoon: hence he was naturally unwilling to move, as he had a certain knowledge that he was outnumbered. It would seem that Wellington had, that morning, an opportunity of crushing his enemy, which he must have regretted to have lost on many subsequent days of the campaign. Marmont’s position was one of very great risk: he had pushed in so close to the British heights, that he might have been attacked and brought to action in half an hour, and could not have got away without fighting. His position was visible from end to end--it had no flank protection, and its only strong points were the two villages of Morisco and Castellanos de Morisco on its left centre. Behind was an undulating sea of cornfields extending to the horizon. Wellington (after deducting the two missing brigades of the 6th Division) could have come down in a general charge from his heights, with 37,000 Anglo-Portuguese infantry, and 3,500 horse--not to speak of Carlos de España’s 3,000 Spaniards. Marmont had only five divisions of infantry (about 28,000 bayonets) on the ground at daybreak, and less than 2,000 horse. He was in a thoroughly dominated position, and it is hard to see what he could have done, had Wellington strengthened his left wing with all his cavalry and delivered a vigorous downhill assault on the unprotected French right. The opportunity for an attack was so favourable that Wellington’s staff discussed with curiosity the reasons that might be preventing it, and formed varying hypotheses to account for his holding back[432]. As a matter of fact, as his dispatch to Lord Liverpool explains[433], the British Commander-in-Chief was still hoping for a second Bussaco. He saw that Marmont was not going to attack till his rear had come up, but hoped that he might do so that afternoon or next morning, when he had all his men in hand. The daring way in which the Marshal continued to hold on to an untenable position, within cannon shot of his enemy’s line, seemed to argue an ultimate intention to bring on an action.

[432] Tomkinson’s _Diary_, p. 166.

[433] Wellington to Liverpool, Salamanca, June 25, in _Dispatches_, ix. p. 252.

Nor was Wellington very far out in his ideas: Marmont was in a state of indecision. When the missing 10,000 men came up he called a council of war--the regular resort of generals in a difficulty. We have concerning it only the evidence of Foy, who wrote as follows in his diary.

‘At dusk on the 21st there was a grand discussion, on the problem as to whether we should or should not give battle to the English. The Marshal seemed to have a desire to do so, but a feeble and hesitating desire. Remembering Vimeiro, Corunna, and Bussaco, I thought that it would be difficult to beat the English, our superiors in number, on such a compact position as that which they were occupying. I had not the first word: I allowed Maucune, Ferey[434], and La Martinière to express their views, before I let them see what I thought. Then Clausel having protested strongly against fighting, I supported his opinion. Because we had left a small garrison in the Salamanca forts, we were not bound to lose 6,000 killed and wounded, and risk the honour of the army, in order to deliver them. The troops were in good spirits, and that is excellent for the first assault: but here we should have a long tough struggle: I doubted whether we had breath enough to keep it up to the end. In short, I saw more chances of defeat than victory. I urged that we ought to keep close to the English, “contain” them, and wait for our reinforcements; this could be done by manœuvring along the left bank of the Tormes above and below Salamanca. Clausel and I set forth this policy from every aspect. The Marshal was displeased: he fancied that his generals were plotting to wreck his plan: he wanted to redeem the blunder which he saw that he had made in leaving a garrison in Salamanca: he dreads the Emperor and the public opinion of the army. He would have liked a battle, but he had not determination enough to persist in forcing it on[435].’

[434] The first two were great fire-eaters, and always urged

## action.

[435] Foy’s _Vie militaire_, ed. Girod de l’Ain, pp. 165-6.

It seems, therefore, certain that Wellington nearly obtained the defensive general action that he had desired and expected, and was only disappointed because Marmont was talked down by his two best divisional generals. If the Marshal had made his attack, it is clear that his disaster would have been on a far more complete and awful scale than the defeat which he was actually to endure on July 22. For he would have had behind him when repulsed (as he must have been) no friendly shelter of woods and hills, such as then saved the wrecks of his army, but a boundless rolling plain, in which routed troops would have been at the mercy of a cavalry which exceeded their own in the proportion of seven to five (or slightly more).

On the morning of the 22nd, the British general, who had now kept his army in position for thirty-six hours on end, began to guess that he was not to be attacked. Was it worth while to advance, since the enemy refused to do so? The conditions were by no means so favourable as at the dawn of the 21st, when Marmont had been short of 10,000 men. But the allied army still possessed a perceptible superiority in numbers, a stronger cavalry, and a dominating position, from which it would be easy to deliver a downhill attack under cover of their artillery.

Wellington, however, made no decisive movement: he threw up some _flèches_ to cover the batteries in front of the 1st and 7th Divisions, of which the latter was pushed a little nearer to the Tormes. He brought up the six heavy howitzers which had been used against the forts, and placed them on this same right wing of his position. Then he commenced a partial offensive movement, which was apparently designed to draw Marmont into a serious bickering, if he were ready to stand. The 7th Division began to make an advance towards Morisco: the skirmishers of the Light Brigade of the King’s German Legion moved down, and began to press in the pickets opposite them, their battalions supporting. Soon after the 51st and 68th, from the other brigade of the division, that of De Bernewitz, were ordered to storm a knoll immediately above Morisco, which formed the most advanced point of the enemy’s line. Wellington directed Graham to support them with the whole 1st and Light Divisions, if the enemy should bring up reinforcements and show fight. But nothing of the kind happened: the two battalions carried the knoll with a single vigorous rush, losing some 30 killed and wounded[436]. But the French made no attempt to recapture it, drew back their skirmishing line, and retired to the village, only 200 yards behind, where they stood firm, evidently expecting a general attack. It was not delivered: Wellington had been willing to draw Marmont into a fight, but was not intending to order an advance of the whole line, and to precipitate a general offensive battle.

[436] The 51st lost 3 killed and an officer and 20 men wounded: the 68th 2 killed and 6 wounded, the K.G.L. Light Battalions 3 killed and 3 officers and 17 men wounded. There are narratives of the combat in the Memoirs of Green of the 68th, and Major Rice and Private Wheeler of the 51st.

There was no more fighting that day, and next morning the whole French army had disappeared save some cavalry vedettes. These being pressed in by Alten’s hussars, it was discovered that Marmont had gone back six miles, to a line of heights behind the village of Aldea Rubia, and was there in a defensive position, with his left wing nearly touching the Tormes near the fords of Huerta. Wellington made no pursuit: only his cavalry reconnoitred the new French position. He kept his army on the San Cristobal heights, only moving down Anson’s brigade of the 4th Division to hold Castellanos, and Halkett’s of the 7th Division to hold Morisco. Hulse’s brigade of the 6th Division was sent back to Salamanca, as were also Dickson’s six howitzers, and Clinton was directed to press the siege of the forts--notwithstanding the unhappy fact that there was scarcely any ammunition left in the batteries.

Marmont had undoubtedly been let off easily by Wellington: yet he hardly realized it, so filled was his mind with the idea that his adversary would never take the offensive. His report to King Joseph shows a sublime ignorance of his late danger. As the document has never been published and is very short, it may be worth quoting.

‘Having concentrated the greater part of this army on the evening of the 19th, I marched on Salamanca the same day. I seized some outlying posts of the enemy, and my army bivouacked within half cannon-shot of the English. Their army was very well posted, and I did not think it right to attack yesterday (June 21) without making a reconnaissance of it. The result of my observations has convinced me that as long as my own numbers are not _at least equal_ to theirs, I must temporize, and gain time for the arrival of the troops from the Army of the North, which General Caffarelli has promised me. If they arrive I shall be strong enough to take an enterprising course. Till then I shall manœuvre round Salamanca, so as to try to get the enemy to divide his army, or to move it out of its position, which will be to my advantage. The Salamanca forts are making an honourable defence. Since we came up the enemy has ceased to attack them, so that I have gained time, and can put off a general action for some days if I think proper[437].’

[437] Marmont to Joseph, night of the 22nd June, from bivouac before San Cristobal. Intercepted dispatch in the Scovell Papers.

Marmont’s plan for ‘manœuvring around Salamanca’ proved (as we shall see) quite ineffective, and ended within a few days in a definite retreat, when he found that the succours promised by Caffarelli were not about to appear.

Meanwhile the siege of the Salamanca forts had recommenced, on the 23rd, under the depressing conditions that the artillery had only 60 rounds (15 apiece!) for the four heavy 18-pounders, which were their effective weapons, and 160 for the six howitzers, which had hitherto proved almost useless. The two light field-guns (6-pounders) were also replaced on the first floor of San Bernardo to shell the enemy’s loopholes--they were no good at all for battering. This time the besiegers placed one of their heavy guns in the right flanking battery near San Bernardo, to get an oblique enfilading fire against the gorge of the San Cayetano fort. The new idea was to leave San Vincente alone, as too hard a nut to crack with the small supply of shot available, and to batter the lesser fort from flank and rear with the few rounds remaining. The entire stock, together with a hundred rounds of shell, was used up by the afternoon, when no practicable breach had been made, though the palisades of San Cayetano had been battered down, and its parapet much injured. Nevertheless Wellington ordered an attempt to storm (or rather to escalade) the minor fort at 10 p.m. on the same evening. It was to be carried out by the six light companies of Bowes’s and Hulse’s brigades of the 6th Division, a force of between 300 and 400 men. ‘The undertaking was difficult, and the men seemed to feel it,’ observes the official historian of the Peninsular sieges[438]. The major of one of the regiments engaged remarks, ‘the result was precisely such as most of the officers anticipated--a failure attended with severe loss of life.’ The storming-column, starting from the ruins near the left flanking battery, had to charge for the gorge of San Cayetano, not only under the fire of that work, but with musketry and artillery from San Vincente taking them in the rear. The casualties from the first moment were very heavy--many men never got near the objective, and only two ladders out of twenty were planted against the fort[439]. No one tried to ascend them--the project being obviously useless, and the stormers ran back under cover after having lost six officers and 120 men, just a third of their numbers[440]. Among the killed was General Bowes, commanding the second brigade of the division, who had insisted on going forward with his light companies--though this was evidently not brigadier’s work. Apparently he thought that his personal influence might enable his men to accomplish the impossible. He was hit slightly as the column started, but bound up his wound, and went forward a second time, only to be killed at the very foot of the ladders, just as his men broke and retired.

[438] Jones, i. p. 281.

[439] The regimental history of the 53rd says that the ladders were so badly made, of green wood, that many of them came to pieces in the hands of their carriers long before they got near the fort.

[440] The loss has got exaggerated in many reports, because the casualties in the 7th Division at Morisco on the preceding day are added to the total.

This, as all engaged in it agreed, was a very unjustifiable enterprise; the escalade was impracticable so long as San Vincente was intact, and able to cover the gorge of San Cayetano with an effective fire from the rear. The siege now had a second period of lethargy, all the shot having been used up. It was only on the morning of the 26th, three days later, that the convoy from Almeida, ordered up on the 20th by Wellington, arrived with 1,000 rounds carried by mules, and enabled the battering to begin once more.

Meanwhile Marmont had been making persistent but ineffective diversions against Wellington. The advantage of the position to which he had withdrawn was that it commanded the great bend, or elbow, of the Tormes, where (at the ford of Huerta) that river turns its general course from northward to westward. Troops sent across the river here could threaten Salamanca from the south, and, if in sufficient strength, might force Wellington to evacuate part of the San Cristobal position, in order to provide a containing force to prevent them from communicating with and relieving the besieged forts. The Marshal’s own statement of his intention[441] was that he hoped, by manœuvring, to get Wellington either to divide his army or to leave his strong ground, or both. He aimed, no doubt, at obtaining the opportunity for a successful action with some isolated part of Wellington’s force, but was still too much convinced of the danger of fighting a general action to be ready to risk much. Moreover he was expecting, from day to day, the 8,000 men of the Army of the North whom Caffarelli had promised him: and it would be reckless to give battle before they arrived--if only they were really coming.

[441] See above, p. 370.

Wellington could see, by his own eyes no less than by the map, for he rode along Marmont’s new front on the 23rd, that the French position gave good possibilities for a passage of the elbow of the Tormes at Huerta: wherefore he detached Bock’s brigade of German Dragoons to the south of the river, with orders to watch the roads debouching from the fords, and to act as a detaining force if any hostile cavalry crossed them. He also threw forward Alten’s hussars to Aldea Lengua, a village and ford half-way between Cabrerizos and Aldea Rubia, with the object of keeping a similar close watch on any attempt of Marmont’s to move north of the river. One brigade of the Light Division came forward to support Alten--the other was écheloned a little back, on hills above Aldea Lengua.

On the late evening of the 23rd Marmont sent a squadron or two across the Huerta fords, which turned back after running into Bock’s vedettes. This was merely an exploring party to test the practicability of the passage; but next morning, in a heavy fog, skirmishing fire and occasional reports of cannon told Wellington that some more important detachment was across the Tormes, and engaged with the Germans. The British head-quarters staff rode to the hill above Aldea Lengua, which commands a wide view over the south bank, and, when the morning vapours rolled up at 7 o’clock, saw Bock retiring across the rolling plain in very good order, pressed by a heavy force of all arms--two divisions of infantry headed by a light cavalry brigade with a horse artillery battery, which was doing some harm to the two dragoon regiments as they retired in alternate échelons across the slopes.

Fortunately there was excellent defensive fighting-ground south of the Tormes, in prolongation of the San Cristobal position north of it. The ravine and brook[442] called the Ribera de Pelagarcia with wooded heights above them, run in front of Santa Marta and its ford, for some miles southward from the Tormes. There was a similar line of high ground facing it, with the villages of Pelabravo and Calvarisa de Ariba on its top, which the French might occupy, but on passing down from them they would run against a formidable position. Along these hills, indeed, Wellington’s first line of defence was to be formed a month later, on the day of the battle of Salamanca. On seeing Bock’s careful retreat in progress, the Commander-in-Chief ordered Graham to cross the Tormes at Santa Marta with the 1st and 7th Divisions, and to occupy the ground in front of him. This was a short move, and easily accomplished while the French detachment was pushing the German dragoons slowly backward. The 4th and 5th Divisions moved down to the north bank of the Tormes, ready to follow if Marmont should support his advanced guard, by sending more men over the Huerta fords. Le Marchant’s heavy brigade crossed the river with a horse artillery battery, and went to reinforce Bock, whom the French could now only push in by bringing forward infantry. Their advance continued as far as the village of Calvarisa de Abaxo, and a little beyond, where the whole 9,000 or 10,000 men deployed, as if intending to attack Graham. But just as observers on the Aldea Lengua heights were beginning to think that serious fighting was probable[443], the whole fell back into column of march, and, retiring to Huerta covered by their _chasseurs_, recrossed the river.

[442] I find the name Ribera de Pelagarcia only in the more modern Spanish maps: contemporary plans do not give it.

[443] Tomkinson, p. 170: ‘Just before they began to retire, I thought that their advance looked serious. Our position was good, and if they had fought with what had crossed, our force would have been the greater.’

The state of affairs at nightfall was just what it had been at dawn. Graham and Le Marchant went back to their old ground north of the river, and south of it cavalry alone was left--this time Alten’s brigade, for Bock’s had had a heavy day, and needed rest. So ended a spectacular but almost bloodless manœuvre--the German dragoons lost three killed and two wounded: the French light horse probably no more.

In a dispatch written the same night Marmont frankly owns that he was foiled by Wellington’s counter-move. This hitherto unpublished document is worth quoting. It is addressed to General Caffarelli, and runs as follows[444]. ‘The movement which I have made toward Salamanca has caused the enemy to suspend his attack on the forts of that town. [An error, as it was not the movement but the lack of ammunition which stopped the bombardment.] This consideration, and the way in which I found him posted to keep me off, and not least your assurance that your powerful reinforcements would reach me very soon, have determined me to suspend the attack which I was about to deliver against him. I stop here with the object of gaining time, and in the expectation of your arrival.’ From this it is clear that if Graham had not been found so well posted, in a position where he could readily be reinforced from San Cristobal, Marmont would have followed up his advanced guard with the rest of his army, and have struck at Salamanca from the South. But finding the ground on the left bank of the river just as unfavourable to him as that on the north, he gave up the game and retired. He risked a serious check, for Wellington might have ordered Graham to follow and attack the retreating divisions, who would have had great difficulty in recrossing the Tormes without loss, if they had been pursued and attacked while jammed at the fords. But Wellington was still in his defensive mood, and took no risks, contented to have foiled most effectively his enemy’s manœuvre.

[444] This is one of the many cipher dispatches in the Scovell Papers, which I have found so illuminating in a period when Marmont’s writings, printed or in the French archives, are very few.

On the 25th Marmont remained stationary, waiting for further advices from Caffarelli, which failed to come to hand. Nor did Wellington make any move, save that of sending orders that the siege of the forts was to be pressed as early and as vigorously as possible. The guns were back in their batteries, waiting for the ammunition which was yet to appear. All that could be done without shot was to push forward a trench along the bottom of the ravine between San Vincente and the other two forts, to cut off communication between them. The French fired fiercely at the workers, where they could look down into the ravine, and killed some of them. But there was much ‘dead ground’ which could not be reached from any point in the forts, and by dawn on the 26th the trench was far advanced, and a picket was lodged safely in it, close under the gorge of San Cayetano.

On the morning of the 26th the convoy of powder and shot from Almeida reached the front, and at three in the afternoon the besiegers recommenced their fire. This time no guns were placed in the original battery opposite the north front of San Vincente; the four 18-pounders all went into the right flank attack, and were concentrated on the gorge of San Cayetano. Four of the howitzers were placed in the left flank battery, near the College of Cuenca, and directed to fire red-hot shot into the roof and upper story of San Vincente. The field-guns in San Bernardo, aided by one howitzer, took up their old work of trying to keep down the fire of the forts.

The battering in of the gorge of San Cayetano made considerable progress, but the most effective work was that of the red-hot shot, which before night had set the tower of San Vincente and several points of its roof in flames. By heroic exertions the garrison succeeded in extinguishing them, but the besiegers’ fire was kept up all night, and from time to time new conflagrations burst out. The governor afterwards informed the British engineers that eighteen separate outbreaks were kept down within the twenty-four hours before his surrender[445]. The fort was very inflammable, owing to the immense amount of timber that had been used for casemating, traverses, barricades, and parapets, inside its walls. Still it was holding out at daybreak, though the garrison was nearly exhausted: the governor signalled to Marmont that he could not resist for more than three days--a sad over-estimate of his power, as was to be shown in a few hours. As a subsidiary aid to the work of the guns two mines were commenced, one from the ravine, destined to burrow under San Cayetano, the other from the cliff by the river, intended to reach La Merced. But neither was fated to be used, other means sufficing.

[445] Jones, _Sieges of the Peninsula_, i. p. 285.

After four hours’ pounding on the morning of the 27th, the gorge of San Cayetano had been battered into a real and very practicable breach, while a new fire had broken out in San Vincente, larger than any one which had preceded it. It reached the main store of gabions and planks within the fort, and threatened the powder magazine. The garrison were evidently flinching from their guns, as the counter-fire from the place, hitherto very lively, began to flag, and the whole building was wrapped in smoke.

Thereupon Wellington ordered San Cayetano to be stormed for the second time. The column charged with the operation crept forward along the trench at the bottom of the ravine, fairly well covered till it had reached the spot immediately below the gorge of the fort. Just as the forlorn hope was about to start out of the trench, a white flag was shown from the breach. The captain commanding in San Cayetano asked for two hours’ truce, to enable him to communicate with his chief in San Vincente, promising to surrender at the end of that time. Wellington offered him five minutes to march out, if he wished to preserve his garrison’s lives and baggage. As the Frenchmen continued to haggle and argue, he was told to take down his white flag, as the assault was about to be delivered. When the stormers ran in, San Cayetano made practically no defence, though a few shots were fired, which caused six casualties in the assaulting column: the greater part of the garrison threw down their muskets and made no resistance.

[Illustration: SALAMANCA FORTS]

At the same moment the white flag went up on San Vincente also: here the conflagration was now burning up so fiercely that the French had been able to spare no attention for the storming-party that captured San Cayetano. The governor, Duchemin, asked for three hours’ suspension of arms, and made a proposal of terms of surrender. Wellington, here as at the smaller fort, refused to grant time, as he thought that the fire would be subdued and the defence prolonged, if he allowed hours to be wasted in negotiations. He sent in the same ultimatum as at San Cayetano--five minutes for the garrison to march out, and they should have all the ‘honours of war’ and their baggage intact. Duchemin, like his subordinate, returned a dilatory message, but while his white flag was still flying, the 9th Caçadores pushed up out of the ravine and entered the battery on the east side of the work. They were not fired on, no one in San Vincente being prepared to continue the defence, and the French standard came down without further resistance.

Not quite 600 unwounded men of the garrison were captured. They had lost just 200 during the siege, including 14 officers[446]. The casualties among the British were, as might have been expected, much heavier, largely owing to the unjustifiable assault of June 23rd. They amounted to 5 officers and 94 men killed, and 29 officers and 302 men wounded. A considerable store of clothing, much powder, and 36 guns of all sorts were found in the three forts. The powder was made over to Carlos de España, one of whose officers, having moved it into the town on the 7th July, contrived to explode many barrels, which killed several soldiers and twenty citizens, besides wrecking some houses[447]. The three forts were destroyed with care, when they had been stripped of all their contents.

[446] The total given by the governor to Warre of Beresford’s staff (see his _Letters_, ed. Dr. Warre, p. 270) were 3 officers and 40 men killed, 11 officers and 140 men wounded. Martinien’s lists show 12 officers hit, 5 in the 65th, 2 each in the 15th and 17th Léger, 1 each in 86th, artillery, and engineers. But these admirable lists are not quite complete.

[447] This is said to have been the result of the escort’s smoking round the store!

The fall of the Salamanca forts happened just in time to prevent Marmont from committing himself to a serious offensive operation for their succour. It will be remembered that, on June 24th, he had used the plea that Caffarelli’s troops must be with him, ere many days had passed, as a justification for not pushing on to attack the British divisions in front of Santa Marta. And this expectation was reasonable, in view of that general’s last dispatch from Vittoria of June 14th[448], which spoke of his appearance with 8,000 men as certain and imminent. On the 26th, however, the Marshal received another letter from the Army of the North, couched in a very different tone, which upset all his plans. Caffarelli, writing on the 20th, reported the sudden arrival on the Biscay coast of Sir Home Popham’s fleet, whose strength he much exaggerated. In co-operation with the English, Longa, Renovales, and Porlier had all come down from their mountains, and Bilbao was in danger from their unexpected and simultaneous appearance. It would probably be necessary to march to drive off the ‘7th Army’ and the British expedition without delay. At any rate the transference of any infantry towards the Douro for the succour of the Army of Portugal had become impossible for the moment. The brigade of light cavalry and the guns might still be sent, but the infantry division had become indispensable elsewhere. ‘I am sorry,’ ended Caffarelli, ‘but I could not have foreseen this development, and when I spoke of marching towards you I was far from suspecting that it could arise.’

[448] Printed in Marmont’s _Mémoires_, iv. p. 410.

This epistle changed the whole aspect of affairs: if the infantry division from Vittoria had been diverted into Biscay for an indefinite period, and if even the cavalry and guns (an insignificant force so far as numbers went, yet useful to an army short of horse) had not even started on June 20th, it was clear that not a single man would be available from the North for many days. Meanwhile the governor of the forts signalled at dawn on the 27th that seventy-two hours was the limit of his power of resistance. Thereupon Marmont came to the desperate resolve to attempt the relief of San Vincente with no more than his own 40,000 men. He tells us that he intended to move by the south side of the Tormes, crossing not at Huerta (as on the 24th) but at Alba de Tormes, seven miles higher up, where he had a small garrison in the old castle, which protected the bridge. This move would have brought him precisely on to the ground where he ultimately fought the disastrous battle of July 22nd. He would have met Wellington with 7,000 men less than he brought to the actual battle that was yet to come, while the Anglo-Portuguese army was practically the same in July as it was in June[449]. The result could not have been doubtful--and Marmont knew that he was taking a serious risk. But he did not fathom its full danger, since he was filled with an unjustifiable confidence in his adversary’s aversion to battle, and thought that he might be manœuvred and bullied out of his position, by a move against his communications[450]. He would have found out his error in front of the Arapiles on June 29th if he had persevered.

[449] If Marmont had marched for Alba de Tormes on the 28th, as he intended to do, Wellington would have had the 6th Division in hand, as well as the rest of his troops, for a battle on the 29th: for the forts fell early on the 27th June.

[450] See his explanation of his intentions in _Mémoires_, iv. pp. 219-20.

But he did not persevere: in the morning of June 27 the firing at Salamanca ceased, and a few hours later it was known that the forts had fallen. Having now no longer any reason for taking risks, the Marshal changed his whole plan, and resolved to remove himself in haste from Wellington’s neighbourhood, and to take up a defensive position till he should receive reinforcements. Two courses were open to him--the first was to retire due eastward toward Arevalo, and put himself in communication, by Avila and Segovia, with the Army of the Centre and Madrid. The second was to retire north-eastward toward Valladolid, and to go behind the strong defensive line of the Douro. Taking this line the Marshal would sacrifice his touch with Madrid and the South, but would be certain of picking up the reinforcement under Bonnet which he was expecting from the Asturias, and would also be able to receive with security whatever succour Caffarelli might send--even if it turned out to be no more than cavalry and guns.

This alternative he chose, probably with wisdom, for in a position on the Douro he threatened Wellington’s flank if he should advance farther eastward, and protected the central parts of the kingdom of Leon from being overrun by the Army of Galicia and Silveira’s Portuguese, who would have had no containing force whatever in front of them if he had kept south of the Douro and linked himself with Madrid. His retreat, commenced before daybreak on the 28th, took him behind the Guarena river that night: on the 29th he crossed the Trabancos, and rested for a day after two forced marches. On the 30th he passed the Zapardiel, and reached Rueda, close to the Douro, on the following morning. From thence he wrote to King Joseph a dispatch which explains sufficiently well all his designs: it is all the more valuable because its details do not entirely bear out the version of his plans which he gives in his _Mémoires_.

‘The Salamanca forts,’ he said, ‘having surrendered, there was no reason for lingering on the Tormes; it was better to fall back on his reinforcements. If he had not done so, he would have been himself attacked, for Wellington was preparing to strike, and pursued promptly. He had detached one division [Foy] towards Toro and the Lower Douro to keep off Silveira, who had passed that river at Zamora. Moreover the Galicians had blockaded Astorga, and crossed the Orbigo. He felt that he could defend the line of the Douro with confidence, being aided by the line of fortified posts along it--Zamora, Toro, and Tordesillas. But to take the offensive against Wellington he must have 1,500 more cavalry and 7,000 more infantry than he actually had in hand--since the Anglo-Portuguese army was nearly 50,000 strong, and included 5,000 English horse.’ This reinforcement was precisely what Caffarelli had promised, but by the 28th not one man of the Army of the North had reached Valladolid. ‘If the general can trump up some valid excuse for not sending me the infantry, there is none for keeping back the cavalry--which is useless among his mountains--or the artillery, which lies idle at Burgos.’ Would it not be possible for the Army of the Centre to lend the Army of Portugal Treillard’s division of dragoons from the valley of the Tagus, since Caffarelli sent nothing? If only the necessary reinforcements, 1,500 horse and 7,000 foot, came to hand, the Army of Portugal could take the offensive with a certainty of success[451]; in eight days Wellington’s designs could be foiled, and Salamanca could be recovered. But without that succour the Marshal must keep to the defensive behind the Douro--’I can combat the course of events, but cannot master them[452].’

[451] In this dispatch and that of July 6 following, Marmont seems to understate his own force at the moment, saying that he can dispose of only 30,000 infantry, and 2,000 cavalry or a trifle over. Allowing for the artillery, engineers and sappers, gendarmerie and train, which the monthly returns show, this would give an army of some 35,000 or 36,000 in all. But the returns (see Appendix) indicate a higher figure for the infantry; after all deductions for detachments, garrisons, and sick have been made, it looks as if there must have been 33,000 or even 34,000 available. Generals with a ‘point to prove’ are always a little easy with their figures.

[452] This is again one of the Scovell intercepted cipher-dispatches, captured and brought to Wellington a day or two after it was written. It was a duplicate, and presumably the other copy reached Madrid.

This interesting dispatch explains all that followed. Marmont was prepared to fight whenever he could show a rough numerical equality with Wellington’s army. He obtained it a few days later, by the arrival of Bonnet with his 6,500 infantry, and the increase of his cavalry by 800 or 900 sabres owing to measures hereafter to be described. On July 15th he had got together nearly 50,000 men of all arms, and at once took the offensive, according to the programme which he had laid down. It is, therefore, unfair to him to say that he declared himself unable to fight till he should have got reinforcements either from Caffarelli or from Madrid, and then (in despite of his declaration) attacked Wellington without having received them. He may have been presumptuous in acting as he did, but at least he gave his Commander-in-Chief fair notice, a fortnight beforehand, as to his intentions. It was the misfortune of the French that some of their dispatches miscarried, owing to the activity of the guerrilleros, while others came to hand very late. Marmont and King Joseph--as we shall see--were very imperfectly and intermittently informed as to each other’s doings. But the Marshal cannot reasonably be accused of betraying or deluding the King out of jealousy or blind ambition. When he had collected a force very nearly equal to Wellington’s in numbers, and far superior in national homogeneity, he cannot be blamed over-much for attacking a foe whose fighting spirit and initiative he much undervalued. That his conception of Wellington’s character and capacity was hopelessly wrong cannot be denied: the estimate was to prove his ruin. But it had not been formed without much observation and experiment: after what he had seen on the Caya, and at Aldea da Ponte, and recently on the heights of San Cristobal, he thought he could take liberties with his opponent. He was to be undeceived in a very rude fashion before July was out.

SECTION XXXIII: