CHAPTER II

THE CONSEQUENCES OF THE FALL OF CIUDAD RODRIGO

The extraordinary speed with which Wellington had in twelve days reduced Ciudad Rodrigo, a fortress that had held out for twenty-four days of open trenches when besieged by Ney in 1810, surprised the captor himself, who had reckoned on taking no shorter time in its leaguer than had the French. But it absolutely appalled his two adversaries, Marmont and Dorsenne, whose whole scheme of operations had rested on the idea that they could count on some three weeks or more for preparation, when the news that the place was invested got to their hands.

Thiébault, the governor of Salamanca, had been warning both the commander of the Army of the North and the commander of the Army of Portugal for some weeks that Wellington might move at any moment[201]. But his reports to the effect that the British were making gabions and fascines, preparing a bridge over the Agueda, and bringing up siege-guns to Almeida, made little or no impression on his superiors, because they had come to the conclusion that it was unlikely that Wellington would undertake a siege at midwinter. His preparations, they thought, were probably intended to force his enemies to concentrate, at a time when roads were bad and food unprocurable: ‘ils n’ont d’autre but que de nous faire faire de faux mouvements,’ said one of Marmont’s aides-de-camp. It was only in the spring that the allied army would become really enterprising and dangerous.

[201] See Thiébault’s _Mémoires_, iv. pp. 551-2. Extracts from two of his letters are printed in Marmont’s _Mémoires_, iv. pp. 280-1, and bear out all that he says in his own book.

Astonishing as it may appear, though Wellington’s troops started on January 2nd, and though Rodrigo was invested and the Redout Renaud stormed on January 8th, the definitive news that the siege had actually begun only reached Salamanca on January 13th. No better proof could be given of the precarious nature of the French hold on the kingdom of Leon. The fact was that the guerrilleros of Julian Sanchez so obsessed all the roads from Salamanca to Rodrigo, that no messenger could pass without a very large escort. Barrié only got the news that he was attacked to Thiébault by entrusting it to a Spanish emissary, who carried his note in disguise, and by a long détour. Marmont and Dorsenne only received it on the 14th: King Joseph at Madrid only on the 25th. On the 13th Marmont was in such a state of blindness as to the actual situation that he was writing to Berthier that ‘si l’armée anglaise passait l’Agueda j’attendrais sur la Tormès la division du Tage et les troupes que le Général Dorsenne pourrait m’amener, _mais sans doute ce cas n’arrivera pas_. Ciudad Rodrigo sera approvisionné jusqu’à la récolte, et à moins d’un siège il ne doit pas être l’objet d’aucune sollicitude[202].’ Wellington, when this was written, had already passed troops over the Agueda some ten days back, and had been beleaguering Ciudad Rodrigo for five. Yet Marmont was dating from Valladolid, which was not much over 100 miles from the hard-pressed fortress. Truly, thanks to the guerrilleros, the ‘fog of war’ was lying heavily round the Marshal.

[202] Marmont to Berthier, Valladolid, Jan. 13, 1812.

Owing to a circumstance of which Wellington could have no knowledge, the moment which he chose for his advance was even more propitious than he guessed. He knew of the march of Montbrun towards Valencia, and had made it the determining factor in his operations. But he was not, and could not be, aware of another fact of high importance. On December 29th Marmont, then at Talavera, had received a dispatch from Paris, dated on the 13th of the same month, informing him that the Emperor had resolved on making a sweeping change with regard to the respective duties and stations of the Armies of the North and of Portugal. Hitherto Dorsenne had been in charge of the whole kingdom of Leon: the troops stationed in it belonged to his army, and on him depended the garrisons of Ciudad Rodrigo, Astorga, and its other fortresses. He was, therefore, responsible for the keeping back of Wellington from all the ground north of the Sierra de Gata. Marmont, with his Army of Portugal, had to ‘contain’ the Anglo-Portuguese army south of that range, and had charge of the valley of the Tagus--northern Estremadura and those parts of New Castile which had been taken away from King Joseph’s direct control. From this central position the Duke of Ragusa had hitherto been supposed to be able to stretch out a hand to Dorsenne, in case of Wellington’s making a move in the valley of the Douro, to Soult in case of his showing himself opposite Badajoz. This indeed Marmont had done: he had brought up his army to Dorsenne’s aid in September, at the time of El Bodon and Aldea da Ponte: he had carried it down to the Guadiana and assisted Soult to relieve Badajoz in June.

Berthier’s dispatch[203], received on December 29th--it had taken sixteen days to reach its destination--informed Marmont that the Emperor had resolved to place the task of ‘containing’ Wellington, when he should operate north of the Tagus, in the hands of one instead of two commanders-in-chief. ‘Considering the importance of placing the command on the whole frontier of Portugal under a single general, His Majesty has decided that the provinces of Avila, Salamanca, Plasencia, Ciudad Rodrigo, the kingdom of Leon, Palencia, and the Asturias, shall belong to the Army of Portugal.’ Along with them were to be handed over to Marmont Souham’s division, then lying in the direction of Zamora, Benavente, and La Baneza, and Bonnet’s division, then in the Asturias--whose central parts (as it will be remembered[204]) that general had reconquered in November 1811. The district of the Army of the North was for the future to be limited to the eastern parts of Old Castile, Santander, Biscay, and Navarre. The real cause of this change, though Berthier’s dispatch lays no stress upon it, was the order recently sent to Dorsenne, which bade him return to France the two strong divisions of the Imperial Guard, which had hitherto formed the most important and effective section of the Army of the North. They were wanted for the probable Russian war, and without them Napoleon rightly judged that Dorsenne would be too weak to ‘contain’ Wellington, hold down all Leon, and observe the Galicians, in addition to hunting Mina and curbing the incursions of Longa and Porlier. Wherefore he resolved to confine the activity of the Army of the North to the lands east and north of Burgos, where its main task would be the crushing of Mina and his compatriots. Marmont should take upon his shoulders the entire responsibility for holding back the Anglo-Portuguese.

[203] Printed in full in Marmont’s _Mémoires_, iv. pp. 271-6.

[204] See vol. iv. p. 586.

But, by the Emperor’s orders, the Army of Portugal, though now charged with a much heavier task than before, was not to get any appreciable increase in numbers. It is true that Marmont was to take over the divisions of Souham and Bonnet, along with the regions that they were occupying. These were strong units, and would have increased his total strength by 16,000 men. But at the same time he was told that Thiébault’s division[205], the other force in the kingdom of Leon, was not to be given him, but to be withdrawn eastward and to remain under Dorsenne. With it were to go other details belonging to the Army of the North, employed in garrison duty in the valley of the Douro, such as the Swiss battalions long garrisoning the city of Leon, Benavente, and Valladolid[206]. Now it was clear that if these garrisons were withdrawn, Marmont would have to find other troops from his own divisions to replace them. Moreover, he was in addition instructed that Bonnet’s division, though now to be regarded as under his command, was not on any excuse to be moved out of the Asturias. ‘It is indispensable that he should remain there, because in that position he menaces Galicia, and keeps down the people of the mountains. You would have to use more troops to guard all the edge of the plain from Leon to St. Sebastian than are required for the Asturias. It is demonstrable in theory, and clearly proved by experience, that of all operations the most important is the occupation of the Asturias, which makes the right of the army rest upon the sea, and continually threatens Galicia.’

[205] 34th Léger, 113th Line, 4th Vistula, Neuchâtel.

[206] Also two cavalry regiments, the 1st Hussars and 31st Chasseurs.

If, therefore, Marmont was forbidden to use Bonnet, and had to replace all the existing garrisons of Leon (including that of Ciudad Rodrigo, as he was specially informed) by troops drawn from his own force, he was given a vast increase of territory to watch, but no appreciable increase of numbers to hold it--no more in fact than the difference between the strength of Souham’s division (placed on the side of gain) and that of the new garrisons (placed on the side of loss). The net profit would be no more than 3,000 or 4,000 men at the most.

In addition the Marshal was restricted further as to the way in which he was to dispose of his army. He was told to leave one division (or, if he chose, two) in the valley of the Tagus, about Plasencia and Almaraz, for the purpose of keeping up his communication with Madrid and Andalusia. The rest of his army was to be moved across the Sierra de Gata into the valley of the Douro, and its head-quarters were to be placed at Valladolid, or if possible at Salamanca. Therefore, if Wellington advanced, only four and a half, or five and a half, divisions out of the eight now comprising the Army of Portugal, could be concentrated against him with promptitude: Bonnet and the troops left in the Tagus valley would be long in arriving. So would the nearest divisions of the Army of the North, of which the most westerly would be as far off as Burgos, the rest still farther towards the Pyrenees. Till he had received some of these outlying succours, Marmont would be too weak to resist Wellington. Five divisions (say 30,000 men) could not keep the Anglo-Portuguese contained--though eight might very possibly suffice.

But on December 29, 1811, Marmont had not eight divisions at his disposition. The Emperor’s misguided order for the Valencian expedition was in progress of being executed, and it was precisely on that same day that Montbrun with two divisions of foot and one of horse was marching off eastward from La Mancha, in an excentric direction, which took him to the shore of the Mediterranean.

Marmont’s available force, after this march began, was as follows:

(1) Souham’s division at La Baneza, Benavente, and Zamora, watching Abadia’s Army of Galicia. This unit had yet to be informed that it had become part of the Army of Portugal.

(2-3) Brennier’s and Maucune’s divisions at Almaraz and Talavera in the valley of the Tagus.

(4) Clausel’s division at Avila.

(5) Ferey’s division in La Mancha, keeping up communication with Montbrun’s expeditionary column.

The other three divisions of the Army of Portugal, as now constituted, those of Bonnet in the Asturias, and of Foy and Sarrut in march for Valencia, were hopelessly out of reach.

Being directed, in very clear and decisive terms, to transfer himself in person to Valladolid or Salamanca, and to move the bulk of his troops thither from the valley of the Tagus, the Marshal had to obey. He directed Brennier’s division alone to remain behind at Almaraz and Talavera. Maucune and Clausel, with Ferey presently to follow, began a toilsome march across the mountains to Leon. They had to abandon the magazines (such as they were) which had been collected for their subsistence in winter-quarters, and to march across bad roads, in the most inclement month of the year, through an unpeopled country, for cantonments where no stores were ready for them.

While Marmont was marching up in the early days of January to occupy his newly-designated positions, Dorsenne was employed in withdrawing his troops eastward, away from the neighbourhood of Wellington, towards the province of Burgos. He himself stopped behind at Valladolid, to see Marmont and hand over in person the charge of the districts which he was ordered to evacuate. His view of the situation at the moment may be judged by an extract from a letter which he directed to Marmont on January 5[207].

[207] Marmont, _Correspondance_, book xv _bis_, p. 287.

‘I have the honour to enclose herewith two letters dated on the 1st and 3rd instant from General Thiébault at Salamanca. I attach no credence to their contents, for during the last six months I have been receiving perpetually similar reports.... If, contrary to my opinion, the English have really made some tentative movements on Ciudad Rodrigo, and if Julian Sanchez has tried to cut our communication with that place, I can only attribute it to your recent movement on Valencia. In that case, the unforeseen reappearance of your Excellency here may make the enemy change his plan of operations, and may prove harmful to him.’

Thiébault had cried ‘Wolf!’ too often to please Dorsenne, and the latter had no real apprehension that Wellington was already on the move. No more had Marmont. On arriving at Valladolid on January 13th he wrote to Berthier (five days after the trenches were opened at Rodrigo!), ‘It is probable that the English may be on the move at the end of February, and then I shall have need of all my troops: I have, therefore, told Montbrun to start on his backward march towards me before the end of January[208].’ By the end of January Rodrigo had already been for twelve days in the hands of the British army.

[208] Ibid., p. 291.

And if Dorsenne and Marmont were blind to the actual situation, so, most of all, was their master. The dispatch which gave over the charge of the kingdom of Leon to Marmont contains the following paragraph:

‘If General Wellington (_sic_) after the rainy season is over (i. e. after February) should determine to take the offensive, you can then unite all your eight divisions for a battle: General Dorsenne from Burgos would support you, by marching up from Burgos to your assistance. But such a move is not to be expected (_n’est pas présumable_). The English, having suffered heavy losses, and experiencing great difficulties in recruiting their army, all considerations tend to make us suppose that they will simply confine themselves to the defence of Portugal.... Your various dispatches seem to prove that it is at present no longer possible for us to take the offensive against Portugal, Badajoz being barely provisioned, and Salamanca having no magazines. It is necessary, therefore, to wait till the crops of the present year are ripe [June!], and till the clouds which now darken the political situation to the North have disappeared. His Majesty has no doubt that you will profit by the delay, to organize and administer the provinces under your control with justice and integrity, and to form large magazines.... The conquest of Portugal and the immortal glory of defeating the English are reserved for you. Use therefore all possible means to get yourselves into good condition for commencing this campaign, when circumstances permit that the order for it should be given.... Suggestions have been made that Ciudad Rodrigo should be dismantled. The Emperor considers that this would be a great mistake: the enemy, establishing himself in that position, would be able to intercept the communications between Salamanca and Plasencia, and that would be deplorable. The English know quite well that if they press in upon Rodrigo, or invest it, they expose themselves to be forced to deliver a battle--that is the last thing they want: however, if they did so expose themselves, it would be your duty to assemble your whole army and march straight at them[209].’

[209] Berthier to Marmont, Dec. 13, as above.

Such being Napoleon’s views at midwinter, it is strange to find Napier asserting that the disasters of the French at this time were caused partly by the jealousies of his lieutenants, partly by their failing to understand his orders in their true spirit, so that they neglected them, or executed them without vigour[210]. Without denying that Marmont, Dorsenne, and Soult were jealous of each other, we may assert that the real fundamental origin of all their disasters was that their master persisted in directing the details of the war from Paris, founding his orders on data three weeks old, and sending those orders to arrive another fortnight or three weeks after they had been written. As a fair example of what was perpetually happening we may cite the following dates. Wellington started to move on January 1st, 1812, as Thiébault wrote to Dorsenne (on the report of a Spanish spy) on January 3rd: on January 27 the general information that the Anglo-Portuguese army had crossed the Agueda, without any details, reached the Emperor, and caused him to dictate a dispatch for Dorsenne, giving him leave to detain the two divisions of the Imperial Guard under orders for France, and to support Marmont with them: the Emperor added that he hoped that by January 18th Montbrun would be nearing Madrid, and that by the end of the month his column would have joined the Army of Portugal. Eight days _before_ this dispatch was written Ciudad Rodrigo was already in Wellington’s hands: the news of its fall on January 19th seems to have reached Paris on February 11th[211], whereupon, as we shall presently see, the Emperor dictated another dispatch to Marmont, giving elaborate instructions on the new condition of affairs. This (travelling quicker than most correspondence) reached Marmont at Valladolid on February 26[212]: but of what use to the Marshal on that day were orders dictated upon the basis of the state of affairs in Leon on January 19th? ‘On ne dirige pas la guerre à trois ou quatre cents lieues de distance,’ as Thiébault very truly observed[213].

[210] _Peninsular War_, iv. p. 134.

[211] Correspondence in King Joseph’s _Letters_, viii. pp. 306-7.

[212] See Marmont’s letter acknowledging its receipt in his _Correspondance_, iv. pp. 342-3.

[213] _Mémoires_, iv. p. 554.

It was precisely Napoleon’s determination to dictate such operations as Montbrun’s Alicante expedition, or the transference of Marmont’s head-quarters from the valley of the Tagus to Valladolid, without any possible knowledge of the circumstances of his lieutenants at the moment when his orders would come to hand, that was the fatal thing. With wireless telegraphy in the modern style he might have received prompt intelligence, and sent directions that suited the situation. But under the conditions of Spain in 1812 such a system was pure madness.

‘The Emperor chose,’ as Marmont very truly observes, ‘to cut down the numbers of his troops in Spain [by withdrawing the Guards and Poles] and to order a grand movement which dislocated them for a time, precisely at the instant when he had increased the dispersion of the Army of Portugal, by sending a detachment of 12,000 men against Valencia. He was undoubtedly aware that the English army was cantoned in a fairly concentrated position on the Agueda, the Coa, and the Mondego. But he had made up his mind--I cannot make out why--that the English were not in a condition to take the field: in every dispatch he repeated this statement.’ In fairness to his master, Marmont should have added that he was of the same opinion himself, that Dorsenne shared it, and that both of them agreed to treat the Cassandra-like prophecies which Thiébault kept sending from Salamanca as ‘wild and whirling words.’

Marmont reached Valladolid, marching ahead of the divisions of Clausel and Maucune, on the 11th or 12th of January. He found Dorsenne waiting for him, and they proceeded to concert measures for the exchange of territory and troops which the Emperor had imposed upon them. After dinner on the evening of the 14th arrived Thiébault’s definite and startling news that Wellington, with at least five divisions in hand, had invested Rodrigo on the 8th, and was bringing up a heavy battering-train. The siege had already been six days in progress.

This was very alarming intelligence. The only troops actually in hand for the relief of Rodrigo were Thiébault’s small division at Salamanca, Souham’s much larger division about La Baneza and Benavente, and Clausel’s and Maucune’s divisions, now approaching Valladolid from the side of Avila. The whole did not make much more than 20,000 men, a force obviously insufficient to attack Wellington, if he were in such strength as Thiébault reported. Dorsenne at once sent for Roguet’s division of the Imperial Guard from Burgos: Marmont ordered Bonnet to evacuate the Asturias and come down by the route of Leon to join him: he also directed Brennier to come up from the Tagus, and Ferey to hurry his march from La Mancha. Aides-de-camp were sent to hunt for Foy, who was known to be on the borders of the Murcian regions, where Montbrun had dropped him on his march to Alicante. Montbrun himself, with the rest of his column, was also to turn back as soon as the orders should reach him.

By this concentration Marmont calculated[214] that he would have 32,000 men in line opposite Wellington by January 26 or 27th, as Bonnet, Brennier, and Dorsenne’s Guards should have arrived by then. And by February 1 Ferey and Foy ought also to be up, and more than 40,000 men would be collected. Vain dates! For Wellington captured Rodrigo on the 19th, seven days before the Marshal and Dorsenne could collect even 32,000 men.

[214] _Mémoires_, iv. p. 184.

Meanwhile Marmont pushed on for Salamanca, where the troops were to concentrate, having with him only the divisions of Clausel and Maucune. On January 21st he had reached Fuente Sauco, one march north of Salamanca, when he received the appalling news that Ciudad Rodrigo had been stormed by Wellington two days before. This was a thunderstroke--his army was caught not half concentrated, and he was for the moment helpless. He advanced as far as Salamanca, and there picked up Thiébault’s division, but even so he had not more that 15,000 men in hand, and dared not, with such a handful, march on Rodrigo, to endeavour to recover it before Wellington should have restored its fortifications. Bonnet had not yet even reached Leon: Ferey and Dorsenne’s Guard division had not been heard of. As to where Foy and Montbrun might be at the moment, it was hardly possible to hazard a guess. The only troops that could be relied upon to appear within the next few days were the divisions of Souham and Brennier. Even with their help the army would not exceed 26,000 or 28,000 men.

Meanwhile Wellington, with seven divisions now in hand, for he had brought up both the 5th and the 7th to the front, was lying on the Agueda, covering the repairs of Ciudad Rodrigo. Marmont had at first thought that, elated by his recent success, the British general might push his advance towards Salamanca. He made no signs of doing so: all his troops remained concentrated on the Portuguese frontier, ready to protect the rebuilding of Rodrigo. Here, on the day after the storm, all the trenches were filled in, and the débris on the breaches removed. Twelve hundred men were then turned to the task of mending the breaches, which were at first built up with fascines and earth only, so as to make them ready within a few days to resist a _coup-de-main_. In a very short time they were more or less in a state of defence, and on February 15th Castaños produced a brigade of Spanish infantry to form the new garrison of the place. The work was much retarded by the weather. Throughout the time of the siege it had been bitterly cold but very dry: but on the 28th the wind shifted to the west, and for the nine days following there was incessant and torrential rain, which was very detrimental to the work. It had, however, the compensating advantage of preventing Marmont from making any advance from Salamanca. Every river in Leon was over its banks, every ford impassable, the roads became practically useless. When, therefore, on February 2nd[215] the Agueda rose to such a height that Wellington’s trestle-bridge was swept away, and the stone town-bridge of Rodrigo was two feet under water, so that the divisions cantoned on the Portuguese frontier were cut off from the half-repaired fortress, there was no pressing danger from the French, who were quite unable to move forward.

[215] Napier says Jan. 29. But Jones, then employed in repairing Rodrigo, gives Feb. 2 in his diary of the work.

Marmont, as we have seen, had reached Fuente Sauco on January 21st, and Salamanca on January 22nd. On the following day Souham, coming in from the direction of Zamora, appeared at Matilla, half way between Salamanca and Ciudad Rodrigo, so that he was in touch with his chief and ready to act as his advanced guard. But no other troops had come up, and on the 24th the Marshal received a hasty note from Dorsenne, saying that the division of the Young Guard from Burgos would not reach the Tormes till February 2[216]. With only four divisions at his disposition (Clausel, Maucune, Thiébault, Souham) Marmont dared not yet move forward, since he knew that Wellington had at least six in hand, and he shrank from committing himself to decisive action with little more than 20,000 men assembled. On the 28th Dorsenne sent in a still more disheartening dispatch than his last: he had now ordered Roguet’s Guards, who had got as far forward as Medina del Campo, to return to Burgos[217]. The reasons given were that Mina had just inflicted a severe blow on General Abbé, the commanding officer in Navarre, by beating him near Pampeluna with a loss of 400 men, that the Conde de Montijo, from Aragon, had laid siege to Soria, and was pressing its garrison hard, and that another assembly of guerrillero bands had attacked Aranda del Duero, and would take it, if it were not succoured in a few days. ‘I therefore trust that your excellency will approve of my having called back Roguet’s division, its artillery, and Laferrière’s horse, to use them for a _guerre à outrance_ against the guerrillas.’ Nothing serious--he added--would follow, as all reports agreed that Wellington was sitting tight near Ciudad Rodrigo, and would make no advance toward Salamanca.

[216] Dorsenne to Marmont, from Valladolid, Feb. 24.

[217] Same to same, from Valladolid, Feb. 27.

No succours whatever, therefore, were to be expected from the Army of the North: Bonnet had only just recrossed the Cantabrian mountains, much incommoded by the bad weather in the passes, and Foy and Montbrun were only expected in the neighbourhood of Toledo early in February. Therefore Marmont abandoned all hope of attacking Wellington before Ciudad Rodrigo should be in a state of defence. The desperately rainy weather of January 28th to February 6th was no doubt the last decisive fact in making the Marshal give up the game. Before the rain had ceased falling, he concluded that all chance of a successful offensive move was gone, for he returned from Salamanca to Valladolid on February 5th.

On February 6th he wrote to Berthier[218] that he had ordered Montbrun and Foy, on their return from the Alicante expedition, to remain behind in the valley of the Tagus, and not to come on to Salamanca. His reason for abandoning all idea of a general concentration against Wellington in the kingdom of Leon, was that he was convinced that the next move of the British general would be to make a dash at Badajoz, and that he wished to have a considerable force ready in the direction of Almaraz and Talavera, with which he could succour the Army of the South, when it should be compelled to march, as in 1811, to relieve that fortress. His forecast of Wellington’s probable scheme of operations was perfectly correct, and his idea that the best way to foil it would be to hold a large portion of his army in the valley of the Tagus was correct also. But he was not to be permitted to carry out his own plan: the orders from Paris, which he so much dreaded, once more intervened to prescribe for him a very different policy[219].

[218] Marmont to Berthier, Valladolid, Feb. 6. Not in Marmont’s _Mémoires_, but printed in King Joseph’s Correspondance, viii. p. 301.

[219] I must confess that all Napier’s comment on Marmont’s doings (vol. iv. pp. 94-5) seems to me to be vitiated by a wish to vindicate Napoleon at all costs, and to throw all possible blame on his lieutenant. His statements contain what I cannot but call a _suggestio falsi_, when he says that ‘Bonnet quitted the Asturias, Montbrun hastened back from Valencia, Dorsenne sent a detachment in aid, and on Jan. 25 six divisions of infantry and one of cavalry, 45,000 men in all, were assembled at Salamanca, from whence to Ciudad is only four marches.’ This misses the facts that (1) Marmont had only _four_ divisions (Souham, Clausel, Maucune, and the weak division of Thiébault); (2) that Bonnet had not arrived, nor could for some days; (3) that Dorsenne sent nothing, and on Jan. 27 announced that nothing would be forthcoming; (4) that Montbrun (who was at Alicante on Jan. 16) was still far away on the borders of Murcia. With 22,000 men only in hand Marmont was naturally cautious.

Wellington during the critical days from January 20th to February 6th was naturally anxious. He knew that Marmont would concentrate against him, but he hoped (as indeed he was justified in doing) that the concentration would be slow and imperfect, and that the Marshal would find himself too weak to advance from Salamanca. His anxiety was made somewhat greater than it need have been, by a false report that Foy and Montbrun were already returned from the Alicante expedition--he was told that both had got back to Toledo by the beginning of January[220]--a most mischievous piece of false news. An equally groundless rumour informed him that Bonnet had left the Asturias, many days before his departure actually took place. On January 21 he wrote to Lord Liverpool that Bonnet had passed Benavente on his way to Salamanca, and that ‘the whole of what had gone eastward’ [i. e. Foy and Montbrun] was reported to be coming up from the Tagus to Valladolid, so that in a few days Marmont might possibly have 50,000 men in hand[221]. To make himself strong against such a concentration he ordered Hill, on January 22, to bring up three brigades of the 2nd Division to Castello Branco, with which he might join the main army at a few days’ notice[222]. At the same time he directed General Abadia to send a force to occupy the Asturias, which must be empty since Bonnet had evacuated it. It was not till some days later that he got the reassuring, and correct, news that Foy and Montbrun, instead of being already at the front in Castile, were not even expected at Toledo till January 29th, and that Bonnet had started late, and was only at La Baneza when February had already begun. But, by the time that he had received this information, it had already become evident that Marmont was not about to take the offensive, and Ciudad Rodrigo was already in a condition to resist a _coup-de-main_; while, since the whole siege-train of the Army of Portugal had been captured therein, it was certain that the Marshal could not come up provided with the artillery required for a regular siege.

[220] See _Dispatches_, viii. p. 547.

[221] I fancy that Wellington’s erroneous statement that Marmont had six divisions collected at Salamanca on the 23rd-24th [misprinted by Gurwood, _Dispatches_, viii. p. 577, as ‘the 6th Division!’] was Napier’s source for stating that such a force was assembled, which it certainly was not, Wellington reckoned that Marmont had Souham, Clausel, Maucune, Thiébault, and two divisions from the East, which last had not really come up--and never were to do so.

[222] Wellington to Hill, Jan. 22, _Dispatches_, viii. p. 566.

By February 12th the real state of affairs became clear, ‘the enemy has few troops left at Salamanca and in the towns on the Tormes, and it appears that Marshal Marmont has cantoned the right of his army on the Douro, at Zamora and Toro, the centre in the province of Avila, while one division (the 6th) has returned to Talavera and the valley of the Tagus.’ This was nearly correct: Marmont, on February 6th, had defined his position as follows--two divisions (those just returned from the Alicante expedition) in the valley of the Tagus; one, the 6th (Brennier), at Monbeltran, in one of the passes leading from the Tagus to the Douro valley; one (Clausel) at Avila; three on the Douro and the Esla (Zamora, Toro, Benavente) with a strong advanced guard at Salamanca. The heavy detachment towards the Tagus, as he explained, was to provide for the probable necessity of succouring Badajoz, to which Wellington was certain to turn his attention ere long.

Marmont was perfectly right in his surmise. Ciudad Rodrigo had hardly been in his hands for five days, when Wellington began to issue orders presupposing an attack on Badajoz. On January 25th Alexander Dickson was directed to send the 24-lb. shot and reserve powder remaining at the artillery base at Villa da Ponte to be embarked on the Douro for Oporto, where they were to be placed on ship-board[223]. Next day it was ordered that sixteen howitzers of the siege-train should start from Almeida overland for the Alemtejo, each drawn by eight bullocks, while twenty 24-pounders were to be shipped down the Douro from Barca de Alva to Oporto, and sent round from thence to Setubal, the seaport nearest to Elvas[224]. On the 28th Dickson himself was ordered to start at once for Setubal, in order that he might be ready to receive each consignment on its arrival, and to make arrangements for its transport to Elvas[225], while a dispatch was sent to Hill[226] definitely stating that, if all went well, the siege of Badajoz was to begin in the second week of March.

[223] _Dickson Papers_, ii. p. 571.

[224] Wellington, _Dispatches_, viii. pp. 568-9.

[225] _Dickson Papers_, ii. p. 576.

[226] Wellington to Hill, _Dispatches_, viii. p. 571.

These plans were drawn up long before it was clear that the army might not have to fight Marmont on the Agueda, for the defence of Ciudad Rodrigo. ‘If they should move this way, I hope to give a good account of them,’ Wellington wrote to Douglas (the British officer attached to the Army of Galicia)[227]: but he judged it more likely that no such advance would be made. ‘I think it probable that when Marmont shall have heard of our success, he will not move at all[228].’ Meanwhile there was no need to march the army southward for some time, since the artillery and stores would take many weeks on their land or water voyage, when roads were bad and the sea vexed with winter storms. So long as seven divisions were cantoned behind the Agueda and Coa, Marmont could have no certain knowledge that the attack on Badajoz was contemplated, whatever he might suspect. Therefore no transference southward of the divisions behind the Agueda was begun till February 19th. But Wellington, with an eye on Marmont’s future movements, contemplated a raid by Hill on the bridge of Almaraz, the nearest and best passage which the French possessed on the Tagus. If it could be broken by a flying column, any succours from the Army of Portugal to the Army of the South would have to take a much longer route and waste much time[229]. The project was abandoned, on Hill’s report that he doubted of its practicability, since a successful _coup-de-main_ on one of the bridge-head forts might not secure the actual destruction of the boats, which the French might withdraw to the farther side of the river, and relay at their leisure[230]. But, as we shall see, the scheme was postponed and not entirely rejected: in May it was carried out with complete success.

[227] Wellington to Sir Howard Douglas, Jan. 22, _Dispatches_, viii. p. 568.

[228] Wellington to Hill, _Dispatches_, viii. p. 567, same day as last.

[229] Wellington to Hill, Jan. 28, _Dispatches_, viii. pp. 571-2 and 586-7.

[230] Wellington to Hill, Feb. 12, _Dispatches_. viii. p. 603.

While Wellington was awaiting the news that his siege artillery was well forward on the way to Elvas, Marmont had been undergoing one of his periodical lectures from Paris. A dispatch sent to him by Berthier on January 23, and received at Valladolid on February 6th--fourteen days only having been occupied by its travels--had of course no reference to Wellington or Ciudad Rodrigo, the news of the investment of that fortress having only reached Paris on January 27th. It was mainly composed of censures on Montbrun’s Alicante expedition, which Napoleon considered to have been undertaken with too large a force--’he had ordered a flying column to be sent against Valencia, a whole army corps had marched.’ But the paragraph in it which filled Marmont with dismay was one ordering him to make over at once 6,000 men to the Army of the North, whose numbers the Emperor considered to be running too low, now that the two Guard divisions had been directed to return to France.

‘Twenty-four hours after the receipt of this dispatch you will start off on the march one of your divisions, with its divisional artillery, and its exact composition as it stands at the moment of the arrival of this order, and will send it to Burgos, to form part of the Army of the North. His Majesty forbids you to change any general belonging to this division, or to make any alterations in it. In return you will receive three provisional regiments of detachments, about 5,000 men, whom you may draft into your battalions. They are to start from Burgos the day that the division which you are ordered to send arrives there. All the Guards are under orders for France, and can only start when your division has reached that place.... The Army of the North will then consist of three divisions: (1) that which you are sending off; (2) Caffarelli’s division (due at Pampeluna from Aragon); (3) a third division which General Dorsenne will organize from the 34th Léger, the 113th and 130th of the line and the Swiss battalions.... By this arrangement the Army of the North will be in a position to aid you with two divisions if the English should march against you[231].’

[231] The ‘third division’ practically represented Thiébault’s old division of the Army of the North, which had long held the Salamanca district. This division was to be deprived of its Polish regiment (recalled to France with all other Poles) and to be given instead the 130th, then at Santander. But the 130th really belonged to the Army of Portugal (Sarrut’s division), though separated from it at the moment. So Marmont was being deprived of one regiment more.

Along with this dispatch arrived another from Dorsenne[232], clamouring for the division which was to be given him--he had already got the notice that he was to receive it, as he lay nearer to France than Marmont. He promised that the three provisional regiments should be sent off, as the Emperor directed, the moment that the ceded division should reach him. The Duke of Ragusa could not refuse to obey such peremptory orders from his master, and ordered Bonnet’s division, from Benavente and Leon, to march on Burgos. His letter acknowledging the receipt of the Emperor’s dispatch was plaintive. ‘I am informed that, according to the new arrangement, the Army of the North will be in a position to help me with two divisions if I am attacked. I doubt whether His Majesty’s intentions on this point will be carried out, and in no wise expect it. I believe that I am justified in fearing that any troops sent me will have to be long waited for, and will be an insignificant force when they do appear. Not to speak of the slowness inevitable in all joint operations, it takes so long in Spain to get dispatches through, and to collect troops, that I doubt whether I shall obtain any help at the critical moment. ... The net result of all is that I am left much weaker in numbers.’

[232] Dorsenne to Marmont, from Uñas, Feb. 5.

Marmont might have added that the three provisional regiments, which he was to receive in return for Bonnet’s division and the 130th Line, were no real reinforcement, but his own drafts, long due to arrive at the front, but detained by Dorsenne in Biscay and Old Castile to garrison small posts and keep open communications. And he was not destined to receive them as had been promised: Dorsenne wrote on February 24 apologizing for not forwarding them at once: they were guarding the roads between Irun and Vittoria, and could not be spared till other troops had been moved into their scattered garrisons to relieve them.

On January 27th the news of the advance of Wellington against Ciudad Rodrigo had at last reached Paris--eight days after the fortress had fallen. It caused the issue of new orders by the Emperor, all exquisitely inappropriate when they reached Marmont’s hands on February 10th. The Marshal had been contemplating the tiresome results of the storm of the fortress for nearly three weeks, but Napoleon’s orders presupposed much spare time before Rodrigo would be in any danger: Dorsenne is to stop the march of the Guards towards France, and to bring up all the forces he can to help the Army of Portugal: Montbrun will be back at Madrid by January 18 [on which day he was really in the middle of the kingdom of Murcia], and at the front in Leon before February 1st. After his arrival the Army of Portugal will be able to take up its definitive line of action. Finally, there is a stab at Marmont, ‘the English apparently have advanced in order to make a diversion to hamper the siege of Valencia; they only did so because they had got information of the great strength of the detachment which the Army of Portugal made in that direction[233].’

[233] Napoleon to Berthier, Jan. 27.

The Marshal could only reply by saying that the orders were all out of date, that he had (as directed) given up Bonnet’s division to the Army of the North, and that, Ciudad Rodrigo having fallen far earlier than any one had expected, and long before any sufficient relieving force could be collected, he had been unable to save it, and had now cantoned his army (minus Bonnet) with four divisions in the valley of the Douro and three in the valley of the Tagus, in expectation of an approaching move on the part of Wellington towards Badajoz.

These dispositions had not long been completed when another dispatch arrived from Paris, dated February 11th, in which the Emperor censured once more all his lieutenant’s actions, and laid down for him a new strategical policy from which he was forbidden to swerve.

‘The Emperor regrets that when you had the division of Souham and three others united [i. e. on January 23] you did not move on Salamanca, to make out what was going on. That would have given the English much to think about, and might have been useful to Ciudad Rodrigo. The way to help the army under the present circumstances is to place its head-quarters at Salamanca, and concentrate your force there, detaching one division to the Tagus valley and also reoccupying the Asturias. [This concentration] will oblige the enemy to remain about Almeida and in the North, for fear of an invasion of Portugal. You might even march on Rodrigo, and, if you have the necessary siege artillery, capture the place--your honour is bound up with it. If want of the artillery or of food renders it necessary to put off such an operation, you could at least make an incursion into Portugal, and advance towards the Douro and Almeida. This menace would keep the enemy “contained”.... Your posture should be offensive, with Salamanca as base and Almeida as objective: as long as the English know that you are in strength at Salamanca they will not budge: but if you retire to Valladolid yourself, and scatter divisions to the rear, and above all if you have not got your cavalry effective by the time that the rainy season ends, you will expose all the north of Spain to misfortunes.

‘It is indispensable to reoccupy the Asturias, because more troops are needed to hold the edge of the plain as far as Biscay than to keep down that province. Since the English are divided into two corps, one in the South and the other opposite you, they cannot be in heavy strength: you ought to outnumber them greatly.... I suppose that you consider the English mad, for you believe them capable of marching against Badajoz when you are at Salamanca, i. e. of allowing you to march to Lisbon before they can get back. They will only go southward if you, by your ill-devised schemes, keep two or three divisions detached on the Tagus: that reassures them, and tells them that you have no offensive projects against them.

‘To recapitulate, the Emperor’s intentions are that you should stop at Salamanca, that you should reoccupy the Asturias, that your army should base itself on Salamanca, and that from thence you should threaten the English.’

It may seem profane to the worshippers of the Emperor to say that this dispatch was purely wrong-headed, and argued a complete misconception of the situation. But it is impossible to pass any other verdict on it. Marmont, since Bonnet’s division had been stolen from him, had seven divisions left, or about 44,000 men effective, including cavalry and artillery. The Emperor tells him to keep one division on the Tagus, to send a second to occupy the Asturias. This leaves him about 34,000 net to concentrate at Salamanca. With this force he is to attempt to besiege Rodrigo, or at least to execute a raid as far as Almeida and the Douro. ‘The English are divided and so must be much numerically inferior to you.’ But, as a matter of fact, the only British detachment that was not under Wellington’s hand at the moment was Hill’s 2nd Division, and he had just brought that up to Castello Branco, and would have had it with him in five days, if Marmont had advanced from Salamanca. The Marshal would have seen 55,000 men falling upon his 34,000 if he had moved on any day before the 20th of February, and Wellington was ‘spoiling for a fight,’ or, in his own quiet phraseology, ‘if the French move this way, I hope to give a good account of them[234].’ Supposing Marmont had, by some evil inspiration, done what the Emperor had wished him to do before the orders came, he would have been crushed by almost double numbers somewhere in the neighbourhood of Rodrigo or Almeida. The battle of Salamanca would have been fought six months too soon.

[234] Wellington to Douglas, _Dispatches_, viii. p. 568.

This is the crucial objection to Napoleon’s main thesis: he underrated Wellington’s numbers and his readiness to give battle. As to details we may observe (1) that there was no siege-train to batter Rodrigo, because the whole of the heavy guns of the Army of Portugal had been captured in that fortress. (2) That Wellington was ‘mad’ enough to march upon Badajoz with his whole army, precisely because he knew that, even if Marmont should invade Portugal, he could never get to Lisbon. He realized, as the Emperor did not, that an army of five or six divisions could not march on Lisbon in the casual fashion recommended in this dispatch, because it would starve by the way. Central Portugal, still suffering from the blight of Masséna’s invasion, could not have sustained 30,000 men marching in a mass and trying to live upon the country in the usual French style. And Marmont, as his adversary well knew, had neither great magazines at his base, nor the immense transport train which would have permitted them to be utilized. The best proof of the impracticability of Napoleon’s scheme was that Marmont endeavoured to carry it out in April, when nothing lay in front of him but Portuguese militia, and failed to penetrate more than a few marches into the land, because he could not feed his army, and therefore could not keep it concentrated.

The Marshal knew long beforehand that this plan was hopeless. He wrote to Berthier from Valladolid on February 26th as follows:

‘Your Highness informs me that if my army is united at Salamanca the English would be “mad” to move into Estremadura, leaving me behind them, and free to advance on Lisbon. But they tried this precise combination in May 1811, though all my army was then quite close to Salamanca, and though the Army of the North was then twice as strong as it is to-day, and though the season was then later and allowed us to find provender for our horses, and though we were then in possession of Ciudad Rodrigo. They considered at that time that we could not undertake such an operation [as a march on Lisbon], and were perfectly right. Will they think that it is practicable to-day, when all the conditions which I have just cited are changed to our disadvantage, and when they know that a great body of troops has returned to France?... Consequently no movement on this side can help Badajoz. The only possible course is to take measures directly bearing on that place, if we are to bring pressure upon the enemy and hope to attain our end. The Emperor seems to ignore the food question. This is the important problem; and if it could be ended by the formation of base-magazines, his orders could be executed with punctuality and precision. But we are far from such a position--by no fault of mine.... When transferred to the North in January, I found not a grain of wheat in the magazines, not a sou in the treasury, unpaid debts everywhere. As the necessary result of the absurd system of administration adopted here, there was in existence a famine--real or artificial--whose severity was difficult to realize. We could only get food for daily consumption in our cantonments by using armed force: there is a long distance between this state of affairs and the formation of magazines which would allow us to move the army freely.... The English army is always concentrated and can always be moved, because it has an adequate supply of money and transport. Seven or eight thousand pack-mules bring up its daily food--hay for its cavalry on the banks of the Coa and Agueda has actually been sent out from England[235]. His Majesty may judge from this fact the comparison between their means and ours--we have not four days’ food in any of our magazines, we have no transport, we cannot draw requisitions from the most wretched village without sending thither a foraging party 200 strong: to live from day to day we have to scatter detachments to vast distances, and always to be on the move.... It is possible that His Majesty may be dissatisfied with my arguments, but I am bound to say that I cannot carry out the orders sent me without bringing about a disaster ere long. If His Majesty thinks otherwise, I must request to be superseded--a request not made for the first time: if I am given a successor the command will of course be placed in better hands[236].’

[235] An exaggeration, but hay was actually brought to Lisbon and Coimbra, and used for the English cavalry brigades, which had been sent to the rear and cantoned on the Lower Mondego.

[236] Marmont to Berthier, Valladolid, Feb. 26. Marmont’s _Mémoires_, iv. pp. 344-5.

[Illustration: _Marshal Marmont, Duke of Ragusa_]

This was an admirable summary of the whole situation in Spain, and might have caused the Emperor to change his policy, if he had not by this time so hardened himself in his false conceptions as to be past conviction. As Marmont complains, his master had now built up for himself an imaginary picture of the state of affairs in the Peninsula, and argued as if the situation was what he wished it to be, not what it actually was. ‘Il suppose vrai tout ce qu’il voudrait trouver existant[237].’

[237] Marmont’s ‘Observations on the Imperial Correspondence of Feb. 1812,’ _Mémoires_, iv. p. 512.

A subsequent letter from Paris, dated February 21st and received about March 2nd, contained one small amelioration of Marmont’s lot--he was told that he might take back Bonnet’s division, and not cede it to Dorsenne, on condition that he sent it at once to occupy the Asturias. But it then proceeded to lay down in the harshest terms the condemnation of the Marshal’s strategy:

‘The Emperor charges me to repeat to you that you worry too much about matters with which you have no concern. Your mission was to protect Almeida and Rodrigo--and you have let them fall. You are told to maintain and administer the North, and you abandon the Asturias--the only point from which it can be dominated and contained. You are getting into a state of alarm because Lord Wellington sends a division or two towards Badajoz. Now Badajoz is a very strong fortress, and the Duke of Dalmatia has 80,000 men, and can draw help from Marshal Suchet. If Wellington were to march on Badajoz [he had done so the day before this letter was written] you have a sure, prompt, and triumphant means of bringing him back--that of marching on Rodrigo and Almeida.’

Marmont replied, with a suppressed rage that can be read between the lines even more clearly than in his earlier letters, ‘Since the Emperor attributes to me the fall of Almeida, which was given up before I had actually taken over the command of this army[238], I cannot see what I can do to shelter myself from censures at large: ... I am accused of being the cause of the capture of Ciudad Rodrigo: it fell because it had an insufficient garrison of inferior quality and a bad commandant. Dorsenne was neither watchful nor prescient. Was it for me to take care of a place not in my command, and separated from me by a chain of mountains, and by the desert that had been made by the six months’ sojourn of the Army of Portugal in the valley of the Tagus?... I am blamed for having cantoned myself in the valley of the Tagus after repulsing Lord Wellington beyond the Coa [at the time of El Bodon], but this was the result of the imperative orders of the Emperor, who assigned me no other territory than the Tagus valley. Rodrigo was occupied by troops of the Army of the North.... I have ordered General Bonnet to reoccupy the Asturias at once, and quite see the importance of the occupying of that province.... I am told that the Emperor thinks that I busy myself too much about the interests of others, and not enough about my own. I had considered that one of my duties (and one of the most difficult of them) was to assist the Army of the South, and that duty was formally imposed on me in some twenty dispatches, and specially indicated by the order which bade me leave three divisions in the valley of the Tagus. To-day I am informed that I am relieved of that duty, and my position becomes simpler and better! But if the Emperor relies with confidence on the effect which demonstrations in the North will produce on the mind of Wellington, I must dare to express my contrary opinion. Lord Wellington is quite aware that I have no magazines, and is acquainted with the immensely difficult physical character of the country, and its complete lack of food resources at this season. He knows that my army is not in a position to cross the Coa, even if no one opposes me, and that if we did so we should have to turn back at the end of four days, unable to carry on the campaign, and with our horses all starved to death[239].’

[238] To be exact, it was on May 10 that Marmont took over the command from Masséna, and Almeida was evacuated by Brennier that same night.

[239] I extract these various paragraphs from Marmont’s vast dispatch of March 2, omitting much more that is interesting and apposite.

This and much more to the same effect had apparently some effect on the mind of the Emperor. But the result was confusing when formulated on paper. Berthier replied on March 12:

‘Your letters of February 27 and 28 and March 2 have been laid before the Emperor. His Majesty thinks that not only must you concentrate at Salamanca, but that you must throw a bridge across the Agueda, so that, if the enemy leaves less than five divisions north of the Tagus, you may be able to advance to the Coa, against Almeida, and ravage all northern Portugal. If Badajoz is captured by two divisions of the enemy its loss will not be imputed to you, the entire responsibility will fall on the Army of the South. If the enemy leaves only two, three, or even four divisions north of the Tagus, the Army of Portugal will be to blame if it does not at once march against the hostile force before it, invest Almeida, ravage all northern Portugal, and push detachments as far as the Mondego. Its rôle is simply to “contain” six British divisions, or at least five: it must take the offensive in the North, or, if the enemy has taken the initiative, or other circumstances necessitate it, must dispatch to the Tagus, by Almaraz, the same number of divisions that Lord Wellington shall have dispatched to conduct the siege of Badajoz.’

This double-edged document reached Salamanca on March 27, _eleven days after Wellington had invested Badajoz_. The whole allied field army had marched for Estremadura in the last days of February, and not a single British division remained north of the Tagus. In accordance with the Emperor’s dispatches of February 11th and of February 18th, Marmont had already concentrated the bulk of his resources at Salamanca, drawing in everything except Bonnet (destined for the Asturias), Souham, who was left on the Esla to face the Army of Galicia, and the equivalent of another division distributed as garrisons in Astorga, Leon, Palencia, Zamora, and Valladolid. With five divisions in hand, or just coming up, he was on the move, as the Emperor had directed, to threaten Rodrigo and Almeida and invade northern Portugal.

The Paris letter of March 12, quoted above, suddenly imposed on Marmont the choice between continuing the attack on Portugal, to which he was committed, or of leading his whole army by Almaraz to Badajoz--it must be the whole army, since he was told to send just as many divisions southward as Wellington should have moved in that direction, and every one of the seven units of the allied army had gone off.

Since Badajoz was stormed on April 6th, only ten days after Marmont received on March 27 the Emperor’s dispatch of March 12, it is clear that he never could have arrived in time to help the fortress. In June 1811 he had accomplished a similar movement at a better season of the year, and when some time had been allowed for preparation, in fifteen days, but only by making forced marches of the most exhausting sort. It could not have been done in so short a time in March or April, when the crops were not ripe, the rivers were full, and the roads were far worse than at midsummer. Moreover (as we shall presently see) Wellington had placed a large containing force at Merida, half-way between Almaraz and Badajoz, which Marmont would have had to drive in--at much expense of time.

The Marshal’s perplexity on receiving the dispatch that came in upon March 27 was extreme. ‘The instructions just received,’ he wrote to Berthier, ‘are wholly contradictory to those of February 18 and February 21, imperative orders which forced me, against my personal conviction, to abandon my own plan, and to make it impossible to do what I regarded as suitable to the interests of the Emperor. The letters of February 18 and February 21 told me that his Majesty thought me a meddler in matters which did not concern me: he told me that it was unnecessary for me to worry about Badajoz, “a very strong fortress supported by an army of 80,000 men.” ... He gave me formal orders to abandon any idea of marching to succour it, and added that if Lord Wellington went thither, he was to be left alone, because by advancing to the Agueda I could bring him back at once. The letters of the 18th and 21st made it quite clear that His Majesty freed me from all responsibility for Badajoz, provided I made a demonstration on the Agueda. ... To-day your Highness writes that I _am_ responsible for Badajoz, if Lord Wellington undertakes its siege with more than two divisions. The concluding paragraph of your letter seems to give me permission to succour the place, by bringing up troops to the Tagus. So, after imperative orders have wrecked my original arrangement, which had prepared and assured an effective help for Badajoz, and after all choice of methods has been forbidden to me, I am suddenly given an option when it is no longer possible to use it.... To-day, when my troops from the Tagus valley have repassed the mountains, and used up the magazines collected there at their departure, when it is impossible to get from Madrid the means to establish a new magazine at Almaraz, my army, if it started from this point [Salamanca], would consume every scrap of food that could be procured before it could possibly reach Badajoz.... The movement was practicable when I was in my original position: it is almost impracticable now, considering the season of the year, and the probable time-limit of the enemy’s operations.... After ripe reflection on the complicated situation, considering that my main task is to hold down the North, and that this task is much greater than that of holding the South, taking into consideration the news that an English force is said to be landing at Corunna (an improbable story, but one that is being repeatedly brought me), considering that the Portuguese and Galician troops threaten to take the offensive from Braganza, remembering that your letters of February 18 and 21 state that Suchet’s Army of Aragon is reckoned able to reinforce the Army of the South, and considering that my dispositions have been made (in spite of immense preliminary difficulties) for a fifteen days’ march on the Agueda, which is already begun, I decide in favour of continuing that operation, though I have (as I said before) no great confidence in its producing any effective result.

‘Accordingly I am putting the division that came up from the Tagus in motion for Plasencia, with orders to spread the rumour that it is to rejoin the army by the pass of Perales and enter Portugal; I start from here with three more divisions for the Agueda; ... if I fought on the Tormes I could put one more division in line, five in all: the number of seven divisions of which the Emperor speaks could only be concentrated if the Army of the North[240] could send two divisions to replace my own two now on the lines of communications and the Esla.’

[240] Marmont writes the Army of the Centre, evidently in confusion for the Army of the North. The nearest posts of the Army of the Centre were 150 miles away from the Esla, while the Army of the North at Burgos was much closer. Moreover, the Army of the Centre had not two infantry divisions, but only one--d’Armagnac’s--and some _Juramentado_ regiments.

The recapitulation of all this correspondence may seem tedious, but it is necessary. When it is followed with care I think that one definite fact emerges. Napoleon was directly and personally responsible for the fall of Badajoz. Down to March 27th Marmont was strictly forbidden to take any precautions for the safety of that fortress, and was censured as a meddler and an alarmist, for wishing to keep a strong force in the valley of the Tagus, ready to march thither. On March 27 he was suddenly given an option of marching to Estremadura with his whole army. It appears to be an option, not a definite order, for Berthier’s sentence introducing the new scheme is alternative--the Army of Portugal is ‘to take the offensive in the North _or_, under certain circumstances, to march for Almaraz.’ But this point need not be pressed, for if taken as a definite order it was impracticable: Marmont received it so late that, if he had marched for Badajoz with the greatest possible speed, he would have reached it some days after the place was stormed. The fact that he believed that he would never have got there at all, because lack of food would have stopped him on the way, is indifferent. The essential point of Napoleon’s responsibility is that he authorized the march too late, after having most stringently forbidden it, in successive letters extending over several weeks.

That a march on Badajoz by the whole Army of Portugal (or so much of it as was not required to contain the Galicians and to occupy Asturias), if it had begun--as Marmont wished--in February or early March, would have prevented Wellington from taking the fortress, is not certain. A similar march in June 1811 had that effect, at the time of the operations on the Caya. But Wellington’s position was much better in February 1812 than it had been eight months earlier. This much, however, is clear, that such an operation had a possible chance of success, while Napoleon’s counter-scheme for a demonstration on the Agueda and an invasion of the northern Beira had no such prospect. The Emperor, for lack of comprehension of the local conditions, misconceived its efficacy, as Marmont very cogently demonstrated in his letters. Northern Portugal was a waste, where the Marshal’s army might wander for a few days, but was certain to be starved before it was many marches from the frontier. Napier, in an elaborate vindication of the Emperor, tries to argue that the Marshal might have taken Rodrigo by escalade without a battering-train, have assailed Almeida in similar fashion, have menaced Oporto and occupied Coimbra[241]. He deliberately ignores one essential condition of the war, viz. that because of the French system of ‘living on the country,’ Marmont had no magazines, and no transport sufficient to enable his army to conduct a long offensive campaign in a devastated and hostile land. His paragraphs are mere rhetoric of the most unfair kind. For example, he says, ‘Wellington with 18,000 men[242] escaladed Badajoz, a powerful fortress defended by an excellent governor and 5,000 French veterans: Marmont with 28,000 men would not attempt to escalade Rodrigo, although its breaches were scarcely healed and its garrison disaffected.’ This statement omits the essential details that Wellington had a large siege-train, had opened three broad breaches in the walls of Badajoz, and, while the enemy was fully occupied in defending them, escaladed distant points of the _enceinte_ with success. Marmont had no siege-train, and therefore could have made no breaches; he would have had to cope with an undistracted garrison, holding ramparts everywhere intact. Moreover, Ciudad Rodrigo and its outworks form a compact fortress, of not half the circumference of Badajoz and its dependencies. If Ney and Masséna, with an adequate siege apparatus, treated Rodrigo with respect in 1810, and proceeded against it by regular operations, Marmont would have been entirely unjustified in trying the desperate method of escalade in 1812. The fortifications, as Napier grudgingly admits, were ‘healed’: an escalade against Carlos de España’s garrison would certainly have met the same fate as Suchet’s assault on Saguntum, a much weaker and unfinished stronghold. But it is unnecessary to follow into detail Napier’s controversial statements, which are all part of a wrong-headed scheme to prove Napoleon infallible on all occasions and at all costs.

[241] See