Just after he had left Segovia[632] King Joseph received a dispatch from Soult, dated July 16. It was a reply to the peremptory orders sent him on July 6th, which had directed him to evacuate part of Andalusia and to send a large detachment to Toledo. This was a strange document, which amounted to an absolute refusal to obey instructions. After stating (quite falsely) that Hill was advancing with 30,000 men in Estremadura, and that in consequence he was himself about to repair thither, he announced that the evacuation of Andalusia would be ruinous to the French cause in Spain. ‘We could not find means to subsist either on the Tagus or in Estremadura, and from one position to another we should retreat as far as the Ebro. There is a way to avoid this; by taking the initiative we can save 6,000 sick and maimed men whom I should probably have to abandon, as well as 200,000 Spaniards (who have declared for your Majesty, and will be lost without hope), also 2,000 guns, and the only artillery arsenal now existing in Spain. A single order by your Majesty can effect this, and shorten the Spanish war by six campaigns. Let your Majesty come to Andalusia in person, with every man that can be collected: if the number is large we can increase the expeditionary force in Estremadura to 25,000 or 30,000 men, and transfer the seat of war to the left bank of the Tagus. The Army of Portugal, being relieved of pressure, will be able to come into line again. Whatever occurs, your Majesty will find yourself at the head of a splendid army, ready to deliver battle. If the worst came, and we were unlucky, there is always the resource of retiring on the Army of Aragon [in Valencia] and so keeping the field.... I have the honour to repeat to your Majesty that I cannot send any detachments beyond the Sierra Morena or the Guadiana, save by evacuating all Andalusia and marching with my whole army. I must have a positive order from your Majesty to that effect[633].’
This was an astonishing letter for a Commander-in-Chief to receive from a subordinate. Instead of obeying a very definite order to move a certain number of troops to a certain point, Soult replies by sending to the King an alternative plan of campaign. And this plan, it is not too much to say, was an absolutely perverse and insane one. It must be remembered that, when Soult was writing, the battle of Salamanca was still six days in the future, and the Army of Portugal was known to be at close quarters with Wellington and in urgent need of reinforcements. Soult urges his master to abandon Marmont to the enemy, to evacuate Madrid, to give up his communication with France, and to retire into Andalusia, where he would be cut off from all the other imperial armies, for it was not possible even to communicate with Suchet and Valencia, since the Spanish Army of Murcia blocked the way. The cardinal sin of this project was that if the French were to hold Spain at all, it was necessary for them to be strong in the North: Soult proposed to deliver over the North to Wellington, by leaving Marmont in the lurch. As Napoleon had observed, five months earlier, ‘a check to the Army of Portugal would be a calamity which would make itself felt all over Spain. A check to the Army of the South might force it back on Madrid or Valencia, but would be of a very different degree of importance[634].’ He had said much the same thing four years before, when first his armies were invading Spain; for he then expressed the opinion that a disaster to Bessières in Castile would be the one ruinous possibility: defeats in the South or East mattered comparatively little. Soult, blinded by his own interest in the viceroyalty of Andalusia, refused to see this obvious fact. Long after he had received the news of Salamanca, he persisted in maintaining that the true policy was to hold on to Seville, even when the British army was at Madrid, and the wrecks of Marmont’s forces were retiring on Burgos. Of this we shall hear more presently.
King Joseph on receiving Soult’s letter returned answer: ‘You will see by my letter of the 29th July the errors that you have been labouring under as to Lord Wellington’s real designs. Hasten, therefore, to carry out the orders which I give you—viz. to evacuate Andalusia and march with your whole army on Toledo[635].’ Even so the King did not obtain exact obedience to his commands, but received a second series of counter-projects: and in the end Soult marched not on Toledo but on Valencia, and only many days after he had been instructed to commence his movement.
Wellington was, of course, unaware of the exact motives which had induced King Joseph to make his flank march to Segovia, but he considered that it might mean that there was some intention on the part of Clausel to bring the Army of Portugal to join the Army of the Centre by way of the Upper Douro [i. e. via Aranda]. He therefore resolved to make such a conjunction impossible, by driving the King over the mountains and towards Madrid[636].
While Anson’s and Arentschildt’s cavalry continued the pursuit of Clausel on the 29th-30th, and the 1st and Light Divisions were brought up to the neighbourhood of Tudela, opposite Valladolid, the rest of the army was turned against King Joseph. It was necessary to find out, as a preliminary, whether he was really making a stand at Segovia. To ascertain this point D’Urban’s Portuguese horse pushed out from Olmedo on the 29th, and found the King’s cavalry in Santa Maria de Nieva, ten miles in front of Segovia. Deserters from the Spanish Guards here came in to D’Urban, and gave him useful information as to the exact strength of the Army of the Centre. On the 30th Wellington placed at D’Urban’s disposal the German Heavy Dragoons, a battalion of Halkett’s brigade of the 7th Division, and a British battery, telling him to drive in the enemy’s screen. The French gave way reluctantly, and on hearing of their attitude Wellington ordered the whole 7th Division to follow D’Urban’s detachment, and other divisions to make ready to move in succession. But the report that Segovia was being firmly held, as the point de rassemblement[637] for Clausel, turned out to be false, for when the flying column approached that city it learnt that the main body of the enemy had left it in the morning for the Guadarrama pass. A considerable rearguard, under General Espert[638], however, was left to guard Segovia till the King should have got a fair start; and its mediaeval walls made it defensible for a short time against a force without heavy artillery. D’Urban could do nothing with his cavalry, but sent to Wellington a request that the 7th Division might move round to intercept Espert’s retreat towards the Guadarrama by a forced march. His chief replied that he had no great faith in the success of any of these attempts to ‘cut the French off,’ and that it did not appear to him more practicable at Segovia than elsewhere. ‘The result of such attempts would merely be to fatigue the troops in getting into Segovia, and it might as well be done without fatiguing them.’ And so it was, for Espert decamped by night on August 3 unmolested, and D’Urban entered the place next morning, followed some time later by the infantry. He at once explored the mountain road toward the pass, and found that the French had completely disappeared: not even at the ‘Puerto’ of the Guadarrama was a vedette to be seen.
Wellington had now to revise his whole plan of campaign, since it had become clear that the two armies opposed to him had retreated in different directions, and could not possibly combine. While it was still conceivable that Clausel might defend the line of the Douro, he had brought up the main body of his infantry to Olmedo. But after his entry into Valladolid on the 30th, and the precipitate retreat of the Army of Portugal toward Burgos, he had been for two days under the impression that King Joseph might stand at Segovia. Not only had he sent on the German dragoons and the 7th Division to follow D’Urban, but on July 31st he moved his own head-quarters and the 3rd Division to Cuellar, while the 4th, 5th, and 6th Divisions were at El Pino on the Cega river, a few miles behind. He wrote next morning (August 1st) that he was in such a position that Joseph and Clausel could not possibly join, and that if the King lingered any longer at Segovia, ‘I can move upon him, and make him go quicker than he will like[639].’ But he imagined that the Army of the Centre would fall back instantly on Madrid—as indeed it was doing at the very moment that he was writing his dispatch.
On receiving the information that Joseph had vanished, Wellington halted for three whole days [August 2nd, 3rd, 4th] with his head-quarters at Cuellar, and his infantry gathered round him in its neighbourhood. The 1st and Light Divisions, which had marched as far as the Douro, came southward to join the rest. But it was only on the 5th that orders were issued for the march of nearly the whole army on Segovia, by the road to Mozencillo. During these three days of halt Wellington had made up his mind as to his general policy. Clausel, whose army was harmless for the present, was to be ignored: only a small containing force was to be left in front of him, while the main body of the Anglo-Portuguese host marched on Madrid.
The strategical purpose that determined this decision was never set forth in full by Wellington. His contemporary dispatches to Lord Bathurst and to Hill are short, and lack explanatory detail—he states his decision, but says little of his reasons for making it. Nor did he, at the end of the campaign, write any long official narrative of his doings, as he had done in 1810 and 1811. The causes that governed his action have to be deduced from scattered opinions expressed in many different documents. We need hardly take seriously the common French dictum, found in many a book written by his exasperated opponents, that he ‘wished to parade himself as conqueror and liberator in the Spanish capital.’ That was not the sort of motive which any serious student of Wellington’s character would dream of imputing to him. Nor, if we translate it into less offensive terms, would it be true to say that it was the political advantage of expelling the King from Madrid, and so demonstrating to all Europe the weakness of the French hold on the Peninsula, that was the determining cause of the march into New Castile and the abandonment of the campaign on the Douro. We must rather look for definite military reasons. And of these the predominant one was that he conceived that the most probable result of the battle of Salamanca would be to force the King to call up Soult and Suchet to Madrid, in order to check the Anglo-Portuguese army, even at the cost of abandoning great tracts of conquered land in Andalusia and Valencia. Such indeed, as we have already seen, was Joseph’s purpose. The order to Soult to evacuate his viceroyalty and to march on Toledo with his whole army had been issued a day or two before Wellington had made up his mind to turn southward. Suchet had been directed at the same time to send all that he could spare toward Madrid. Though the pursuit of Clausel to the Ebro offered many advantages, it would be a ruinous move if the enemy should concentrate 70,000 men at Madrid, and then march on Valladolid, to take the allied army in the rear and cut it off from Portugal.
It was quite uncertain whether Soult or Suchet would make this move. But that it was the correct one is certain. Wellington was aware that Soult had been summoned to send troops northward. Hitherto he had found excuses for refusing to obey, as his last intercepted dispatch of July 8th sufficiently showed. But the results of Salamanca might probably render further disobedience impossible: and the moment that Soult should hear of that tremendous event, it was reasonable to suppose that he would abandon his viceroyalty, and march to join the King with every available man. If he found Joseph and his army still in possession of Madrid, they would have a central base and magazines from which to operate, and a very favourable strategic position. It was true that Wellington could call up Hill’s 18,000 men, but this was the only succour on which he could count: neither Ballasteros nor the numerous garrison of Cadiz would ever appear in New Castile, if old experience was to be trusted. If some Spaniards did arrive, they would be very uncertain aid. Granted, therefore, that Soult marched on Toledo and Madrid to join the King, Wellington must take almost every man of the Salamanca army to face them, even allowing for the certain junction of Hill. He could only afford to leave a small ‘containing force’ to look after Clausel.
But there was another possibility which made the situation still more doubtful. Would Suchet also push up to join King Joseph with the Army of Valencia, or the greater part of it? If he should do so, the odds would be too great, and a defensive campaign to cover Portugal, and so much as was possible of the newly regained Spanish provinces, would be the only resource. But Suchet’s action depended upon a factor over which Wellington had some influence, though not a complete and dominating control. When he had started on the Salamanca campaign he had been relying on Lord William Bentinck’s Sicilian expedition to keep the French in Valencia engaged: an attack on Catalonia would draw Suchet northward with all his reserves, and nothing would be left which O’Donnell and the Spanish army of Murcia could not ‘contain.’ It will be remembered that a few days before the battle of Salamanca[640] Wellington had received the disheartening news that Bentinck had countermanded his expedition, and was turning himself to some chimerical scheme for invading Italy. This had left Suchet’s attention free for the moment, and he might conceivably have sent troops to join the Army of the Centre. Fortunately he had not done so—only Palombini’s division and the small garrison of Cuenca had been swept up by King Joseph, without the Marshal’s consent and much to his disgust.