The only reason for enlarging on this deplorable theme is that there

was a close connexion in the minds of all soldiers of the early

nineteenth century, from the highest to the lowest ranks, between the

idea that an over-obstinate garrison had forfeited quarter, and the

idea that the town they had defended was liable to sack. This may be

found plainly stated in Lannes’s summons to Palafox at Saragossa in

January 1809[277], in the capitulation-debate before the surrender

of Badajoz in 1811, in Augereau’s address to the inhabitants of

Gerona[278], in Leval’s summons to the governor of Tarifa[279], and

with special emphasis in Suchet’s threatening epistle to Blake on the

day before the fall of Valencia: ‘in a few hours a general assault

will precipitate into your city the French columns: if you delay till

this terrible moment, it will not be in my power to restrain the fury

of the soldiery, and you alone will be responsible before God and man

for the evils which will overwhelm Valencia. It is the desire to avert

the complete destruction of a great town that determines me to offer

you honourable terms of capitulation[280].’ It was hardly necessary

in the Napoleonic era to enlarge on the connexion between storm and

sack--it was presupposed. Every governor who capitulated used to put in

his report to his own government a mention of his ‘desire to spare the

unfortunate inhabitants the horrors of a storm.’

[277] See Belmas, ii. p. 381.

[278] Ibid., ii. pp. 844-5.

[279] Text in the _Defence of Tarifa_, p. 64, and in Arteche.

[280] Belmas, iv. p. 202.

This idea, sad to say, was as deeply rooted in the minds of British as

of French soldiers. It is frankly confessed in many a Peninsular diary.

‘The men were permitted to enjoy themselves (!) for the remainder of

the day,’ says Kincaid in his narrative of the fall of Badajoz, ‘and

the usual frightful scene of plunder commenced, which officers thought

it prudent to avoid for the moment by retiring to the camp[281].’ ‘The

troops were, of course, admitted to the immemorial privilege of tearing

the town to pieces,’ says another writer on another occasion[282]. The

man in the ranks regarded the connexion of storm and sack as so close

that he could write, ‘the prisoners being secured and the gates opened,

we were allowed to enter the town _for the purpose of plundering

it_[283].’ But perhaps the most eye-opening sentence on the subject is

Wellington’s official order of April 7, 1812, issued late in the day,

and when the sack had already been going on for fifteen or eighteen

hours, ‘It is now full time that the plunder of Badajoz should cease;

an officer and six steady non-commissioned officers will be sent from

each regiment, British and Portuguese, of the 3rd, 4th, 5th, and Light

Divisions into the town, at 5 a.m. to-morrow morning, to bring away any

men still straggling there[284].’

[281] Kincaid, p. 39.

[282] Leith Hay, ii. pp. 256-7.

[283] Memoirs of Donaldson of the 94th, p. 158.

[284] Wellington, _Supplementary Dispatches_, vii. p. 311.

It was unfortunately the fact that Badajoz was a Spanish and not a

French town, and this adds a special shame to the lamentable outrages

which were perpetrated in its streets for many hours after the storm.

It is comparatively seldom in war that an army takes by assault a town

which does not belong to the hostile power. The only parallel of recent

years to the sack of Badajoz had been that of Lübeck in November

1806. Blücher’s Prussian corps, retiring before the pursuing French,

trespassed on neutral territory by seizing on the old Hanseatic city,

which lay in its way, and endeavouring to defend it. The magistrates

protested, but were powerless, as they had no armed force at their

disposition. Then the French came upon the scene, and, after a fierce

fight, won their way over wall and ditch and took the place. They

sacked it from end to end with every circumstance of atrocity[285]:

Marshal Bernadotte, when importuned by the Burgomaster to stay the

horrors, said that he was sorry, but that his troops only recognized

the fact that they were in a stormed town--he and his officers

could only succeed in calling them off after the city had been half

destroyed. This was sufficiently horrible; but to sack a town belonging

to a friendly nation is a shade worse than to sack a neutral place--and

this the British troops did.

[285] It is said on good first-hand authority that all the

inmates of an asylum for female lunatics were raped. See

Lettow-Vorbeck, _Geschichte des Krieges von 1806-7_, ii. p. 384.

Two short quotations from eye-witnesses may serve to show the kind of

scenes that prevailed in Badajoz from the early hours of the morning on

April 7th down to the following night.

‘Unfortunate Badajoz,’ writes one narrator[286], ‘met with the usual

fate of places taken at the point of the bayonet. In less than an

hour after it fell into our possession it looked as if centuries had

gradually completed its destruction. The surviving soldier, after

storming a town, considers it as his indisputable property, and thinks

himself at liberty to commit any enormity by way of indemnifying

himself for the risking of his life. The bloody strife has made him

insensible to every better feeling: his lips are parched by the

extraordinary exertions that he has made, and from necessity, as well

as inclination, his first search is for liquor. This once obtained,

every trace of human nature vanishes, and no brutal outrage can be

named which he does not commit. The town was not only plundered of

every article that could be carried off, but whatever was useless or

too heavy to move was wantonly destroyed. Whenever an officer appeared

in the streets the wretched inhabitants flocked round him with terror

and despair, embraced his knees and supplicated his protection. But it

was vain to oppose the soldiers: there were 10,000 of them crowding

the streets, the greater part drunk and discharging their pieces in

all directions--it was difficult to escape them unhurt. A couple of

hundred of their women from the camp poured also into the place, when

it was barely taken, to have their share of the plunder. They were,

if possible, worse than the men. Gracious God! such tigresses in the

shape of women! I sickened when I saw them coolly step over the dying,

indifferent to their cries for a drop of water, and deliberately search

the pockets of the dead for money, or even divest them of their bloody

coats. But no more of these scenes of horror. I went deliberately into

the town to harden myself to the sight of human misery--and I have had

enough of it: my blood has been frozen with the outrages I witnessed.’

[286] Hodenberg of the K.G.L. See his letters published in

_Blackwood’s Magazine_ for March 1913, by myself.

Another eye-witness gives a passing glimpse of horrors. ‘Duty being

over, I chanced to meet my servant, who seemed to have his haversack

already well filled with plunder. I asked him where the regiment was:

he answered that he did not know, but that he had better conduct me to

the camp, as I appeared to be wounded. I certainly was hit in the head,

but in the excitement of the escalade had not minded it, nor had I felt

a slight wound in my leg: but, as I began to be rather weak, I took

his advice, and he assisted me on. In passing what appeared to be a

religious house I saw two soldiers dragging out an unfortunate nun, her

clothes all torn: in her agony she knelt and held up a cross. Remorse

seized one of the men, who appeared more sober than the other, and he

swore she should not be outraged. The other soldier drew back a step

and shot his comrade dead. At this moment we found ourselves surrounded

by several Portuguese: they ordered us to halt, and presented their

muskets at us. I said to my servant, “throw them some of your plunder:”

he instantly took off his haversack and threw it among them: some

dollars and other silver coin rolled out. They then let us pass--had

he not done so they would have shot us--as they did several others. We

got safe to the bastion, and my servant carried me on his back to the

camp, where I got a draught of water, fell asleep instantly, and did

not waken till after midday[287].’

[287] _Recollections of Col. P. P. Nevill, late Major 63rd_ [but

with the 30th at Badajoz], pp. 15-16.

‘In justice to the army’--we quote from another authority[288]--’I

must say that the outrages were not general: in many cases they were

perpetrated by cold-blooded villains who had been backward enough in

the attack. Many risked their lives in defending helpless women, and,

though it was rather a dangerous moment for an officer to interfere,

I saw many of them running as much risk to prevent inhumanity as they

did in the preceding night while storming the town.’ The best-known

incident of the kind is the story of Harry Smith of the 95th, who saved

a young Spanish lady in the tumult, and married her two days later,

in the presence of the Commander-in-Chief himself, who gave away the

bride. This hastily-wedded spouse, Juana de Leon, was the Lady Smith

who was the faithful companion of her husband through so many campaigns

in Spain, Belgium, and South Africa, and gave her name to the town in

Natal which, nearly ninety years after the siege of Badajoz, was to be

the scene of the sternest leaguer that British troops have endured in

our own generation. Harry Smith’s narrative of the Odyssey of himself

and his young wife in 1812-14, as told in his autobiography, is one of

the most romantic tales of love and war that have ever been set down on

paper.

[288] Donaldson of the 94th, p. 159.

It was not till late in the afternoon of the 7th that Wellington, as

has been already mentioned, came to the rather tardy conclusion that

‘it was now full time that the plunder of Badajoz should cease.’ He

sent in Power’s Portuguese brigade to clear out those of the plunderers

who had not already gone back exhausted to their camps, and erected

a gallows in the cathedral square, for the hanging of any criminals

who might be detected lingering on for further outrages. Authorities

differ as to whether the Provost Marshal did, or did not, put his

power in action: the balance of evidence seems to show that the mere

threat sufficed to bring the sack to an end. The men were completely

exhausted: Napier remarks that ‘the tumult rather subsided than was

quelled.’

SECTION XXXII: CHAPTER V

OPERATIONS OF THE FRENCH DURING THE SIEGE OF BADAJOZ

Before proceeding to demonstrate the wide-spreading results of the

fall of the great Estremaduran fortress, it is necessary to follow

the movements of the French armies which had been responsible for its

safety.

Soult had been before Cadiz when, on March 11, he received news from

Drouet that troops were arriving at Elvas from the North, and on March

20 the more definite information that Wellington had moved out in

force on the 14th, and invested Badajoz on the 16th. The Marshal’s

long absence from his head-quarters at Seville at this moment, when

he had every reason to suspect that the enemy’s next stroke would be

in his own direction, is curious. Apparently his comparative freedom

from anxiety had two causes. The first was his confidence that Badajoz,

with its excellent governor and its picked garrison, could be relied

upon to make a very long defence. The second was that he was fully

persuaded that when the time of danger arrived he could count on

Marmont’s help--as he had in June 1811. On February 7 he wrote to his

colleague[289] that he had just heard of the fall of Rodrigo, that

Wellington’s next movement would naturally be against Badajoz, and that

he was glad to learn that Montbrun’s divisions, on their return from

Alicante, were being placed in the valley of the Tagus. ‘I see with

pleasure that your excellency has given him orders to get in touch

with the Army of the South. As long as this communication shall exist,

the enemy will not dare to make a push against Badajoz, because at

his first movement we can join our forces and march against him for

a battle. I hope that it may enter into your plans to leave a corps

between the Tagus and the Guadiana, the Truxillo road, and the Sierra

de Guadalupe, where it can feed, and keep in touch with the troops

which I keep in the Serena [the district about Medellin, Don Benito,

and Zalamea, where Daricau was cantoned]. I am persuaded that, when

the campaigning season begins, the enemy will do all he can to seize

Badajoz, because he dare attempt nothing in Castille so long as that

place offers us a base from which to invade Portugal and fall upon his

line of communications.... I am bound, therefore, to make a pressing

demand that your left wing may be kept in a position which makes the

communication between our armies sure, so that we may be able, by

uniting our disposable forces, to go out against the enemy with the

assurance of success.’

[289] The letter is printed in Marmont’s _Correspondance_, iv.

pp. 304-5.

This was precisely what Marmont had intended to do. He was convinced,

like Soult, that Wellington’s next move would be against Badajoz, and

he placed Montbrun and the divisions of Foy, Brennier, and Sarrut about

Talavera, Monbeltran, and Almaraz, precisely in order that they might

be in easy touch with Drouet. On February 22 he wrote to his colleague

explaining his purpose in so doing, and his complete acquiescence in

the plan for a joint movement against Wellington, whenever the latter

should appear on the Guadiana[290]. His pledge was quite honest and

genuine, and in reliance on it Soult made all his arrangements. These,

however, appear to have been rather loose and careless: the Marshal

seems to have felt such complete confidence in the combination that he

made insufficient preparations on his own side. No reinforcements were

sent either to Badajoz or to Drouet, whose 12,000 men were dispersed

in a very long front in Estremadura, reaching from Medellin and Don

Benito on the right to Fregenal on the left. This is why Graham, when

he moved forward briskly on March 17th, found no solid body of the

enemy in front of him, but only scattered brigades and regiments, which

made off in haste, and which only succeeded at last in concentrating

so far to the rear as Fuente Ovejuna, which is actually in Andalusia,

and behind the crest of the Sierra Morena. We may add that having been

advised by Drouet as early as March 11th[291] that British troops were

accumulating behind Elvas, Soult ought to have taken the alarm at once,

to have moved back to Seville from Santa Maria by Cadiz, where he lay

on that date, and to have issued orders for the concentration of his

reserves. He did none of these things, was still in front of Cadiz on

March 20[292], and did not prescribe any movement of troops till, on

that day, he received Drouet’s more definite and alarming news that

Wellington was in person at Elvas, and had moved out toward Badajoz on

the 16th. Clearly he lost nine days by want of sufficient promptness,

and had but himself to blame if he could only start from Seville

with a considerable field-force on March 30. All that he appears to

have done on March 11 was to write to Marmont that the long-foreseen

hypothesis of a move of Wellington on Badajoz was being verified, and

that they must prepare to unite their forces. Jourdan has, therefore,

some justification for his remark that he does not see why Soult should

have been before Cadiz, amusing himself by throwing shells into that

place[293] as late as March 20th.

[290] This Soult quotes in his recriminatory letter to Marmont of

April 8, and in his angry dispatch to Berthier of the same date

(printed in King Joseph’s _Correspondance_, viii. p. 355).

[291] The date is proved by the letter from Soult to Marmont of

March 11, printed in Marmont’s _Mémoires_, iv. p. 359.

[292] The date is proved by Soult’s letter to the Emperor of

that date from Santa Maria, in which he announces his intention

to start, and says that he is writing to Marmont, to get him to

unite the armies as soon as possible.

[293] See his _Mémoires_, p. 377.

From the 20th to the 30th of that month Soult was busily engaged in

organizing the relief-column which, after picking up Drouet on the

way, was to march to the succour of Badajoz. He could not venture

to touch the divisions of Conroux and Cassagne, which together were

none too strong to provide for the manning of the Cadiz Lines and the

fending off of Ballasteros from their rear. But he called off the

whole division of Barrois, nearly 8,000 strong[294], Vichery’s brigade

of infantry from Leval’s division in the province of Granada[295],

and six regiments of Digeon’s and Pierre Soult’s dragoons. This, with

the corresponding artillery, made a column of some 13,000 men, with

which the Marshal started from Seville on the 30th March, crossed

the Guadalquivir at Lora del Rio next day, and moved on Constantina

and Guadalcanal. An interesting complication would have been caused

if Graham had been allowed to stop with his 19,000 men at Azuaga and

Llerena, where he was directly between Soult and Drouet’s position at

Fuente Ovejuna, and if Hill from Merida had moved against Drouet’s

corps. But as Wellington had withdrawn Graham’s column to Villafranca

on March 31, there was nothing left to prevent Drouet from coming in

from his excentric position, and joining his chief at Llerena on April

4th, with the 12,000 men of his own and Daricau’s divisions. This

gave the Marshal some 25,000 men[296] in hand, a force which would be

manifestly incapable of raising the siege of Badajoz, for he knew that

Wellington had at least 45,000 men in hand, and, as a matter of fact,

the arrival of the 5th Division and other late detachments had raised

the Anglo-Portuguese army to something more like 55,000 sabres and

bayonets.

[294] To be exact, 7,776 officers and men on March 1. He also

brought with him some ‘bataillons d’élite’ of grenadier companies

from Villatte’s division.

[295] The 55th, three battalions about 1,500 strong, the fourth

being left at Jaen. Soult says in his dispatch of April 8 that

he took a whole _brigade_ from Leval, but the states of April 14

show the 32nd and 58th regiments of Leval’s division, and three

of the four battalions of the 43rd, all left in the kingdom of

Granada. Apparently three battalions of the 55th and one of the

43rd marched, about 2,200 strong.

[296] Though he calls them only 21,000 in his dispatches. But the

figures [see Appendix no. VIII] show 23,500. The total in the

monthly reports indicate 25,000 as more likely.

Wellington’s orders, when he heard that Soult was in the passes, and

that Drouet was moving to join him, directed Graham to fall back on the

Albuera position, and Hill to join him there by the route of Lobon and

Talavera Real, if it should appear that all the French columns were

moving directly to the relief of Badajoz, and none of them spreading

out eastward towards the Upper Guadiana[297]. These conditions were

realized, as Soult moved in one solid body towards Villafranca and

Fuente del Maestre: so Hill evacuated Merida, after destroying its

bridge, and joined Graham on the old Albuera ground on April 6th. They

had 31,000 men, including four British divisions and four British

cavalry brigades, and Wellington could have reinforced them from the

lines before Badajoz with two divisions more, if it had been necessary,

while still leaving the fortress adequately blockaded by 10,000 or

12,000 men. But as Soult did not appear at Fuente del Maestre and

Villafranca till the afternoon of April 7th, a day after Badajoz had

fallen, this need did not arise. The Marshal, learning of the disaster,

hastily turned back and retired towards Andalusia, wisely observing

that he ‘could not fight the whole English army.’ It is interesting to

speculate what would have happened if he had lingered five days less

before Cadiz, had issued his concentration orders on the 14th or 15th

instead of the 20th March, and had appeared at Villafranca on the 2nd

instead of the 7th of the next month. His dispatch of April 17 states

that he had intended to fight, despite of odds, to save Badajoz: if he

had done so, and had attacked 40,000 Anglo-Portuguese with his 25,000

men, he must inevitably have suffered a dreadful disaster. He must

have fought a second battle of Albuera with much the same strength

that he had at the first, while his enemy would have had six British

divisions instead of two, and an equal instead of a wholly inferior

cavalry. The result of such a battle could hardly have failed to be

not only a crushing defeat for the French, but the prompt loss of all

Andalusia; for thrown back on that kingdom with a routed army, and

unable to gather in promptly reserves scattered over the whole land,

from the Cadiz Lines to Granada and Malaga, he must have evacuated his

viceroyalty, and have retreated in haste either on La Mancha or on

Valencia.

[297] The orders to Hill issued by Wellington on April 4 and 5

(_Dispatches_, ix. p. 30) contemplate two possibilities: (1)

Soult is marching with his whole force on Villafranca, and Foy is

remaining far away: in this case Hill is to move _en masse_ on

Albuera. This is the case that actually occurred; (2) if Foy is

moving toward the Upper Guadiana, and Soult is showing signs of

extending to join him, Howard’s British and Ashworth’s Portuguese

brigades and Campbell’s Portuguese horse will stay at Merida as

long as is prudent, in order to prevent the junction, and will

break the bridge at the last moment and then follow Hill.

Wellington, when he wrote his first orders of the 4th to

Hill, was intending to storm Badajoz on the 5th, and knew, by

calculating distances, that Soult could not be in front of

Albuera till the 7th. He ultimately chanced another day of

bombardment, running the time limit rather fine. But there was no

real risk with Graham and Hill at Albuera: Soult could not have

forced them.

It is most improbable, however, that Soult would really have ventured

to attack the Albuera position[298], in spite of the confident

language of his ex-post-facto dispatches. His whole plan of operations

depended on his being joined by the Army of Portugal, in accordance

with Marmont’s promise of February 22nd. And he was well aware, by a

letter sent by Foy to Drouet on March 31st, and received on April 6th,

that he could expect no help from the North for many weeks, if any

came at all. That Badajoz was never relieved was due, not to Soult’s

delay in concentrating (though this was no doubt unwise), nor to his

over-confidence in Phillipon’s power of resistance, which was (as it

turned out) misplaced. He wrote to Berthier that ‘the garrison wanted

for nothing--it had still food for two months, and was abundantly

provided with munitions: its total strength was 5,000 men: it had

victoriously repulsed three assaults: the men were convinced that,

however great a hostile force presented itself before the breaches,

it would never carry them: Phillipon had been informed on March 28th

that I was marching to his help: the troops were in enthusiastic

spirits, though they had already lost 500 men in successful sorties: my

advanced guard was at only one long day’s march from the place, when it

succumbs!’ It was indeed an _évènement funeste_!

[298] He says in his letter to Berthier of April 8 that he had

intended (but for the fall of Badajoz) to move by his right that

morning, to the lower course of the Guadajira river--which would

have brought on an action near Talavera Real, lower down the

stream of the Albuera than the battle-spot of May 1811.

But Soult’s late arrival and miscalculation of the time that the

siege would take, were neither of them the causes of the fall of

Badajoz. It would have fallen none the less if he had arrived on the

Albuera upon April 2nd. The fate of the place was really settled

by Napoleon’s dispatches to Marmont, with which we dealt at great

length in an earlier chapter[299]. The orders of February 11 and

February 21 (received by the Duke of Ragusa on February 26 and March

2 respectively) forbade him to worry about Badajoz, ‘a very strong

fortress supported by an army of 80,000 men,’ and told him to withdraw

to Salamanca two of the three divisions which he was keeping in the

valley of the Tagus, and to reply to any movement of Wellington into

Estremadura by invading Northern Portugal. The plan which Soult and

Marmont had concerted for a joint relief of Badajoz was expressly

forbidden by their master, on his erroneous hypothesis that a thrust at

Ciudad Rodrigo and Almeida must bring Wellington home again. Marmont’s

promise of co-operation, sent off on February 22nd to Seville, was

rendered impossible--through no fault of his--by the imperial dispatch

received four days later, which expressly forbade him to stand by it.

‘The English will only go southward if you, by your ill-devised scheme,

keep two or three divisions detached on the Tagus: that reassures them,

and tells them that you have no offensive projects against them.’ So

Marmont, protesting and prophesying future disaster, was compelled to

withdraw two divisions from the central position on the Tagus, and to

leave there only Foy’s 5,000 men--a negligible quantity in the problem.

Nor was this all--he was not even allowed to send them back, since the

whole Army of Portugal was ordered to march into the Beira.

[299] See chapter ii above, pp. 54, 55.

Soult, therefore, was justified in his wrath when he wrote to Marmont

that he had been given a promise and that it had been broken, ‘if there

had been the least attempt to concert operations between the armies of

Portugal and the South, the English army would have been destroyed, and

Badajoz would still be in the power of the Emperor. I deplore bitterly

the fact that you have not been able to come to an arrangement with me

on the subject.’ But the wrath should have been directed against the

Emperor, not against his lieutenant, who had so unwillingly been forced

to break his promise. The only censure, perhaps, that can be laid upon

Marmont is that he should have made it more clear to Soult that, by

the new directions from Paris, he was rendered unable to redeem his

pledge. Soult was not, however, without warnings that something of the

kind might happen: Berthier had written to him on February 11th, and

the letter must have arrived by the middle of March, that the Emperor

was displeased to find him appealing for troops of the Army of Portugal

to be moved to Truxillo, and that he ought to be more dependent on his

own strength[300]. It would have been better if the Emperor’s trusty

scribe had explained to Soult that Marmont was expressly forbidden, in

a dispatch written that same day, to keep more than one division on the

Tagus, or to worry himself about the danger of Badajoz.

[300] Berthier to Soult, Feb. 11. The same date as the fatal

dispatch sent to Marmont, who was given a copy of that to Soult

as an enclosure.

Marmont’s original plan for joining Soult via Almaraz might have

failed--he himself confesses it in one of his replies to Berthier. But

it was the only scheme which presented any prospect of success. By

making it impossible Napoleon rendered the fall of Badajoz certain. For

it is no defence whatever to point out that his dispatch of March 12th,

which reached Salamanca on March 27, finally gave Marmont the option

of going southward. By that time it was too late to try the move:--if

the Duke of Ragusa had marched for Almaraz and Truxillo next morning,

he would still have been many days too late to join Soult before April

6th, the date on which Badajoz fell.

Summing up the whole operation, we must conclude that Wellington’s

plan, which depended for its efficacy on the slowness with which

the French always received information, and the difficulty which

they always experienced in concentrating and feeding large bodies of

troops in winter or early spring, was bound to be successful, unless

an improbable conjunction of chances had occurred. If Marmont and

Soult had both taken the alarm at the earliest possible moment, and

had each marched with the strongest possible field army, Soult with

the 25,000 men that he actually collected, Marmont with the three

divisions that lay on the Tagus on March 1st, and three more from

Castile[301], they might have met east of Merida somewhere about the

last days of March. In that case their united strength would have been

from 50,000 to 55,000 men: Wellington had as many, so that he would

not have been bound down to the mere defensive policy that he took

up on the Caya in June-July 1811, when his numbers were decidedly

less than now. But the chance that both Marmont and Soult would do

the right thing in the shortest possible time was unlikely. They

would have had terrible difficulties from the torrential rains that

prevailed in the last ten days of March, and the consequent badness of

the roads. Marmont’s (if not Soult’s) food-problem would have been a

hard one, as he himself shows in several of his letters. Soult got his

first definite alarm on March 11th: Marmont could hardly move till

he had learnt that Wellington had started for Estremadura in person:

till this was certain, he could not be sure that the main body of the

Anglo-Portuguese army was not still behind the Agueda. Wellington only

left Freneda on March 5th, and Marmont did not know of his departure

till some days later. If the two marshals had each issued prompt

concentration orders on March 11, it still remains very doubtful if

they would have met in time to foil Wellington’s object. As a matter of

fact Soult (as we have seen) delayed for nine days before he determined

to concentrate his field-force and march on Badajoz, and this lateness

would have wrecked the combination, even if Marmont had been more ready

than his colleague.

[301] More probably he would have brought only _two_ divisions

from north of the mountains, as he had to leave Bonnet to look

after the Asturians, and Souham’s single division would hardly

have sufficed to contain the Galicians, the Portuguese, and the

Guerrilleros.

Still there was some chance that the armies might have joined, if

Napoleon had not intervened with his misguided refusal to allow Marmont

to keep three divisions in the valley of the Tagus or to ‘worry about

affairs that did not concern him.’ Wellington could not know of these

orders: hence came his anxieties, and his determination to hurry the

siege of Badajoz to a conclusion at the earliest possible date. He was

never--as it turned out--in serious danger, but he could not possibly

be aware of the fact that Marmont was fettered by his instructions. It

was only the gradual accumulation of reports proving that the Army of

Portugal was moving against Ciudad Rodrigo, and not on Almaraz, that

finally gave him comparative ease of mind with regard to the situation.

As to Soult, he somewhat over-estimated his force, taking it at 30,000

or even 35,000 men rather than the real 25,000: this was, no doubt,

the reason why he resolved to fight with his ‘covering army’ ranged on

the Albuera position, and not farther forward. If he had known that

on April 1 Soult had only 13,000 men at Monasterio, and was still

separated from Drouet, he might possibly have been more enterprising.

No signs of Marmont’s arrival being visible, Wellington could afford

to contemplate with great equanimity Soult’s position at Villafranca

on April 7th. If the Marshal moved forward he would be beaten--but it

was almost certain that he would move back at once, for, as it will

be remembered, precautions had been taken to give him an alarming

distraction in his rear, by means of the operations of Penne Villemur

and Ballasteros[302]. This combination worked with perfect success,

far more accurately than Blake’s similar raid on Seville had done in

June 1811. Ballasteros, it is true, did much less than was in his

power. He started from his refuge under the guns of Gibraltar, passed

down from the Ronda mountains, and reached Utrera, in the plain of the

Guadalquivir less than twenty miles from Seville, on April 4. But he

then swerved away, having done more to alarm than to hurt the French,

though he had a force of 10,000 infantry and 800 horse[303], sufficient

to have put Seville in serious peril. But Penne Villemur and Morillo,

though they had not half the numbers of Ballasteros, accomplished all

that Wellington required: having slipped into the Condado de Niebla

almost unobserved, they pushed rapidly eastward, and occupied San Lucar

la Mayor, only twelve miles from Seville, on April 4, the same day

that Ballasteros appeared at Utrera. Their cavalry pushed up so boldly

toward the suburbs that they had to be driven off by cannon-shot from

the _tête-de-pont_ at the bridge of Triana. General Rignoux, governor

of Seville, had a very motley and insufficient garrison, as Wellington

had calculated when he sent Penne Villemur forth. The only organized

units were a battalion of ‘Swiss’ Juramentados--really adventurers

of all nations--and a regiment of Spanish horse, making 1,500 men

altogether: the rest consisted of convalescents and weakly men

belonging to the regiments in the Cadiz Lines, and of 600 dismounted

dragoons. These made up some 2,000 men more, but many were not fit to

bear arms. In addition there were some companies of the recently raised

‘National Guards.’ The enormous size of Seville, and the weakness

of its old wall, compelled Rignoux to concentrate his force in the

fortified Cartuja convent, leaving only small posts at the gates and

the bridge. He sent at once, as Wellington had hoped, pressing appeals

to Soult, saying that he was beset by 14,000 men, and that the citizens

would probably rise and let in the enemy.

[302] See above, p. 229.

[303] Infantry divisions of Cruz Murgeon (5,400 men) and the

Prince of Anglona (4,300 men) and five squadrons of horse,

besides irregulars.

On the 6th Ballasteros received false news that Conroux was marching

against him with the troops from the Cadiz Lines, and drew back into

the mountains. It is said that he was wilfully deceived by persons in

the French interest; at any rate he must have been badly served by his

cavalry and intelligence officers, who ought to have been able to tell

him that there was no foundation for the report. Penne and Morillo,

however, though disappointed at failing to meet their colleague’s army,

made a great parade of their small force under the walls of Seville,

and skirmished with the French at the bridge-head of Triana, and under

the walls of the Cartuja, so boldly that Rignoux expected a serious

attack. They could only have accomplished something more profitable if

the people of Seville had risen, but no disturbance took place. After

remaining in front of the place all the 7th and 8th of April, they

disappeared on the 9th, having received news of the fall of Badajoz,

and drawn the correct deduction that Soult would turn back to hunt them

when freed from his other task. Wellington, indeed, had written to give

them warning to that effect on the very morning that they retired[304]:

but they anticipated the danger, and were safely behind the Rio Tinto

when Soult turned up in hot haste at Seville on the 11th, after four

days of exhausting forced marches.

[304] Wellington to Col. Austin from Badajoz, April 9.

The Marshal had left the two divisions of Drouet and Daricau with

Perreymond’s cavalry in Estremadura, to act as an observing force,

and had marched with his remaining 13,000 men to save Seville, which

owing to Ballasteros’s timidity had never been in any real danger. But

the Spanish diversion had nevertheless had precisely the effect that

Wellington had expected and desired. During Soult’s short absence of

twelve days great part of the open country of Andalusia had fallen out

of his control, the communications with La Mancha and King Joseph had

been cut off, and the guerrilleros had blockaded all the smaller French

posts. The hold of the invaders upon the kingdom was never so secure as

it had been before the fall of Badajoz.

Ballasteros, after his fiasco in front of Seville, made two fruitless

attempts against isolated French garrisons. He failed at the Castle

of Zahara on April 11th. One of his columns in an assault on Osuna

two days later got into the town and killed or captured 60 of the

defenders, but failed to take the citadel, where the remainder defended

themselves till Pierre Soult was reported to be at hand, and the

Spaniards withdrew[305]. He ended his campaign of raids, however, with

a more successful stroke. Hearing that the brigadier Rey, with three

battalions and some dragoons, was marching from Malaga to relieve

the garrison of Ronda, he fell upon him at Alhaurin on the 14th with

his main body, encompassed him with fourfold strength, and drove him

in rout back to Malaga, capturing his two guns and inflicting more

than 200 casualties upon him[306]. Ballasteros then hoped to seize

on Malaga, where the French were much alarmed, and prepared to shut

themselves up in the citadel of Gibalfaro. But the news that Pierre

Soult and Conroux were approaching with a strong column caused the

Spaniards to retire to the mountains above Gibraltar [April 19th]. Thus

the operations in Andalusia, which had opened with Soult’s march to

Badajoz, came to an end, with no ruinous disaster to the French, but

with a diminution of their prestige, and a distinct weakening of their

hold on the kingdom. In the Condado de Niebla Soult made no attempt

to reoccupy lost ground, and east of Granada his line of posts had

recoiled considerably on the Murcian side: Baza and Ubeda had been

abandoned for good. It was but a vain boast when the Marshal wrote to

Berthier that, after he had set all things to rights in the central

parts of Andalusia, he intended to organize a general concentration to

crush Ballasteros, and that his next task would be to lay siege for a

second time to Tarifa, ‘the loss of which place would be more injurious

to the English and the Insurgents than that of Alicante, or even that

of Badajoz--against which last-named fortress I ought to make no

attack till I shall have finished matters on the Tarifa side, and so

have nothing to fear on my left flank[307].’

[305] Napier, I know not on what authority, says that Osuna was

only defended by ‘Juramentados’ who made a gallant resistance

against their own countrymen. But Soult, in a letter to Berthier

dated April 21 from Seville, says that Osuna was held by some

companies of the 43rd Line and a detachment of the 21st Dragoons.

He cannot be wrong. Moreover, the 43rd shows losses at Osuna,

April 13, in Martinien’s tables.

[306] Martinien’s tables show three officers killed and nine

wounded at ‘Alora near Malaga’ on this date, in the 43rd, 58th

Line, and 21st Dragoons. Soult’s dispatch makes out that only

Rey’s advanced guard under Maransin was cut up, and that the

main body defeated the Spaniards. If so, why did they retreat on

Malaga?

[307] Soult to Berthier from Seville, April 17, 1812.

To complete the survey of the fortunes of the Army of the South in

April, it only remains that we should mention the doings of Drouet,

now left once more with his two old divisions to form the ‘corps of

observation’ opposite the Anglo-Portuguese. Soult during his retreat

had dropped his lieutenant at Llerena, with orders to give back on

Seville without fighting any serious action, if the enemy should

pursue him in force, but if he were left alone to hold his ground,

push his cavalry forward, and keep a strong detachment as near the

Upper Guadiana as possible. For only by placing troops at Campanario,

Medellin, and (if possible) Merida, could communication be kept up via

Truxillo and Almaraz with the Army of Portugal.

As it turned out, Drouet was not to be permitted to occupy such a

forward position as Soult would have liked. He was closely followed

by Stapleton Cotton, with Le Marchant’s and Slade’s heavy and

Ponsonby’s[308] light cavalry brigades, who brought his rearguard

to action at Villagarcia outside Llerena on April 11th. This was

a considerable fight. Drouet’s horse was in position to cover the

retirement of his infantry, with Lallemand’s dragoons in first line,

and Perreymond’s hussars and chasseurs in support. Lallemand evidently

thought that he had only Ponsonby’s brigade in front of him, as Le

Marchant’s was coming up by a side-road covered by hills, and Slade’s

was far out of sight to the rear. Accordingly he accepted battle on an

equal front, each side having three regiments in line. But, just as the

charge was delivered, the 5th Dragoon Guards, Le Marchant’s leading

regiment, came on the ground from the right, and, rapidly deploying,

took the French line in flank and completely rolled it up[309]. The

enemy went to the rear in confusion, and the pursuit was continued

till, half-way between Villagarcia and Llerena, the French rallied on

their reserve (2nd Hussars) behind a broad ditch. Cotton, who had not

let his men get out of hand, re-formed Anson’s brigade and delivered

a second successful charge, which drove the French in upon Drouet’s

infantry, which was in order of battle to the left of Llerena town. It

was impossible to do more, as three cavalry brigades could not attack

12,000 men of all arms in a good position. But a few hours later the

whole French corps was seen in retreat eastward: it retired to Berlanga

and Azuaga on the watershed of the Sierra Morena, completely abandoning

Estremadura.

[308] This officer was in command of the brigade of Anson, then

absent on leave, which at this time consisted of the 12th, 14th,

and 16th Light Dragoons.

[309] There is a good account of all this in the admirable diary

of Tomlinson of the 16th, which I so often have had to cite. He

has an interesting note that the 16th in their charge found a

stone wall in their way, and that the whole regiment took it in

their stride, and continued their advance in perfect order (p.

150).

The French (outnumbered, if Slade’s brigade be counted, but it was far

to the rear and never put in line) lost 53 killed and wounded and 4

officers and 132 rank and file taken prisoners. Cotton’s casualties

were 14 killed and 2 officers and 35 men wounded: he insisted that

his success would have been much greater if Ponsonby had held back a

little longer, till the whole of Le Marchant’s squadrons came on the

field--Lallemand would then have been cut off from Llerena and his line

of retreat, and the greater part of his brigade ought to have been

captured, though the light cavalry in the second line might have got

off[310]. However, the affair was very creditable to all concerned.

[310] Soult only acknowledges a loss of three officers and

about 110 men in his dispatch of April 21 to Berthier, adding

the ridiculous statement that the British had 100 killed and

many more wounded, and that the 5th Dragoon Guards had been

practically destroyed. Martinien’s tables show four French

officers wounded and one killed, but (of course) take no account

of unwounded prisoners. The British lost two missing, men who had

ridden ahead in the pursuit into the French infantry.

Hill’s infantry did not follow the retreating French, and had halted

about Almendralejo and Villafranca, only the cavalry having gone on in

pursuit to Llerena. The rest of the Anglo-Portuguese army was already

in movement for the North, as Wellington had given up the idea, which

had somewhat tempted him at first, of pursuing Soult to Seville and

trying to upset the whole fabric of French power in Andalusia. Of this

more in its due place. Suffice it to say here that he fell back on his

old partition of forces, leaving Hill in Estremadura as his ‘corps of

observation’, with precisely the same force that he had been given in

1811, save that one British cavalry brigade (that of Slade) was added.

The rest of the corps consisted of the 2nd Division, Hamilton’s two

Portuguese brigades, Long’s British and John Campbell’s Portuguese

horse[311]. The whole amounted to about 14,000 men, sufficient not

only to hold Drouet in check, but also to keep an eye upon the French

troops in the valley of the Tagus, against whom Wellington was now

meditating a raid of the sort that he had already sketched out in his

correspondence with Hill in February.

[311] This was the brigade formerly under Barbaçena, 4th and 10th

regiments.

So much for the Army of Andalusia and its fortunes in April 1812. We

must now turn to those of Marmont and the Army of Portugal during the

same critical weeks.

The Duke of Ragusa, as it will be remembered, had been caught at

Salamanca, on March 27th, by Napoleon’s dispatch giving him an

over-late option of detaching troops to the relief of Badajoz. But

being already committed to the invasion of Portugal prescribed by

the Emperor’s earlier letters, and having his field-force and his

magazines disposed for that project, he had resolved to proceed with

it, though he had no great belief in the results that would follow

from his taking the offensive[312]. As he informed his master, there

was nothing at which he could strike effectively. ‘It would seem that

His Majesty thinks that Lord Wellington has magazines close behind the

frontier of northern Portugal. Not so. These magazines are at Abrantes,

or in Estremadura. His hospitals are at Lisbon, Castello Branco, and

Abrantes. There is nothing of any importance to him on the Coa.’ And

how was Almeida or Ciudad Rodrigo to be assailed in such a way as to

cause Wellington any disquietude, when the Army of Portugal had not a

single heavy gun left? ‘General Dorsenne had the happy idea of leaving

in Rodrigo, a fortress of inferior character on the front of our line,

the whole siege-train prepared for this army at great expense, so that

new guns of large calibre must actually be brought up from France.’

[312] Mes dispositions étant faites pour une marche de quinze

jours sur l’Agueda, déjà commencée, je continue ce mouvement,

sans cependant (je le répète) avoir une très grande confiance

dans les résultats qu’il doit donner.’ Marmont to Berthier, March

27.

Marmont’s striking force was not so large as he would have wished.

Bonnet was, by the Emperor’s orders, beginning his advance for the

reoccupation of the Asturias. Foy was in the valley of the Tagus.

Souham had to be left on the Esla, to observe the Army of Galicia.

This left five divisions for active operations: but the Marshal came

to the conclusion that he must split up one more (Ferey’s) to hold

Valladolid, Salamanca, Zamora, Toro, Avila, Benavente, and other

places, which in an elaborate calculation sent to Berthier he showed

to require 4,910 men for their garrisons. He therefore marched with

four infantry divisions only [Clausel, Maucune, Sarrut, Brennier]

and 1,500 light cavalry, about 25,000 men in all: his division of

dragoons was left behind in Leon, to keep open communication between

his various garrisons. A rather illusory help was sought by sending to

Foy, who then lay at Almaraz, orders to the effect that he might push

a detachment to Plasencia, and give out that he was about to join the

main army by the pass of Perales. But Foy’s real concern, as he was

told, was to keep up communication with the Army of the South, and to

give any help that was possible on the side of Truxillo, if (by some

improbable chance) the Army of the Centre should be able to lend him

the aid of any appreciable number of battalions.

On the 30th the French army appeared in front of Rodrigo, and Carlos

de España, leaving 3,000 men as garrison there, under General Vives,

retired with the small remainder of his division towards the Portuguese

frontier. He was pursued and molested by the enemy’s cavalry, not

having been covered or assisted, as Wellington had directed, by Victor

Alten’s regiment of German Hussars. That officer, neglecting his

orders in the most flagrant fashion, did not retire slowly and in a

fighting posture, when the French drove in his line of vedettes in

front of Rodrigo, but collected his regiment and rode hard for Castello

Branco, without concerning himself in the least as to the safety of the

Spanish and Portuguese forces in his neighbourhood, or the procuring of

intelligence as to the strength and the purpose of the French army. His

carelessness or shirking of responsibility, which was to be displayed

in still worse form as the campaign went on, drew on him such a sharp

and bitter rebuke from Wellington that it is a wonder that he was not

sent home forthwith[313].

[313] Wellington to V. Alten, April 18, ‘You were desired “not to

be in a hurry,” to give them (España and General Baccelar) your

countenance so far as might be in your power, and to tell them

that you were left in the front for a particular object.... I

beg you to observe that if you had assembled the 1st Hussars at

Pastores on March 30 and April 1, the Agueda being then scarcely

fordable for cavalry, you could have kept open the communications

between Almeida and Ciudad Rodrigo.... You wrote on the seventh

from Castello Branco that you knew nothing about the enemy! and

instead of receiving from you (as I had expected) a daily account

of their operations, you knew nothing, and, from the way in which

you made your march, all those were driven off the road who might

have given me intelligence, and were destined to keep up the

communication between me and Carlos de España.’

Marmont looked at Rodrigo, but refused to attempt anything against

it, though he was informed that the garrison was undisciplined and

dispirited. Without siege artillery he held that it was useless

to attack the place. After sending in a formal summons to Vives

(who gave the proper negative answer in round terms), and throwing

into the streets a few shells from the howitzers attached to his

field-batteries, he told off Brennier’s division to blockade Rodrigo,

as also to guard a flying bridge which he cast across the Agueda at La

Caridad, a few miles up-stream.

His next move was to send forward Clausel with two divisions to

investigate the state of Almeida. He had heard that its walls were

unfinished, and thought that there might be some chance of executing a

_coup-de-main_ against it. The general, however, came back next day,

reporting that he thought the scheme impossible. He had apparently been

deterred from pressing in upon the place both by the defiant attitude

of the governor, Le Mesurier, whose outposts skirmished outside the

walls for some time before allowing themselves to be driven in, and

still more by the sight of a considerable force of Portuguese troops

encamped close to the town on the other side of the Coa. This was

Trant’s militia, the first detachment that had got to the front of the

various bodies of troops which Wellington had told off for the defence

of the Beira. They had taken up the strong position behind the bridge

of the Coa, which Craufurd had so obstinately defended against Ney in

July 1810.

On the alarm being given on March 29th that Marmont was marching

against that province, and not against Galicia or the Tras-os-Montes,

Wellington’s orders suiting that contingency were carried out with more

or less accuracy. Silveira, with the Tras-os-Montes militia and his

small body of regular cavalry, began to move on Lamego, where Baccelar,

the chief commander in the North, had concentrated the regiments from

the Oporto region and the Beira Alta, even before Marmont had left

Salamanca. General Abadia had been requested to press forward against

the French on the Esla, so as to threaten the flank and rear of the

invading army. He did not accomplish much, being convinced that the

forces left opposite him were too strong to be lightly meddled with.

But he directed a raid to be made from the Western Asturias towards

the city of Leon, and the division at Puebla de Senabria threatened

Benavente. Both movements were executed too late to be of any

importance in affecting the course of the campaign.

Baccelar had been ordered to avoid committing himself to a general

action with any large body of the enemy, but to show such a mass of

troops concentrated that Marmont would have to keep his main body

together, and to act cautiously on the offensive. His primary duty was

to cover, if possible, the large magazines at São João de Pesqueira and

Lamego on the Douro, and the smaller ones at Villa da Ponte, Pinhel,

and Celorico. To these Wellington attached much importance, as they

were the intermediate dépôts from which his army drew its sustenance

when it was on the northern frontier, and he knew that he would be

requiring them again ere many weeks had passed. As long as Marmont

remained near Almeida, it was necessary to keep a force as far forward

as possible, behind the very defensible line of the Coa, and Trant

was advanced for this purpose, though he was directed not to commit

himself. His presence so close to Almeida was very valuable, as he

would have to be driven off before the Marshal formally invested the

place. Le Mesurier, the governor, was not at all comfortable as to

his position: though he had a proportion of British artillery left

with him, the whole of the infantry of the garrison consisted of Beira

militia, who had no experience under arms. On taking over charge of the

place, on March 18, the governor had complained that though the walls

were in a sufficient state of repair, and there were plenty of guns

forthcoming, yet few or none of them were mounted ready for service,

the powder magazines were insufficiently sheltered, and many details

of fortification (palisades, platforms, &c.) had to be completed in a

hurry[314]. However, the place looked so sound for defence when Clausel

reconnoitred it, that--as we have seen--he made no attempt to invest

it, and promptly withdrew, reporting to his chief that Almeida was not

to be taken by a _coup-de-main_.

[314] For complaints by Le Mesurier as to the defects of the

place when he took over charge of it on March 18, see his letter

of the 28th of the same month, to Wellington, in the Appendix to

Napier, iv. pp. 450-1.

Marmont then made the move which Wellington had most desired, and

which in his dispatch to Baccelar he had specified as the happiest

thing that could come about. Instead of sitting down before Almeida or

Ciudad Rodrigo, or making a push against the dépôts on the Douro, he

turned southward towards the Lower Beira, and (leaving Brennier behind

to guard communications) marched with three divisions to Sabugal via

Fuente Guinaldo. This policy could have no great results--the Marshal

might ravage the country-side, but such a movement with such a force

could not possibly alarm Wellington overmuch, or draw him away from

the siege of Badajoz if he were determined to persevere in it. There

was nothing of importance to him in central Beira--only minor dépôts

at Celorico and Castello Branco, much less valuable than the larger

ones at Lamego and São João de Pesqueira on the Douro. ‘He can do no

more,’ as an acute observer on the Portuguese staff remarked, ‘than

drive off some cattle, burn some cottages, and ruin a few wretched

peasants[315].’ For the country about the sources of the Zezere and

round Castello Branco is one of the most thinly peopled districts of

Portugal.

[315] The observation comes from D’Urban’s unpublished Journal.

To meet Marmont’s southern move Baccelar brought up Trant’s and

Wilson’s militia by a parallel march to Guarda, while Le Cor, with

the two regiments of the Beira Baixa, held on at Castello Branco till

he should be evicted from it. To Wellington’s intense disgust[316],

Victor Alten, whose orders directed him to fall back no farther than

that town, continued his precipitate retreat with the German Hussars

to the bridge of Villa Velha on the Tagus, and began to take measures

to destroy that all-important link of communications between north and

south. Fortunately he was stopped before he had done the damage. The

bridge was only taken over to the south bank, not committed to the

flames.

[316] Wellington to Alten, _Dispatches_, ix. p. 69. ‘You were

positively ordered by your instructions to go to Castello Branco

and no farther. The reason for this instruction was obvious.

First the militia of Lower Beira would be there in the case

supposed [that of Marmont’s making an invasion south of the

Douro], and they _were_ there. Secondly, as soon as I should

be informed of the enemy’s approach to the Coa, it would be

necessary for me to assemble a force at Castello Branco--of

which the foundation would be the 1st Hussars K.G.L. Yet

notwithstanding my orders you marched from Castello Branco on

the 8th, and crossed the Tagus on the 9th. Till I received your

letter I did not conceive it possible that you could so far

disregard your instructions.’

Halting at Sabugal, on April 8th, Marmont sent out flying columns,

which ravaged the country-side as far as Penamacor, Fundão, and

Covilhão, and dispatched Clausel with a whole division against Castello

Branco, the one important place in the whole region. Le Cor evacuated

it on April 12th, after burning such of the magazines as could not be

removed in haste: and Clausel--who occupied it for two days--did not

therefore get possession of the stores of food which his chief had

hoped to find there. In revenge the town and the small proportion of

its inhabitants who did not take to the hills were badly maltreated:

many buildings, including the bishop’s palace, were burnt.

Hearing that Marmont had dispersed the larger portion of his army with

flying columns, and was lying at Sabugal, on the 12th, with only a few

thousand men, Trant conceived the rash idea that it would be possible

to surprise him, at his head-quarters, by a night march of his own

and Wilson’s combined divisions from Guarda. The distance was about

twenty miles over mountain roads, and the scheme must have led to

disaster, for--contrary to the information which the militia generals

had gathered--the Marshal’s concentrated main body was still stronger

than their own, despite of all his detachments[317]. ‘You could not

have succeeded in your attempt, and you would have lost your division

and that of General Wilson[318],’ wrote Wellington to Trant, when

the scheme and its failure were reported to him a week later. It was

fortunately never tried, owing to Baccelar’s having made objections to

his subordinate’s hare-brained plan.

[317] I cannot resist quoting here, as an example of Trant’s

over-daring and reckless temperament, his letter to Wilson,

urging him to co-operate in the raid, which was lent me by

Wilson’s representative of to-day:--

GUARDA, 11th _April_, 1812.

MY DEAR WILSON,--I arrived last night. Hasten up your division:

there never was a finer opportunity of destroying a French corps,

in other words and in my opinion, their 2nd Division: but I have

no certainty of what force is the enemy. At any rate send me

your squadron of cavalry, or even _twenty_ dragoons. I am very

ill-treated by Baccelar in regard to cavalry. Push on yourself

personally. You know how happy I shall be in having you once more

as the partner of my operations. Order up everything you can from

Celorico to eat: here there is _nothing_.--Yrs. N. T.

The French 2nd Division was Clausel’s, as it chanced, the one

that was precisely _not_ at Sabugal, but executing the raid on

Castello Branco.

[318] Wellington to Trant, _Dispatches_, ix. p. 73.

But the best comment on the enterprise is that on the very night

(April 13-14) which Trant had fixed for his march, he was himself

surprised by Marmont, so bad had been his arrangements for watching the

country-side. The Marshal had learnt that there was an accumulation

of militia at Guarda threatening his flank, and resolved to give it

a lesson. He started with a brigade each from Sarrut’s and Maucune’s

divisions and five squadrons of light cavalry--about 7,000 men--and

was, at dawn, on the 14th, at the foot of the hill of Guarda, where

he had the good luck to cut off all Trant’s outposts without their

firing a shot--so badly did the militia keep their look-out. ‘Had he

only dashed headlong into the town he might have captured Wilson’s and

my divisions without losing probably a single man,’ wrote Trant. But

the ascent into Guarda was long and steep, and Marmont, who had only

cavalry up, did not guess how careless were his adversaries. He took

proper military precautions and waited for his infantry: meanwhile the

Portuguese were roused, almost by chance as it seems. ‘My distrust of

the militia with regard to the execution of precautions,’ continues

Trant, ‘had induced me at all times to have a drummer at my bedroom

door, in readiness to beat to arms. This was most fortunately the case

on the night of April 13, 1812, for the first intimation that I had of

the enemy being near at hand was given me by my servant, on bringing me

my coffee at daybreak on the 14th. He said that there was such a report

in the street, and that the soldiers were assembling at the alarm

rendezvous. I instantly beat to arms, and the beat being as instantly

taken up by every drummer in the place, Marmont, who was at that very

moment with his cavalry at the entrance of the town, held back. I was

myself the first man out of the town, and he was not then 400 yards

away[319].’

[319] Narrative of Trant in Napier’s Appendix to vol. iv. p. 451.

The Marshal, in his account of the affair, says that the Portuguese

formed up on the heights by the town, apparently ready to fight, but

drew off rapidly so soon as he had prepared for a regular attack on

the position. Wise not quite in time, the two militia generals sent

their men at a trot down the steep road at the back of the place, with

the single troop of regular dragoons that they possessed bringing

up the rear. It had now begun to rain in torrents, and Trant and

Wilson having obtained two or three miles start, and being able to

see no distance owing to the downpour, thought that they had got

off safe. This was not the case: Marmont realized that his infantry

could not catch them, but seeing their hurry and disorder ordered his

cavalry--his own escort-squadron and the 13th Chasseurs--to pursue

and charge the rearguard of the retreating column. They overtook it

by the bridge of Faya, three miles outside Guarda, where the road to

Celorico descends on a steep slope to cross the river. The leading

French squadron scattered the forty dragoons at the tail of Trant’s

division, and rode on, mixed with them, against the rearguard battalion

(that of Oporto). The militiamen, startled and caught utterly by

surprise, tried to form across the road and to open fire: but the rain

had damped their cartridges, and hardly a musket gave fire. Thereupon

the battalion went to pieces, the men nearest the French throwing

down their guns and asking for quarter, while those behind scattered

uphill or downhill from the road, seeking safety on the steep slopes.

The charge swept downhill on to the battalion of Aveiro, and the other

successive units of the Oporto brigade, which broke up in confusion.

Five of their six colours were taken, and 1,500 prisoners were cut off,

while some tumbled into the Mondego and were drowned, by losing their

footing on the steep hillside. Hardly a Frenchman fell, and not very

many Portuguese, for the _chasseurs_, finding that they had to deal

with helpless militiamen who made no resistance, were sparing with

the sabre[320]. The greater part of the prisoners were allowed, in

contempt, to make off, and only a few hundred and the five flags were

brought back to Marmont at Guarda. The pursuit did not penetrate so

far as Wilson’s division, which got across the Mondego while Trant’s

was being routed, and formed up behind the narrow bridge, where the

_chasseurs_, being a trifling force of 400 men, did not think fit to

attack them. The French infantry had marched over twenty miles already

that day, and were dead beat: Marmont did not send them down from

Guarda to pursue, in spite of the brilliant success of his cavalry.

[320] There is an account of this rout from the French side in

the _Mémoires_ of Parquin, of the 13th Chasseurs, an officer

mentioned in Marmont’s dispatch as having taken one of the flags.

Parquin calls it that of the regiment of _Eurillas_. There was

no such corps: those which lost standards were Aveiro, Oliveira,

and Penafiel. A lengthy account may be found also in Beresford’s

_Ordens do Dia_ for May 7, where blame and praise are carefully

distributed, and the curious order is made that the disgraced

regiments are to leave their surviving flags at home, till they

have washed out the stain on their honour by good service in the

field.

The day after the ‘Rout of Guarda’ Marmont pushed an advanced guard

to Lagiosa, half-way to Celorico, where Trant and Wilson had taken

refuge, with their ranks short of some 2,000 men scattered in the

hills. Thereupon the militia generals set fire to the stores, and

evacuated Celorico, falling back into the hills towards Trancoso. But

finding that the French were not coming on, they halted; and when they

ascertained that the enemy was actually returning to Guarda, they came

back, extinguished the fires, and rescued great part of the magazines.

Marmont’s unexpected forbearance was caused by the fact that the news

of the fall of Badajoz reached him on the 15th, along with a report

from Clausel (who had just evacuated Castello Branco) that Wellington’s

army had already started northward, and that its advanced guard was

across the Tagus at Villa Velha.

This was startling, nay appalling, intelligence. Badajoz had been

reckoned good for a much longer resistance, and the news had come

so slowly--it had taken nine days to reach Marmont--that it was

possible that the British army was already in a position to cut off

his expeditionary force from its base on the Agueda. Wherefore Marmont

hastily evacuated Guarda, and was back at Sabugal by the 16th, where

Clausel and the other dispersed fractions of his army joined him. Here

he regarded himself as reasonably safe, but determined to retire behind

the Spanish frontier ere long, raising the blockade of Ciudad Rodrigo.

‘My troops,’ he wrote to Berthier on that day, ‘have used up the little

food to be gathered between the Tagus and the Zezere; and now that

the enemy is on the Tagus I cannot possibly remain on the Mondego, as

I should be leaving him on my line of communications. I shall fall

back to the right bank of the Agueda. If the enemy resolves to pursue

me thither I shall fight him. If not I shall fall back on Salamanca,

because of the absolute impossibility of feeding an army between the

Agueda and the Tormes.’

Marmont remained at Sabugal and its neighbourhood for nearly a week--by

the 22nd he had drawn back a few miles to Fuente Guinaldo--with about

20,000 men. His position was more dangerous than he knew; for on the

18th the heavy rains, which began on the day of the combat of Guarda,

broke his bridge over the Agueda at La Caridad, so that he was cut off

from Brennier and from Salamanca. He was under the impression that

Wellington had only brought up a couple of divisions against him, and

that these were still south of Castello Branco[321], whereas as a

matter of fact seven had marched; and on the day that he wrote this

incautious estimate Wellington’s headquarters were at Penamacor, the

Light and 3rd Divisions were closing in on Sabugal, the 4th and 5th

were a full march north of Castello Branco, and the 1st, 6th, and 7th

were at Losa, quite close to that city. Thirty-six hours more of delay

would have placed Marmont in the terrible position of finding himself

with a broken bridge behind him, and 40,000 enemies closing in upon his

front and flank.

[321] Marmont to Berthier: Fuente Guinaldo, April 22. ‘Les

rapports des prisonniers sont que trois divisions de l’armée

anglaise reviennent sur le Coa. Mais cette nouvelle ayant été

donnée avec affectation par les parlementaires, et n’ayant vu

jamais autre chose que le seul 1er de Hussards Allemands, qui

était précédemment sur cette rive, et point d’infanterie, ni rien

qui annonce la présence d’un corps de troupes, je suis autorisé

à croire que c’est un bruit qu’on a fait courir à dessein, et

qu’il n’y a pas d’Anglais en présence. Je suis à peu près certain

qu’il a parti de Portalègre deux divisions, qui se sont portées

à Villa Velha: mais il me paraît évident qu’elles ne se sont

beaucoup éloignées du Tage.’ The actual situation was 1st Hussars

K.G.L. Quadraseyes in front of Sabugal: Light Division, Sabugal:

3rd Division, Sortelha: 4th Division, Pedrogão, 5th Division,

Alpedrinha; 1st, 6th, 7th Divisions, Losa: Pack’s Portuguese,

Memoa. The map will show what a fearful situation Marmont would

have been in had he halted for another day.

To explain the situation, Wellington’s movements after the capture

of Badajoz must now be detailed. It had been his hope, though not

his expectation, that Soult might have remained at Villafranca after

hearing of the disaster of the 6th April; in this case he had intended

to fall upon him with every available man, crush him by force of

numbers, and then follow up his routed army into Andalusia, where the

whole fabric of French occupation must have crumpled up. But Soult

wisely retreated at a sharp pace; and the idea of following him as far

as Seville, there to find him reinforced for a general action by all

the troops from the Cadiz Lines and Granada, was not so tempting as

that of bringing him to battle in Estremadura. On the day after the

fall of Badajoz Wellington formulated his intentions in a letter to

Lord Liverpool. ‘It would be very desirable that I should have it in

my power to strike a blow against Marshal Soult, before he could be

reinforced.... But it is not very probable that he will risk an action

in the province of Estremadura, which it would not be difficult for

him to avoid; and it is necessary for him that he should return to

Andalusia owing to the movements of General Ballasteros and the Conde

de Penne Villemur ... if he should retire into Andalusia I must return

to Castille[322].’

[322] Wellington to Liverpool, April 7, _Dispatches_, ix. p. 43.

The reason given by Wellington for his resolve to turn north again was

that Carlos de España had informed him that Ciudad Rodrigo, though

otherwise tenable enough, had only provisions for twenty-three

days, partly from what Wellington called the general policy of

‘Mañana’[323]--of shiftless procrastination--partly from the definite

single fact that a very large convoy provided from the British

magazines on the Douro had been stopped at Almeida on March 30th. This,

in Wellington’s estimation, was the fault of Victor Alten, who, if he

had held the outposts beyond the Agueda for a day longer, might have

covered the entry of the convoy into Ciudad Rodrigo[324]. Marmont’s

operations on the Coa and the Agueda would have been quite negligible

from the strategic point of view but for this one fact. He might

ravage as far as Guarda or Castello Branco without doing any practical

harm, but it could not be permitted that he should starve Rodrigo into

surrender: even allowing for a firm resistance by the garrison, and a

judicious resort to lessened rations, the place would be in danger from

the third week of April onward. Wherefore, unless Marmont withdrew into

Spain by the middle of the month, he must be forced to do so, by the

transference of the main body of the Anglo-Portuguese Army to the North.

[323] Wellington to Henry Wellesley, April 4, _Dispatches_, ix.

p. 29.

[324] Wellington to Alten, April 18, _Dispatches_, ix. p. 68,

‘I beg to observe that if you had assembled the 1st Hussars at

Pastores on the 30th March and 1st April ... you would have kept

open the communication between Almeida and Ciudad Rodrigo, and

the convoy would probably have got into the latter place.’

The Marshal, during the critical days following the fall of Badajoz,

showed no such intention. Indeed he advanced to Sabugal on the 8th,

seized Castello Branco on the 12th, and executed his raid on Guarda

upon the 13th-14th. Ignorant of the fall of Badajoz, he was naturally

extending the sphere of his operations, under the belief that no

serious force was in his front. While he was overrunning Beira Baixa,

Ciudad Rodrigo continued to be blockaded by Brennier, and its stores

were now running very low.

On April 11th[325] Wellington made up his mind that this state of

things must be brought to an end, and he determined that no mere

detachment should march, but a force sufficient to overwhelm Marmont

if he could be brought to action. The movement began with the march

of the 11th Light Dragoons and Pack’s and Bradford’s Portuguese to

Elvas on the afternoon of the 11th April, all being ordered to move

on Arronches and Portalegre. On the 12th a larger force started off

from the camps around Badajoz and on the Albuera position: the 3rd

and Light Divisions moved (following Pack and Bradford) on Portalegre

via Arronches, the 4th and 5th, making a shorter move, to Campo Mayor

on the same road, the 7th from Valverde to Elvas. The 1st and 6th

under Graham, bringing up the rear, went off on the 13th from Valverde

and Elvas northward. Orders were sent to Stapleton Cotton, then in

pursuit of Drouet in southern Estremadura, to come with Anson’s and Le

Marchant’s cavalry brigades to join the main army, leaving only Slade’s

and Long’s to Hill. Bock’s Heavy Dragoon brigade of the King’s German

Legion was also directed to take part in the general movement.

[325] The date can be fixed from D’Urban’s Journal: ‘Marmont has

blockaded Rodrigo, reconnoitred Almeida, and has now made an

inroad as far as Fundão: all this obliges a movement toward him.

April 11.’

Only Hill, with the troops that had served under him since the summer

of 1811, plus one new cavalry brigade, was left behind in Estremadura

to ‘contain’ Drouet. It was highly unlikely that Soult would be heard

of in that province, as he had his own troubles in Andalusia to keep

him employed. Indeed Wellington in his parting message to this trusty

lieutenant told him that it was ‘impossible’ that the enemy could

assemble enough troops to incommode him at present, and explained that

his chief duty would be to cover the repairing of Badajoz, into which

three Portuguese line regiments[326] under Power, hitherto forming the

garrisons of Elvas and Abrantes, were thrown, to hold it till Castaños

should provide 3,000 Spaniards for the purpose.

[326] 5th and 17th from Elvas, 22nd from Abrantes.

The movement of the army marching against Marmont was rapid and

continuous, though it might have been even more swift but for the fact

that the whole long column had to pass the bridge of Villa Velha, the

only passage of the Tagus that lay straight on the way to the Lower

Beira: to send troops by Abrantes would have cost too much time. On

the 16th the Light and 3rd Divisions crossed the bridge, on the 17th

some cavalry and Pack’s and Bradford’s Portuguese, while the 4th, 5th,

and 6th Divisions were now close to the river at Castello de Vide and

Alpalhão, and only the 1st was rather to the rear at Portalegre[327].

Alten’s German Hussars, picked up at Castello Branco on the 18th by the

head of the column, were the only cavalry which Wellington showed in

his front. This was done on principle: Marmont knew that this regiment

was in his neighbourhood, and if it pressed in upon his outposts, it

told him nothing as to the arrival of new troops opposite him. As we

have already seen, when quoting one of his dispatches[328], he drew the

inference that Wellington intended, and so late as the 22nd believed

that his adversary’s main army was still behind the Tagus, and that at

most two divisions had come up to Villa Velha--but probably no further.

[327] All these movements are taken from the elaborate tables in

D’Urban’s Journal for these days.

[328] See above, p. 288.

Steadily advancing, the column, with the 3rd and Light Divisions

leading, reached Castello Branco on the 17th. They found that it had

been reoccupied on the 15th by Alten’s Hussars and Le Cor’s militia;

but it was in a dreadful state of dilapidation owing to the ravages

of Clausel’s troops during the two days of their flying visit. Clear

information was received that Marmont was still at Sabugal, and his

vedettes lay as far south as Pedrogão. The British staff were in hopes

that he might be caught. ‘His ignorance (as we hope) of the real force

in march against him may end in his destruction,’ wrote D’Urban to

Charles Stewart on the 18th, ‘for he has put the Agueda in his rear,

which the late rains have made impassable: his situation is very

critical. If he discovers his error at once, he may get off by his

left down the Perales road, and so reach Plasencia: but if he does

not, and waits to be _driven_ out of the ground he holds, I don’t see

how he is to get away. Lord Wellington will be all closed up by the

21st; meanwhile he shows little to his front, and avoids giving serious

alarm: the fairest hopes may be entertained of a decisive blow[329].’

[329] Letter in the D’Urban Papers.

It looked indeed as if Marmont was waiting over-long: on the 17th-18th

his exploring parties came as far south as Idanha Nova, where by an ill

chance they captured Wellington’s most famous intelligence-officer,

Major Colquhoun Grant, who there commenced that extraordinary

series of adventures which are told in detail in the life of his

brother-in-law, Dr. McGrigor, Wellington’s chief medical officer. He

escaped at Bayonne, and returned to England via Paris and the boat of a

Breton fisherman[330].

[330] See the _Life of Surgeon-General Sir Jas. McGrigor_, pp.

284-96. I have before me, among the Scovell papers, Grant’s

original signed parole as far as Bayonne, witnessed by General

Lamartinière, the chief of Marmont’s staff. It was captured by

_Guerrilleros_ in Castile, and sent to Wellington. Accompanying

it is the General’s private letter, commending Grant to the

attention of the French police, with the explanation that he was

only not treated as a spy because he was captured in British

uniform, though far in the rear of the French outpost line.

The rear of the column had dropped behind somewhat, owing to the

incessant rains which had set in from April 14th, and which had broken

Marmont’s bridge four days later. Wellington had given the 4th Division

leave to halt for a day, because of the state of the roads and the

entire want of cover for the night in the desolate tract between

Villa Velha and Abrantes[331]. It reached Castello Branco, however,

on the 20th, on which day only (by some extraordinary mismanagement)

Wellington got the tardy news of Trant’s disaster at Guarda on the

morning of the 14th. And this news was brought not by any official

messenger, but by a fugitive ensign of militia, who garnished it with

all manner of untrue additions--whereupon Beresford had him tried

and shot, for deserting his troops and spreading false intelligence.

Clearly Trant, Wilson, and Baccelar between them should have got the

true narrative to head-quarters before six days had elapsed.

[331] Wellington to Graham, Castello Branco, April 18,

_Dispatches_, ix. p. 70.

The 21st April was the critical day of this campaign. Marmont was

still at Fuente Guinaldo, on the wrong side of the Agueda, and his

bridge at La Caridad was still broken and not relaid. Though unaware

that Wellington was close upon him with an overwhelming force, whose

existence he denied (as we have seen) in a letter sent off so late as

the 22nd, he was yet feeling uncomfortable, both because of his broken

communications, and because he had used up his food. Wherefore he gave

orders that his artillery, using very bad side-roads, should pass the

Agueda by the bridge of Villarubia, a small mountain crossing quite

near its source, which would take it, not by the ordinary route past

Ciudad Rodrigo, but by Robledo to Tamames, through a very difficult

country.[332] He himself with the infantry stood fast on the 21st and

22nd, unaware of his dangerous position.

[332] Marmont to Berthier, Fuente Guinaldo, April 22 [original

intercepted dispatch in Scovell Papers]: ‘J’ai eu la plus grande

peine à faire arriver mon artillerie sur la rive droite de cette

rivière. Les ponts que j’avais fait construire sur l’Agueda

ayant été détruits par les grandes crues d’eau, et n’ayant pas

la faculté de les rétablir, je n’ai su d’autre moyen que de la

diriger par les sources de cette rivière, et les contreforts

des montagnes.’ The wording of Wellington’s intercepted copy

differs slightly from that of the duplicate printed in Ducasse’s

_Correspondence of King Joseph_, viii. pp. 404-10.

For the allies were closing in upon him--the head-quarters of

Wellington were on the 21st at Pedrogão, the 1st German Hussars,

covering the advance, had reached Sabugal, and the Light and 3rd

Divisions were close behind, as were Pack’s and Bradford’s Portuguese,

while the 4th and 5th were both beyond Castello Branco. On the morning

of the 22nd the head of the infantry column had passed Sabugal, and the

Hussars were in front of them, pushing in Marmont’s vedettes. A delay

of twenty-four hours more on the part of the French would have brought

the armies into collision, when Marmont gave orders for his infantry

to retreat across the Agueda by the fords near Ciudad Rodrigo, where

the water on that day had at last fallen enough to render the passage

possible, though difficult and dangerous. The leading division marched

on the 22nd, the rest on the 23rd: by the night of the latter day all

were across the river, and retiring rapidly on Salamanca; for, as

Marmont truly observed, there was not a ration of food to be got out of

the devastated country between Rodrigo and the Tormes.

The odd part of this sudden, if long-deferred, retreat was that it

was made without the slightest knowledge that it was imperative,

owing to Wellington’s near approach; in the letter announcing it

to Berthier the Marshal reiterates his statement that he does not

believe that Wellington has a man north of Castello Branco save the

1st Hussars K.G.L. The retreat is only ordered because it is clear

that, with 20,000 men only in hand, it is useless to continue the

tour of devastation in the Beira. ‘Your highness may judge that the

result of the diversion which I have sought to make in favour of the

Army of the South has been practically nil. Such a movement could only

be effective if carried out with a force great enough to enable me

to march against the enemy with confidence, and to offer him battle,

even if he had every available man collected. With 18,000 or 19,000

men (reduced to 15,000 or 16,000 because I have to leave detachments

to keep up communications) I could not move far into Portugal without

risk, even if I have no one in front of me, and the whole hostile

army is on the farther bank of the Tagus. For if I passed the Zezere

and marched on Santarem, the enemy--master of Badajoz and covered by

the Guadiana--could pass the Tagus behind me, and seize the defiles

of Zarza Major, Perales, and Payo, by which alone I could return....

There are several places at which he could cross the Tagus, above and

below Alcantara, and so place himself by a rapid and secret movement

that my first news of him would be by the sound of cannon on my line of

communications--and my position would then be desperate[333].’

[333] Intercepted dispatch in the Scovell Papers, Fuente

Guinaldo, April 22, quoted above.

The real danger that was threatening him, on the day that he wrote this

dispatch, Marmont did not suspect in the least, indeed he denied its

existence. But he moved just in time, and was across the Agueda when,

on the 24th, Wellington had his head-quarters at Alfayates, and three

divisions at Fuente Guinaldo, which the French had only evacuated on

the preceding day, with three more close behind. Only the 1st and 6th,

under Graham, were still at Castello Branco and Losa. Evidently if the

fords of the Agueda had remained impassable for another twenty-four

hours, Marmont’s four divisions would have been overwhelmed by superior

numbers and driven against the bridgeless river, over which there would

have been no escape. As it was, he avoided an unsuspected danger, and

returned to Salamanca with his army little reduced in numbers, but with

his cavalry and artillery almost ruined: his dispatch of the 22nd says

that he has lost 1,500 horses, and that as many more needed a long rest

if they were ever again to be fit for service.

On the 24th Wellington bade all his army halt, the forced marches

which they had been carrying out for the last ten days having failed

to achieve the end of surprising and overwhelming Marmont, who had

obtained an undeserved escape. On the 26th he paid a flying visit to

Ciudad Rodrigo, whose safety he had at least secured, and commended

General Vives for his correct attitude during the three weeks of

the late blockade. The next movements of the allied army belong to

a different series of operations, and must be dealt with in a new

section.

SECTION XXXIII

THE SALAMANCA CAMPAIGN

CHAPTER I

KING JOSEPH AS COMMANDER-IN-CHIEF

On March 16, 1812, the day on which Wellington opened his trenches

before Badajoz, the Emperor Napoleon took a step of no small importance

with regard to the control of his armies in Spain. He had now made up

his mind that the long-threatened war with Russia must begin within a

few months, and that he must leave Paris ere long, and move forward

to some central point in Germany, from which he could superintend the

preparations for a campaign, the greatest in scale of any which he

had hitherto undertaken. He was persuaded that war was inevitable:

the Czar Alexander had dared to dispute his will; and in the state of

megalomania, to which his mind had now accustomed itself, he could

tolerate no opposition. Yet he was aware, in his more lucid moments,

that he was taking a great risk. On March 7th Colonel Jardet, Marmont’s

confidential aide-de-camp, was granted an interview, in which he

set forth all the difficulties of the Army of Portugal. The Emperor

heard him out, and began, ‘Marmont complains that he is short of many

resources--food, money, means, &c.... Well, here am I, about to plunge

with an immense army into the heart of a great country which produces

_absolutely nothing_.’ And then he stopped, and after a long silence

seemed suddenly to rouse himself from a sombre reverie, and looking

the colonel in the face asked, ‘How will it all end?’ Jardet, thrown

off his balance by such a searching query, stammered that it would of

course end in the best possible fashion. But he went out filled with

gloomy forebodings, inspired by his master’s evident lack of confidence

in the future[334].

[334] See Marmont’s _Mémoires_, iv. p. 202. Jardet’s long report

to Marmont was captured on its journey out to Salamanca from

Paris, and lies among the Scovell Papers.

Some weeks were yet to elapse before the Emperor’s actual departure

from France; but, ere he went, he had to set in good working order the

conduct of his policy during his absence, and of all its complicated

machinery the Spanish section was one of the most puzzling and the

most apt to get out of order. It was clearly impossible that he should

continue to send from Dresden or Wilna elaborate orders every five

or ten days, as he had been wont to do from Paris. If it took three

weeks to get an order to Seville in February, it might take five or

six in July, when the imperial head-quarters might be in some obscure

Lithuanian hamlet. Something must be done to solve the problem of

continuous policy, and of co-operation between the five armies of

Spain, and after much consideration the Emperor dictated to Berthier

the solution which he thought least bad--’Send by special messenger a

dispatch to the King of Spain, informing him that I confide to him the

command of all my Spanish armies, and that Marshal Jourdan will serve

as his Chief-of-the-Staff. You will send, at the same time, a similar

intimation to that marshal. You will inform the King that I shall keep

him advised of my political intentions through my ambassador at Madrid.

You will write to Marshal Suchet, to the Duke of Dalmatia, and the

Duke of Ragusa that I have entrusted the King with the charge of all

my armies in his realm, and that they will have to conform to all the

orders which they may receive from the King, to secure the co-operation

of their armies. You will write, in particular, to the Duke of Ragusa

that the necessity for obtaining common action between the Armies of

the South, of Valencia, and of Portugal, has determined me to give

the King of Spain control over all of them, and that he will have to

regulate his operations by the instructions which he will receive.

To-morrow you will write in greater detail to the King, but the special

messenger must start this very night for Bayonne[335].’

[335] King Joseph had been prepared for the formal proposal by a

tentative letter sent off to him about three weeks earlier, on

February 19, inquiring whether it would suit him to have Jourdan

as his Chief-of-the-Staff, supposing that the Emperor went off to

Russia and turned over the command in Spain to him. See Ducasse’s

_Correspondence_, ix. p. 322.

Of the bundle of dispatches that for the King was delivered at Madrid

on March 28th, after twelve days of travel. Marmont got his a little

later, as he had started on his Portuguese expedition when it reached

Salamanca. Communication between his field-force and his base being

difficult, owing to the activity of Julian Sanchez, it appears to have

been on March 30, when before Ciudad Rodrigo, that he became aware

that he had a new commander-in-chief[336]. Soult was apprised of the

situation much later, because, when preparing for his expedition to

relieve Badajoz, he had ordered his posts in the Sierra Morena to be

evacuated, and the communication with La Mancha to be broken off for

the moment. It seems that he must have got Berthier’s dispatch quite

late in April, as on the 17th of that month he was only acknowledging

Paris letters of February 23rd[337], and the first courier from Madrid

got through only some time later. Suchet would appear also to have been

advised of the change of command very late--he published the imperial

decree in his official gazette at Valencia only on May 10, giving as

its date the 29th instead of the 16th of March[338], which looks as

if the first copy sent to him had miscarried, and the repetition made

thirteen days later had alone reached him. These dates are only worth

giving as illustrations of the extreme difficulty of getting orders

from point to point in Spain during the French occupation, even when

Andalusia and Valencia were supposed to be thoroughly subdued.

[336] This is proved by Berthier’s letter to King Joseph of April

16 (Ducasse’s _Correspondence of King Joseph_, viii. p. 382),

which says that he has just received Marmont’s dispatch of March

30 acknowledging his own of March 16, and that the Marshal now

knows that he must obey orders from Madrid.

[337] Soult to Berthier from Seville, April 17.

[338] A copy of this print is among the Scovell Papers: it does

credit to the Valencian press by its neat appearance.

It will be noted that in Napoleon’s instructions to Berthier no mention

is made of either the Army of Catalonia or the Army of the North[339];

and it might have been thought that, clinging to the theory of his

paper annexation of Spain north of the Ebro, he was deliberately

exempting from the King’s control the troops in the districts on which

he had resolved to lay hands for his own benefit. But a supplementary

dispatch of April 23rd placed Decaen and the garrison of Catalonia

under the general charge of Suchet, and as that marshal had been

directed to obey King Joseph’s military instructions, the four new

‘French’ departments on the Ebro were now theoretically under the same

general command as the rest of Spain. As to the Army of the North,

Dorsenne wrote (April 19th), with evident glee, to say that he was

exempted from obedience to the King, by not being included in the list

of recipients of the dispatch of March 16, and that he regretted his

inability to carry out a series of orders which Jourdan had sent him.

But he had not many more days to serve in his present capacity, and his

successor, Caffarelli, though equally recalcitrant in spirit, presently

received a formal notice that he was under King Joseph’s command.

[339] The question about the Army of the North is a very curious

one. The authorized copy of the dispatch of May 16, printed in

Napoleon’s correspondence and in Ducasse’s _Correspondence of

King Joseph_, certainly omits its name. But the King declared

that in his original copy of it Dorsenne and his army were

mentioned as put under his charge. In one of the intercepted

dispatches in the Scovell Papers, Joseph writes angrily to

Berthier, giving what purports to be a verbatim duplicate of the

document, and in this duplicate, which lies before my eyes as I

write this, the Army of the North _is_ cited with the rest.

Napoleon’s general policy in placing the supreme control of all the

Spanish armies in the hands of one chief, and bringing to an end (in

theory at least) the system of separate viceroyalties was undoubtedly

the right one. And it cannot be disputed that one second-rate

commander-in-chief is more effective than four good ones, working each

for his own private and local profit and glory. But in this particular

case the new arrangement was not likely to bring about any great change

for the better, owing to the personal equation. During the last three

years Napoleon had been inflicting affronts at short intervals upon his

brother, had annexed integral portions of his realm, had disregarded

most of his complaints and suggestions, and had allowed him to become

the butt of the viceroys, whose insults and injuries he had never

been allowed to resent. They had raided the districts assigned to his

personal governance[340], had plundered his magazines, imprisoned his

officials, and set up courts of justice of their own to supersede the

regular magistracy of the land. The Emperor had never punished such

proceedings; at the most he had ordered that they should cease, when

they were injurious to the progress of the French arms in Spain. It

was useless to issue a sudden order that for the future the marshals

were under Joseph’s control, and that ‘he must make them obey him,’

as the phrase ran in one letter to Madrid. As the King’s minister,

Miot de Melito, wrote, ‘What chance was there of success when all

the individuals concerned were at variance with each other? The

marshals had been accustomed for three years to absolute independence.

The new Chief-of-the-Staff, in spite of his acknowledged capacity,

was known to be out of favour with the Emperor, and in consequence

could exercise no moral authority over the masters of the armies.

The apparent testimonial of confidence which was given to the King,

by making him Commander-in-Chief, was a matter to cause disquietude

rather than satisfaction[341].’ The plain fact was that Napoleon was

over-busy, worried with other problems, and he merely took the easiest

and simplest method of throwing the burden of the Spanish war on to the

shoulders of another. The consequences, be they what they might be,

were now of little importance, compared with the success or failure of

the impending Russian campaign.

[340] One of Marmont’s colonels in the province of Segovia was

at this moment threatening to use armed force against the King’s

troops for resisting his requisitions. See Miot, iii. p. 222.

[341] See Miot de Melito’s _Mémoires_, iii. p. 215.

Jourdan sums up the situation in much the same terms. ‘The King

for two years had been allowed to have no direct relations with

the generals-in-chief: he had no exact knowledge of the military

situation in each of their spheres of command, nor was he better

informed as to the strength, organization, and distribution of the

troops under their orders. Unable to use his new authority till he had

got together detailed statements as to these data, he directed his

chief-of-the-staff to ask for reports. Dorsenne replied that he should

not send any at present, because Berthier, when announcing to him that

the Armies of the South, of Portugal, and of Aragon had been put under

the King’s orders, had informed him that the Emperor would let him know

in due course what was to be done with the Army of the North. Marshal

Suchet demonstrated that he had received special instructions from the

Emperor, which presently were seen to make the King’s authority over

the Army of Aragon quite illusory. Soult had removed all the posts

on the lines of communication when he marched to relieve Badajoz, and

showed so little zeal in reopening them, that even in May it was not

known at Madrid whether he was yet aware that he was under the King’s

orders. Marmont was the only one who sent without delay the report

which had been asked for--but he announced at the same time that, in

obedience to the Emperor’s earlier orders, he was already operating

beyond the Agueda, to make a diversion for the relief of Badajoz[342].’

[342] Jourdan’s _Mémoires_, p. 384.

Of what use was it to send orders to the marshals, when they

could plead that the execution of them was rendered impossible by

instructions received directly from the Emperor, which prescribed a

different policy? Unfortunately for King Joseph each commander-in-chief

still preserved his direct communication with the Minister of War at

Paris: even after the Emperor had started for Poland in May, each

continued to send in his own plans, and to demonstrate how far superior

they were to those prescribed by King Joseph. Soult, in particular,

generally commenced a dispatch by demonstrating that the directions

received from Madrid could not possibly be executed, and then produced

an elaborate scheme of his own, which would be beneficial for the Army

of Andalusia, but impracticable for those of Portugal, Valencia, and

the Centre. When his suggestions were rejected, he wrote privately to

Paris, declaring that Joseph and Jourdan were absolutely incapable,

and sometimes adding that the King was trying to serve his private

interests rather than those of his brother and suzerain. It was the

accidental receipt by Joseph of an intercepted letter of Soult’s to the

Minister of War, in which he was accused of absolute treason to the

Emperor, that brought about the final rupture between the King and the

Marshal, and led to the recall of the latter to France[343].

[343] Oddly enough this letter was in duplicate, and while

one copy fell into Joseph’s hands, the other was captured by

guerrilleros and sent to Wellington. The cipher was worked out

by Scovell, and the contents gave Wellington useful information

as to the relations between Soult and the King. See below, pages

530-39.

King Joseph, though liable to fits of depression and despair, was, on

the whole, of a mercurial and self-sufficient temperament. A few weeks

before the receipt of the Emperor’s dispatch granting him the command

of the Spanish armies, all his letters had been full of complaints and

threats of abdication. But the decree of March 16th filled him with a

sudden confidence--at last his military talents should be displayed and

recognized; he would, as his brother desired, ‘make the marshals obey

him;’ for the future the armies should all act together for a single

end, and not be guided by the selfish interests of their leaders. He

accepted the position of Commander-in-Chief with undisguised pleasure,

and proceeded to draw out schemes of his own, with Jourdan as his

adviser in technical matters of military logistics.

It cannot be denied that the ‘_Mémoire_ of May 1812[344],’ in which

Jourdan set forth the situation after the fall of Badajoz, and

the policy which he considered that it demanded, is a document of

much greater merit than might have been expected. It is by far the

best summary of the position of the French power in Spain that was

ever drawn up, and it recognizes with great clearness the two main

limitations of that power, which were (1) that the imperial troops

were an army of occupation rather than a genuine field army, and (2)

that the Napoleonic system, by which hosts were supposed to ‘live

on the countryside,’ might be applicable for a short campaign in

Lombardy or Bavaria, but was impossible for protracted manœuvres in an

exhausted and thinly-peopled land like central Spain. Jourdan’s note

on the _Mémoire_ sums up the situation in a few lines--’Two measures

were indispensable: one was to render the army mobile, by giving it

ample transport, and by establishing large magazines on all lines of

communication: without these all permanent concentration of heavy

forces, and all continuous operations were impossible. The second was

to abandon the deplorable system of occupying as much territory as

possible--of which the real object was double: firstly, to enable the

armies to live on the country-side; secondly, to appear in the eyes of

Europe to be dominant over the whole of Spain.’

[344] Printed whole in Jourdan’s _Mémoires_, pp. 386-94.

The _Mémoire_ itself is worth analysing. Its gist runs as follows:--

‘(1) The recent departure of the Imperial Guards, the Poles, and other

troops, and the lack of any adequate system of transport or magazines,

renders the Imperial Army--though still 230,000 men strong--incapable

of undertaking any offensive operations. The present situation is

exceptionally trying, because of the successes of Wellington, and the

deplorable effect on Spanish public opinion of the recent annexation

[Catalonia], the arbitrary government of the generals, and the famine

which has lately prevailed. The discontent thereby engendered has led

to the enormous increase in the number of the guerrillero bands. It has

also encouraged the government at Cadiz to multiply its levies and its

military energy.

(2) It is not yet certain whether the Emperor intends the Army of the

North to be at the King’s disposal. General Dorsenne refuses to send in

reports or to accept orders. But since its recent reduction in numbers

[by the departure of the Imperial Guard, and the transfer of Souham’s

and Bonnet’s divisions to the Army of Portugal] it is believed that

it has not more than 48,000 men under arms, and it appears to be a

fact that it can do no more than hold down the wide regions committed

to its charge, and guard the line of communications with France. Even

if placed at the King’s disposition, it can furnish no important

reinforcements to other armies. Nevertheless it should be put under his

control, as it might under certain circumstances be called upon to lend

a moderate force for a short time.

(3) As to the Army of Aragon [60,000 men, including the divisions in

Catalonia]: the King was informed that Marshal Suchet was placed under

his command, and that if he needed reinforcements he might draw on the

troops in Valencia. He therefore [during the siege of Badajoz] ordered

the Marshal to send a division to join the Army of the Centre for an

indispensable operation[345]. The Marshal sent a formal declaration in

reply, to the effect that he could not execute this order, and that

he was even about to withdraw from Cuenca the regiment that he had

placed there, as its absence imperilled the safety of Valencia. He

says that the Emperor has placed Catalonia under his charge, and that

he is authorized to employ his whole force for the protection of the

provinces entrusted to him. Apparently, then, the Army of Aragon cannot

co-operate in operations outside its own sphere, and the Marshal’s

special instructions place him in an exceptional position. His

relations with the King consist in a polite exchange of views, not in

the giving and taking of orders--his Majesty’s control over this army

is purely illusory.

[345] i. e. for the collection of troops in the valley of the

Tagus, to join Foy and operate for the relief of Badajoz.

(4) As to the Army of the South, Marshal Soult has about 54,000 men

effective [not including _Juramentados_, &c.]. The Cadiz Lines and the

garrisons pin down a large force to fixed stations. The Marshal has

also to keep a considerable flying column in hand, to hunt Ballasteros

and other partisans. For operations outside the bounds of Andalusia he

can only collect a field-force of 24,000 men; this is the total figure

of the corps that tried to relieve Badajoz, and in its absence Seville

was nearly lost. The posts in the Sierra Morena were called in at that

time, and have never come back: correspondence with the Army of the

South is therefore precarious and slow.

(5) The Army of Portugal has 52,000 men effective. It holds the front

line against Wellington; its divisions are much scattered, because it

has to live on the country, and has also to furnish several important

garrisons. One division of 6,000 men is fixed down in the Asturias by

the Emperor’s special orders. The garrisons of Astorga, Valladolid,

Salamanca, Leon, Palencia, &c., absorb 6,000 or 7,000 men more. Only

29,000 infantry [or a total of 35,000 of all arms] are available

as a field-force to use against the English, if they attack on the

front of the Tormes. If Marshal Marmont has to march out of his own

sphere, to join in a combined operation against Wellington [e. g. in

Estremadura], he can bring a still smaller force--say 25,000 men.

The Army of Portugal is many months in arrear of its pay, and has

hardly any transport or magazines: the troops have become terrible

marauders--largely from necessity.

(6) Lastly we come to the Army of the Centre. It consists of 9,500 men

borne on the Imperial muster-rolls, and 5,800 troops belonging to the

King [his Guards and Hugo’s _Juramentados_, horse and foot]. There

are also at present in Madrid 3,200 drafts for the Army of the South,

temporarily retained--so that the whole makes up 18,500 men. But only

15,000 are effective, the remainder consisting of dépôts, dismounted

cavalry, train, &c. Having to hold down the extensive provinces of

Madrid, Segovia, Guadalajara, Toledo, La Mancha, and Cuenca, this

force is a mere “army of occupation.” It can provide no troops for

expeditions outside its own territory, and is spread so thin that even

Madrid would be in danger without the Royal Guards. The pay is eight

months in arrear.

(7) Civil administration is still localized: the commanders of the

armies levy their own taxes, and nothing comes to Madrid. The King has

to feed the Army of the Centre, and to maintain his civil service,

from the revenues of New Castile alone. None of the marshals will help

another with money or stores. The claim of the King to rule all Spain

seems absurd to the people, so long as he cannot exercise any civil

control outside the _arrondissement_ of the Army of the Centre.

(8) Conclusion. All offensive operations are impossible, as long as

the imperial armies have to hold down the entirety of the occupied

provinces. If Lord Wellington concentrates all his forces, he can march

with 60,000 men [not including Spaniards] against either the Army of

Portugal or the Army of the South. Neither of them can assemble a

sufficient force to resist him, unless they abandon whole provinces.

The King has ordered Soult and Marmont to march to each other’s aid if

either is attacked. But they have to unite, coming from remote bases,

while the enemy can place himself between them and strike at one or the

other. The lines of communication between them are long and circuitous.

It is easily conceivable that one of them may be attacked and beaten

before the other is even aware of the danger. A catastrophe is quite

possible if Lord Wellington should throw himself suddenly, with his

whole force, upon either the Army of Portugal or that of the South.

The only possible way of dealing with this danger is to collect

a central reserve of 20,000 men at Madrid, which can be promptly

transferred to right or left, to join either Soult or Marmont as

the conditions of the moment dictate. The Army of the Centre cannot

serve this purpose--it is not a field-force, but an immovable army of

occupation. If the Emperor could send a new corps of this size from

France, Marmont could be reinforced up to a strength sufficient to

enable him to face Wellington, and to besiege Ciudad Rodrigo.

But the present posture of European affairs [the Russian war]

probably makes it impossible to draw such a corps from France. This

being so, the central reserve must be obtained from troops already

existing in the Peninsula. The only way to find them is for the Emperor

to consent to the evacuation of Andalusia. Thirty thousand men of the

Army of the South can then be placed to cover Madrid, in La Mancha:

this force would be ample against any Spanish levies that might come

up to the Sierra Morena from Cadiz and elsewhere. The remainder of

the Army of the South must form the central reserve, and prepare to

reinforce Marmont. The Army of Portugal would then be so strong that

Wellington could not dare to take the offensive--he would be hopelessly

outnumbered. If this scheme is approved by the Emperor, he may be

certain that, when he comes back from Poland, his Spanish armies will

be in the same secure defensive position in which he leaves them now.

The right wing rests on the Bay of Biscay in the Asturias: the left on

the Mediterranean in Valencia.

When Andalusia is evacuated, the remaining provinces in French

occupation will not be able to pay or feed the 54,000 men of the Army

of the South, in addition to the armies already stationed in them; a

liberal subsidy from Paris will be necessary. In addition the King

must, for the sake of his prestige, be given real civil authority over

all the provinces.

It will only be when all authority, civil, military, and

administrative, is concentrated in one hand, that of the King, and when

His Majesty shall have received from the Emperor instructions suiting

the present posture of affairs, that he can be fully responsible for

Spain.’

On the whole this is a very well-reasoned document. It was perfectly

true that the offensive power of the French in the Peninsula had shrunk

to nothing, because no province could be held down without a large

garrison. If left unoccupied, it would burst into revolt and raise an

army. This was the inevitable nemesis for a war of annexation directed

against a proud and patriotic people. There were 230,000 French

troops in Spain; but so many of them were tied down to occupation

duty, that only about 50,000 or 60,000 could be collected to curb

Wellington, unless some large province were evacuated. Either Andalusia

or else Valencia must be abandoned. The former was the larger and the

more wealthy; but it was more remote from the strategical centre of

operations in Madrid, much more infested by the bands of the patriots,

and it lay close to the sphere of operations of Wellington--the great

disturbing element in French calculations. Moreover its evacuation

would set free a much larger field army. Against this was to be set

the adverse balance in loss of prestige: as long as Cadiz appeared to

be beleaguered, the national government of Spain looked like a handful

of refugees in a forlorn island. To abandon the immense lines in front

of it, with their dependent flotilla (which must be burnt, since it

could not be removed), would be a conclusive proof to all Europe that

the main frontal offensive against the Spanish patriots had failed.

Seville and Granada, great towns of world-wide fame, would also have to

be abandoned. Andalusia was full of _Afrancesados_, who must either be

shepherded to Madrid, or left to the vengeance of their countrymen.

But to weigh prestige against solid military advantage, though it might

appeal to Napoleon--whose reputation as universal conqueror was part

of his political stock-in-trade--did not occur to the common-sense

intellect of Jourdan. He voted for the evacuation of Andalusia: so did

his friend and master, King Joseph. Possibly their decision was not

rendered more unwelcome by the fact that it would certainly be most

distasteful to Soult, whom they both cordially detested. The Viceroy

should pay at last for the selfish policy of the General: his realm,

for the last two years, had been administered with much profit and

glory to himself, but with little advantage to the King at Madrid, or

the general prosperity of the French cause in Spain. Whether personal

motives entered into the decision of Joseph and Jourdan we need not

trouble to consider: it was certainly the correct one to take.

Permission to evacuate Andalusia was therefore demanded from the

Emperor: King Joseph did not dare to authorize it on his own

responsibility. Meanwhile, long before the _Mémoire_ of May 1812 had

been completed or sent off, to Napoleon, he issued the orders which

he thought himself justified in giving in the interim, to act as a

stop-gap till the permission should be granted. Marmont was told to

fall back on his own old policy of keeping a large detachment in the

Tagus valley, in order that he might get into touch with Drouet and

Soult’s Estremaduran corps of observation. He was directed to send

two divisions of infantry and a brigade of light cavalry to join Foy,

who was still in the direction of Almaraz and Talavera. They were to

be ready to act as the advance of the Army of Portugal for a march on

Truxillo and Merida, if Wellington’s next move should turn out to be

an attack on Soult in Andalusia. In a corresponding fashion, Soult was

ordered to reinforce Drouet up to a force of 20,000 men, and to push

him forward to his old position about Almendralejo, Zalamea, Merida,

and Medellin, in order that he might march via Truxillo to join the

Army of Portugal, in case the Anglo-Portuguese army should choose

Salamanca, not Seville, as its next objective. The small part of the

Army of the Centre that could be formed into a field-force--three

battalions and two cavalry regiments, under General d’Armagnac--was

directed to move to Talavera, to relieve Foy there if he should be

called to move either north to join Marmont on the Tormes, or south

to join Soult on the Guadiana[346]. To replace these troops, drawn

from the provinces of Cuenca and La Mancha, Joseph--as we have already

seen[347]--requested Suchet to send ‘a good division’ from Valencia by

Cuenca, on to Ocaña in La Mancha[348]. In this way the King and Jourdan

thought they would provide for active co-operation between the Armies

of Portugal and Andalusia, whether Wellington should make his next move

to the South or the North.

[346] See Jourdan to Berthier of April 3, 1812.

[347] See Jourdan’s _Mémoire_, quoted above, p. 304.

[348] Jourdan to Suchet, April 9, 1812.

It is curious, but perhaps not surprising, to find that these orders,

the first-fruits of Joseph’s new commission as Commander-in-Chief, were

obeyed neither by Suchet, by Soult, nor by Marmont.

The former, as we have already seen, when analysing Jourdan’s _Mémoire_

of May 1812, not only refused to send a division to Ocaña, but stated

that he should be obliged to withdraw the regiment that he was keeping

at Cuenca, because he was authorized by the Emperor to reserve all his

own troops for the defence of his own sphere of action, in Valencia,

Aragon, and Catalonia. Soult declared that it was impossible for him

to reinforce Drouet--‘he could not keep 20,000 men on the Guadiana

unless he received large reinforcements: all that he could promise

was that the force in Estremadura should move up again to Medellin

and Villafranca, possibly even to Merida, if Wellington had really

gone northward with his main army. Drouet, with his 10,000 or 12,000

men, might serve to “contain” Hill and the British detachment in

Estremadura, and his position would prevent the enemy from making any

important movement in the valley of the Tagus. Meanwhile he himself

must, as an absolute necessity, lay siege to Tarifa for the second

time, and make an end of Ballasteros: no more troops, therefore, could

be sent to Drouet: but when Tarifa and Ballasteros had been finished

off, the siege of Cadiz should be pressed with vigour.’ This reply

is not only a blank refusal to obey the King’s orders, but amounts

to a definite statement that the local affairs of Andalusia are more

important than the general co-operation of the French armies in Spain.

As we shall presently see, Soult was ready to formulate this startling

thesis in the plainest terms--he was, ere long, to propose that the

King and the Army of the Centre should evacuate Madrid and retire upon

Andalusia, when things went wrong with the Army of Portugal.

As to Marmont, his reply to King Joseph’s dispatch was couched in

terms of less open disobedience, but it was by no means satisfactory.

He wrote from Salamanca, on April 29th, after his return from the

raid to Sabugal and Guarda, that he had now learnt (what he did not

know ten days before), that Wellington had been pursuing him with

five divisions. This force was still in the Beira, and the British

general himself had been at Ciudad Rodrigo on the 26th. It was,

therefore, quite clear that Soult had not ‘the whole English army on

his shoulders.’ This being so, it was not necessary to send into the

valley of the Tagus such a large force as was asked. But one division

should move to Avila at once, and could drop down on to Talavera

in two days, if it turned out to be necessary. Two more should be

cantoned about Arevalo and the Pass of Piedrahita [20 miles north-west

of Avila] respectively, points from which they could be transferred

to the valley of the Tagus in a few days. Marmont then proceeded to

warn Jourdan against any scheme for concentrating any considerable

force in the direction of La Mancha, urging that he must be able to

collect as many of his divisions opposite Wellington as possible, in

case of an advance by the Anglo-Portuguese army towards the Tormes. All

that was necessary on the Tagus was to have the forts at Almaraz well

garrisoned and provided with stores, so that troops dropping down from

Avila on a southward march should find a base and magazines ready for

them. Summing up, he ends with a dictum that ‘if we defend Andalusia

by sacrificing the Army of Portugal, we may save that province for

the moment, but the North will be in danger: if a disaster occurs

there, Andalusia will soon be lost also. If, on the contrary, we make

its defence in the North, the South may be lost, but the North still

remains secure.’ By these somewhat cryptic words, Marmont seems to mean

that, looking at the affairs of Spain at large, Andalusia may be lost

without any shock to the imperial domination in Leon and Old Castile.

But a disaster in Leon or Old Castile entails inevitably the loss of

Andalusia also. This was true enough, though Soult refused to see it.