Salamanca, 1812
Page 36
The allies did not gain a final victory in the Peninsula in 1812. The Battle of Salamanca transformed the prospects facing Wellington, rendered Marmont's army incapable of serious operations for a time, and opened the road to Madrid; but it altered rather than completely overturned the balance of forces in Spain. There were still more than 200,000 French soldiers south of the Pyrenees, far more than Wellington could hope to oppose if they combined against him. The allied army was suffering from the demands of the campaign and the heat of the summer, with a staggering total of twenty thousand British rank and file (not counting the Portuguese) reported sick on 25 August.25 By occupying Madrid, Wellington had gained the central position, but found himself faced with the difficult, perhaps insoluble, problem of how best to use it. He considered marching south and attacking Soult in Andalusia, but decided to postpone this until a cooler season. In the meantime he returned to the north, leaving part of his army, including his best troops, in Madrid. His intention was to drive Clausel beyond the Ebro and, if possible, discover some means of protecting this flank when he again moved south. This led him to undertake the siege of Burgos, despite lacking the necessary means. The siege proved a protracted, costly and demoralizing failure; but even if it had succeeded, it is unlikely that it would have achieved his objectives, for the French were collecting vast forces against him.
The allied capture of Madrid had posed a threat to all the French armies in Spain and forced the marshals to subordinate their local concerns to the requirements of the wider war. Soult reluctantly raised the siege of Cadiz, evacuated his rich viceroyalty and marched through Granada and Murcia to join Joseph and Suchet in Valencia. In the north, Souham took command of the Army of Portugal, which had regained its order and composure, if not its confidence, remarkably quickly; while Caffarelli, threatened by Wellington's advance on Burgos, lent him ample support. Although Wellington brought Hill up from Estremadura to Madrid and received a small reinforcement from the garrison of Cadiz, the allies were soon heavily outnumbered on both fronts. Wellington raised the siege of Burgos on 22 October and began to withdraw, and nine days later Hill's rearguard abandoned Madrid. On 8 November the allied army was reunited on the Tormes, but the retreat continued. A week later, at Salamanca, Wellington offered battle, but although the combined French army under Joseph and Soult was much stronger than the allies, it would not fight and instead sought to manoeuvre Wellington into retreat by threatening his lines of communication. This time there was no mistake, and the allied army, weary and dispirited, was forced back to the Portuguese frontier, where it sullenly went into winter quarters after eleven months of extremely demanding operations.
And so the year ended as it began, with the allied army cantoned along the frontier. At first glance nothing appeared to have been gained except the capture of Ciudad Rodrigo and Badajoz in the winter and spring. But such glances are apt to mislead; much more than this had been accomplished. The allied army had proved its ability to manoeuvre and attack in the open field, and had established a psychological superiority over the French which remained unchallenged for the rest of the war. The siege of Cadiz had been raised and Andalusia had been freed. Madrid had been liberated and, although lost again, the authority of Joseph Bonaparte's government had been dealt an irreparable blow. French garrisons had been weakened or withdrawn from almost every province in Spain, opening the way for the guerrillas who flourished in their absence. The lasting results of this were not yet apparent, but would become obvious as the insurrection in northern Spain gained strength in the winter and spring, paving the way in turn for Wellington's rapid advance to Vitoria in 1813. Beyond the Pyrenees all Europe had seen that the French war in Spain was not going well; that it was no mere matter of suppressing brigands; and that there was no end in sight for the conscription needed to sustain Napoleon's armies. Of course, if Napoleon had triumphed in Russia, the new year might have seen another great influx of reinforcements such as that which had followed the defeat of Austria in 1809. Wellington might have been driven back to the Lines of Torres Vedras, Andalusia might have been overrun again and Joseph's authority might have been supported by another 200,000 French soldiers. But all the gains of the past three years (except Valencia) had been lost, and the Spanish guerrillas were vastly more confident, experienced and well organized. The war was not being won, and the cost of continuing it – in men, in money and above all in domestic support in France – was rising steeply. The illusion was shattered, and the Emperor's Spanish policy stood revealed in all its nakedness.
Appendix I
Casualties Suffered on 18 July 1812
ALLIED CASUALTIES
From the return in PRO WO 1/255, pp. 91–2.
officers men total
killed wounded killed wounded missing
Fourth Division (Cole)
3/27th Foot 2 1 11 58 – 72
1/40th Foot – 1 8 59 1 69
1/7th Foot – 1 1 14 3 19
1/23rd Foot – – – 2 2 4
1/48th Foot – – – 5 1 6
Fifth Division (Leith)
3/istFoot – – – 2 – 2
detachments of 5/60th – – – 1 2 3
Portuguese (all arms) 1 6 33 90 27 157
G. Anson's brigade
11th Light Dragoons – 2 3 11 – 16
12th Light Dragoons – 1 5 12 1 19
16th Light Dragoons – 1 3 8 3 15
V. Alten's brigade
1st Hussars KGL – 4 7 45 4 60
14th Light Dragoons – 3 14 49 9 75
Bock's brigade
1st Dragoons KGL – – – 1 – 1
2nd Dragoons KGL – – 5 1 1 7
Le Marchan's brigade
3rd Dragoons – 1 – 9 – 10
Horse artillery – 1 2 2 – 5
German artillery – – – 2 – 2
* * *
Total 3 22 92 371 54 542
FRENCH CASUALTIES
Officer casualties come from Martinien; those for the men are based on the overall losses given in Lamartinière's return (printed in Sarramon, La Bataille des Arapiles, pp. 423–5) for the whole campaign divided in the same proportions as overall officer losses. For example, if a unit loses ten officer and one hundred rank-and-file casualties in the campaign as a whole, and Martinien identifies three of these officer casualties as occurring on 18 July, it is assumed that thirty of the casualties among the men also occurred on the 18th. For further discussion see Appendix III.
2nd Division (Clausel)
25th Léger lost 1 officer killed and 5 wounded = 6 officer casualties, suggesting 102 men, a total loss of approx. 108, or 7.0%
6th Division (Taupin)
17th Léger lost 2 officers killed, and 6 wounded (1 mortally) = 8 officer casualties, suggesting 184 men, a total loss of approx. 192, or 17.1%
65th Ligne lost 2 officers killed, and 1 wounded = 3 officer casualties, suggesting 87 men, a total loss of approx. 90, or 5.7%
22nd Ligne lost 5 officers wounded (1 mortally) = 5 officer casualties, suggesting 188 men, a total loss of approx. 193, or 12.5%
Light Cavalry Division (Curto)
3rd Hussars lost 2 officers wounded, suggesting 22 men, a total loss of approx. 24, or 9.7%
14th Chasseurs lost 3 officers wounded, suggesting 29 men, a total loss of approx. 32, or 9.9%
Dragoon Division (Boyer)
Brigadier Carrié was wounded and captured
15th Dragoons lost 1 officer killed and 1 wounded = 2 officer casualties, suggesting 26 men, a total loss of approx. 28 or 8.2%
25th Dragoons lost 4 officers wounded, suggesting 24 men, a total loss of approx. 28, or 8.4%
Total Loss
Infantry 22 officers and approx. 561 men, total 583
Cavalry 12 officers and approx. 101 men, total 113
Total 34 officers and approx. 662 men, total 696
Increasing these figures to allow for unwounded prisoners and slightly wounded men who had returned to the ranks by 8 August (see Appendix III for
more details) produces an estimate of total losses on 18 July of 40 officers and 794 men, or 834 casualties in all.
Appendix II
Allied Strength and Losses
The following table gives the strength of the allied army on 15 July, the date of the last return before the battle. The figures for the British are based on the original manuscript weekly state in the Public Record Office at Kew (WO 1/255), and are much the same as those given by Oman, with a few minor corrections.* The figures for the Portuguese come from a less detailed morning state dated 12 July, also in the PRO (WO 1/255). This agrees very closely with the figures printed by Oman from a return in the archives in Lisbon;† unfortunately, neither gives the strength of individual regiments or battalions, only the total for each brigade.
It is worth explaining that the original weekly state in the PRO is far more detailed than the version published here. There are almost fifty columns for each regiment, giving the number of officers of each rank; while sergeants, drummers and rank and file are each subdivided into ‘Present’, ‘Sick – Present’, ‘Sick – Absent’, ‘Command’ (men not currently serving in their normal place in the regiment, although fit for duty) and ‘Prisoners of War and Missing’. Then follow columns for horses and for ‘Alterations since last Return’, which is further broken down under headings such as ‘Joined’, ‘Deserted’, ‘Transferred’, ‘Promoted’ and so on. Comparison shows that Oman disregarded most of these columns when calculating the effective strength of the regiment, and his practice has been followed here. The number of officers given is simply the total number of officers in the regiment: presumably sick or absent officers were returned separately. The number of ‘men’ is a little more complicated. For the infantry, Oman just adds together the sergeants, drummers and rank and file listed as ‘Present’ (not including ‘Sick – Present’). But when dealing with the cavalry he looks instead to the column showing the number of horses present, which he equates with the number of men the regiment could put into the field. This decision is justified, for it gives a total strength for each brigade equal to the final column in the return: ‘Effective Rank and File for Each Brigade and Division’. The phrase ‘rank and file’ here may cause some concern, but as the figure is based on the number of horses returned as present with the regiment, and as NCOs were certainly both present and mounted in action, it seems that it must be used loosely. On the other hand, the same column for the infantry gives simply the total rank and file listed as present and does not include sergeants or drummers.
The artillery returns present a puzzle. The manuscript ‘state’ does not give figures for individual batteries and includes units which were with Hill's corps in Estremadura and possibly also in the rear, and so is a poor guide to the strength of Wellington's artillery in the battle. The Dickson Manuscripts clearly identifies which batteries were present and the number and type of guns, but not the number of men; while a return referred to in the Ordnance Papers in the PRO is no longer present.* However, Oman gives reasonably detailed, plausible figures, and for want of any useful original returns these have been adopted, even though they may be an underestimate (see chapter two, commentary, p. 45).
The quite substantial reinforcements which joined the allied army between 15 and 22 July (such as the 1/38th) are included in the return, but no attempt has been made to deduct the losses which were suffered in this week – men falling out from exhaustion and the heat, and casualties, principally those suffered on 18 July. A rough estimate puts this at about 1,200 men, most of whom would, of course, soon rejoin the army.*
The figures for British casualties in the battle come from the official return in the PRO (WO 1/255) and differ only slightly from those given in Oman.† However, there are two sets of figures for the Portuguese losses. Wellington reported in his official dispatch that the Portuguese suffered 2,038 casualties: 304 killed, 1,552 wounded and 182 missing; but Oman has found a detailed return in the archives in Lisbon which gives a lower figure of 1,627 casualties: 506 killed, 1,035 wounded and 86 missing. Oman notes the difference, but believes that there is a prima facie case for preferring ‘the later and more carefully detailed document’.‡ However, there are two powerful arguments against this choice: first, the sheer improbability of any general overstating his losses – or the losses of a significant part of his army – by 25 per cent; second, the proportion of killed to wounded in the later return is highly unusual. In Napoleonic battles for which we have reliable statistics, the killed normally amount to between 10 and 20 per cent of the casualties. Of the British casualties at Salamanca, 12.4 per cent were killed: the first Portuguese return shows 14.9 per cent killed; the second, 31.1 per cent. It is hardly likely that Portuguese soldiers fighting side by side with their British comrades were two and a half times as likely to be killed rather than wounded. A much more plausible explanation is that the second return was prepared some time – probably a few weeks – after the battle. Many of the slightly wounded men who had returned to the ranks were not counted (hence the fall in the overall total), while the number killed was swollen, both by the discovery of the bodies of men originally listed as missing and by the death of some of the most seriously wounded.
Unfortunately, this first return gives only the total numbers of Portuguese casualties, not those suffered by individual units; but the later return, which Oman prints, does give the losses of each brigade. If the explanation of the difference between the two returns given above is correct, one would expect the understatement of losses in this second return to be spread fairly evenly across the army. Therefore a rough calculation of the true losses of each brigade can be made by multiplying Oman's figure by 1.253, which, with a little rounding, produces the desired total of 2,038. Naturally the resulting figures cannot pretend to be precise, but they are plausible and appear to be the best that we can do. In the table, detailed figures for each Portuguese brigade are taken from Oman's return, with an adjusted total in brackets after the column for total casualties.
The strength of the Spanish division is given by Oman as 160 officers, 3,200 men, total 3,360. The round figures and exact proportion of one officer to twenty men suggest that this may be no more than an approximation. Recent Spanish research, using the papers of Don Julián Sánchez, states that he had a surprisingly powerful brigade of cavalry in the field: two regiments of lancers and a company of artillery (two 4-pounders), amounting to 950 men. Although this brigade was nominally subordinated to Don Carlos de España, it is said to have received its orders direct from Wellington's headquarters.* The strength and composition of the remainder of the Spanish force are uncertain: the assumption made here is that Oman's overall figure is correct, which means that the infantry under España's command amounted to 2,410 all ranks. It is not known how the six recorded Spanish casualties (two killed, four wounded) were distributed between the two elements.
No figures, whether for the strength or the losses of any army in any battle, are absolutely accurate. They include men who left the ranks early in the day and who took no part in the fighting, and fail to include the officially sick or absent soldier who went out of his way to join his comrades for the great occasion. Many officers refused to report themselves wounded for fear that their name appearing in the papers at home would alarm and distress their families, while others made the most of every scratch in the search for glory, promotion and renown. Some officers might understate their unit's losses by only recording men with comparatively serious wounds, while others would inflate their unit's casualty roll. These problems should deter us from regarding these statistics as being incontrovertible or above reproach; but in most cases their impact was probably slight, and the figures remain one of the most useful tools for seeking to understand what happened in the battle.
Summary of Allied Army
(Using amended – higher – figures for Portuguese casualties. No deduction has been made from the strength of the army for losses suffered on 18 July, or for men who fell out in the days before the bat
tle.) strength casualties % casualties
First Division 6,428 153 2.4%
Third Division 5,875 582 9.9%
Fourth Division 5,218 1,116 21.4%
Fifth Division 6,710 659 9.8%
Sixth Division 5,551 1,803 32.5%
Seventh Division 5,175 146 2.8%
Light Division 3,538 48 1.4%
Pack's Brigade 2,607 471 18.1%
Bradford's Brigade 1,894 21 1.1%
Spanish Division 2,410 6 0.2%
Le Merchant's Brigade 1,032 105 10.2%