Salamanca, 1812
Page 12
Commentary
It is worth drawing attention to the fact that 22 July 1812 was indeed a Wednesday, as a surprising number of the less good secondary sources make it a Sunday, apparently for no better reason than the belief that all Wellington's great victories were preceded by a thunderstorm and fought on the sabbath.41
On the evening of the 21st Wellington learnt that General Chauvel with a brigade of cavalry from the Army of the North had arrived at Polios. Oman puts this force at fewer than 800 cavalry plus a battery of artillery, while Sarramon makes it 900 cavalry and eight guns, which, including the artillerymen, would make a total force of rather more than 1,000 men. This seems much more plausible than Fortescue's 1,700 cavalry with 20 guns, which probably reflects Caffarelli's promise rather than his performance.42
There is a vivid picture of the state of Salamanca on the day before the battle in the manuscript narrative by Andrew Leith Hay which differs significantly from his published account:
I went into Salamanca[,] which town appeared in a very different light from what it did when the Army left it to advance. Many people had shut their shops and fled, and the Countenances of those that remained bespoke terror and consternation – they asked questions which could not be satisfactorily answered as to the movements of the Army, and any assurances of their safety from the French were received with distrust. They seemed to think we were going to leave them to their fate, and a wretched prospect these thoughts laid open to the unfortunate people of Salamanca – but this gloom was but the forerunner of a brilliant day.43
No wonder Charles Boutflower reflected that ‘Events may yet turn out well, but at present I fear we have gained but little popularity by our irruption into the North of Spain.’44
The city of Salamanca stands on high ground on the north bank of the Tormes, 803 metres above sea level according to the Spanish equivalent of the Ordnance Survey Map. The ford at Santa Marta a little upstream is 781 metres, which gives a better idea of the level of the river. Alba de Tormes, some miles further upstream, is 836 metres. On the battlefield itself, the village of Arapiles is 841 metres; the hill behind it, the Teso de San Miguel, rises to 864 metres, while the Lesser Arapile is 901 metres, and the Greater Arapile 906 metres. Thus, the summit of the Greater Arapile is some 200 feet higher than the village, which is about a mile distant. The ground further west is generally rather lower: the village of Las Torres to the rear of the allied position is 820 metres; Aldea Tejada where the Third Division and D'Urban waited is 795 metres; and the hamlet of Miranda de Azan beyond the western end of Monte de Azan is 824 metres. The highest point of Monte de Azan is 877 metres.45
Of the well-known maps of the battlefield, the one which appears the most detailed and authoritative is that in Fortescue's History of the British Army. It covers the whole area from Salamanca to Alba de Tormes, including Huerta, where the river turns sharply west, and it shows the rise and fall of the land with contour lines marking every ten-foot change in elevation. This is a little disconcerting at first, for it makes the country appear more rugged than is the case, but it promises a degree of precision and accuracy lacking in other maps. However, a close comparison between it and the modern Spanish ordnance map reveals significant differences in the distance between points. Both maps are on a scale of 1:50,000.
ordnance map Fortescue
Salamanca bridge to Los Arapiles 140mm (7.0 km) 137mm (6.85 km) 97.9%
Santa Marta to Los Arapiles 126mm (6.3 km) 120mm (6.0 km) 95.2%
Aldea Tejada to Calvarrasa de Arriba 174mm (8.7 km) 157mm (7.85 km) 90.2%
Las Torres to Calvarrasa de Ariba 103mm (5.15 km) 88mm (4.4 km) 85.4%
Las Torres to Los Arapiles 36mm (1.8 km) 29mm (1.45 km) 80.6%
Greater Arapile to Los Arapiles 35mm (1.75 km) 37mm (1.85 km) 105.7%
Greater Arapile to Lesser Arapile 17mm (850 m) 19mm (950 m) 111.8%
There are also a number of other discrepancies, of which the most important is that Fortescue's map shows the Teso de San Miguel lying immediately behind the village, while the ordnance map (confirmed by personal observation) places its highest point further east, filling part of the gap between the village and the Lesser Arapile. Nor do the gradient lines always appear accurate when compared to the terrain on the spot.
Detail from Fortescue's map of the battle, reduced by approximately twenty per cent. The whole original is much larger (330 × 350mm) and is coloured brown, green, blue and red making it much clearer. It is both attractive and impressive, but close examination casts doubt on its accuracy.
Of the other maps of the battlefield, Oman's appears surprisingly accurate, with no significant difference from the ordnance map in the distances tested above, while Sarramon's is also generally reliable, although in neither case is the depiction of the hills entirely convincing. The grandest and most elaborate map is that specially surveyed by Sir Thomas Mitchell in 1815 and published in 1841 in Maps and Plans Showing the Principal Movements, Battles and Sieges in which the British Army Was Engaged during the War from 1808 to 1814 in the Spanish Peninsula and the South of France (commonly known, after its publisher, as Wyld's Atlas). This large elephant folio is a tour de force of surveying and engraving and still provides the most accurate depiction of the terrain of the battlefield; however, it is extremely rare and rightly much prized by collectors. Two beautiful large maps accompany Marindin's 1906 The Salamanca Campaign (a handsomely produced book whose text is almost worthless), and there is a fascinating but not entirely accurate map in Arteche's Guerra de la Independencia – historia Militar de España de 1808 a 1814 (14 vols, 1868–1901, reprinted c. 2000), which may have provided the basis for Fortescue's map.
The maps drawn for the present study naturally take their fixed points from the modern ordnance map, with details of the terrain also taken from it where possible, supplemented by Wyld, Oman and personal observation. They are provided as a visual guide and cannot pretend to be absolutely accurate, while the troop locations marked on them are often highly conjectural. There is a common, largely unconscious inclination to accord information displayed on maps greater authority than to that given in the text, but in the end both depend on the same uncertain, partial and often contradictory sources.46
In his published Narrative Andrew Leith Hay writes:
The day was fine; the sun shone bright, nor was there any atmospheric obstruction to prevent a clear view of passing events; no haze withheld a distinct observation to the very extremity of the plains upon which the enemy was in movement. Occasional smoke from the firing, and dust, alone created a temporary uncertainty in the view of any movement of either army.47
But this was early in the day, and in his manuscript account of the battle Leith Hay says, ‘The Dust was so great as to cover the Troops, and appeared one immense moving mass of smoke. The ground was so dry that the instant we began to move it flew up in all directions.’48 Given the fury of the storm of the previous night, it is surprising to read that the ground was dry, but many other accounts also mention the dust: presumably the hot July sun dried out the ground during the morning and early afternoon. Perhaps also the ground near the Arapiles – village and hills – received less rain than that near the river where most of the allied army camped overnight. Most troops, whether closely engaged or not, could see little of the battle, and only experienced and – usually – mounted officers could hope to form any clear idea of the progress of the day as a whole. One British Guards officer, John Mills, wrote home to his mother:
I saw our line advance, and occasionally could see through the smoke the contests in the different places, but there was so much noise and such confusion that it was quite impossible to make anything clearly out. I had not the most distant conception but that it was a small body on both sides fighting hard for some ground that both wanted.49
In Oman's account of the skirmishing around Nuestra Señora de la Peña, he identifies the Portuguese light infantry involved in the fighting as the 2nd Caçadores from the Seventh Division (
and I followed him when discussing the incident in my Tactics and the Experience of Battle in the Age of Napoleon). However, Sarramon points out that Wellington in his official dispatch, explicitly names the 4th Caçadores from Pack's brigade as the unit engaged, and there is no reason to doubt this.50 Wellington goes on to say that the caçadores maintained their position throughout the day. If taken literally, this would suggest that Pack's brigade was without them when it made its attack on the Greater Arapile, but this is directly contradicted by Synge's account, which on this point must be preferred. Presumably, therefore, the 4th Caçadores were withdrawn when the Light Division relieved the Seventh Division in the middle of the afternoon, or possibly even sooner.51
Oman's map of the opening stage of the battle. The original is rather larger (230 × 185mm) and is printed in brown, red and green. Many of the troop locations and some details of the terrain are open to dispute, but it is much more accurate than Fortescue's map.
It seems surprising that the allied troops retained their foothold on the heights near the chapel, for their supports and reserves in the open valley beyond must have been dreadfully exposed to French observation and artillery fire, and they were almost half a mile in advance of their main position. However, Wellington's statement is explicit and is supported by first-hand British accounts (such as Private Green's quoted in the text), while it is not contradicted by Marmont, Foy or Sarramon.52
Victor Alten's rashness in leaving Salamanca despite his wound was not entirely without harmful effects. When James McGrigor, the chief of the medical staff, reported the progress of the wounded to Wellington a few days after the battle, he told him that Alten ‘has some unpleasant symptoms but … I still do not conceive [him] to be in danger’.53 Alten's rapid recovery is evidence of the good care he received, and is one example among many of the toughness of senior officers on both sides in the Peninsula.
When Wellington discouraged J. W. Croker from writing a history of the Battle of Waterloo, he compared a battle to a ball, then continued: ‘Some individuals may recollect all the little events of which the great result is the battle won or lost; but no individual can recollect the order in which, or the exact moment at which, they occurred, which makes all the difference to their value or importance.’54 Questions of sequence and timing are indeed among the most intractable in studying any Napoleonic battle. Salamanca is not nearly as difficult as Waterloo in this respect and, once the fighting begins, the course of the action is mostly fairly clear, although a few points remain debatable. But the events of the morning and early afternoon, and the time at which actions occurred, remain difficult or impossible to determine. The problem is normally not a shortage of evidence, but a plethora of contradictory statements. There are two explanations for these contradictions. The first is that honest, well-disposed and intelligent eyewitnesses are simply mistaken in their recollection; with the stress and anxiety of battle, time passes unexpectedly quickly or slowly and, with all the other impressions made by the sights and sounds of combat, the time at which events happened does not stick fast in the memory. The second is more prosaic but no less important: there was no radio broadcast of Big Ben striking the hour against which officers could set their watches, and there seems to have been no habit of synchronizing watches or giving orders which depended on the time (except, perhaps, in sieges). In all probability our witnesses could not have agreed on the time within an hour had they met on the battlefield with no other purpose than to discuss this very point. For these reasons all the times which can be given in this or any other Napoleonic battle should be regarded as no more than rough approximations, and histories which divide the day into precise segments, or which pin arguments on fine matters of timing, should be regarded with lively scepticism.
The following collection of times gives some idea of the range and variation, but also of the probable sequence of events on the morning and early afternoon of 22 July:
The skirmishing began at 6 am according to John Mills; 7 am according to John Aitchison.55
The French seized the Greater Arapile at around 8 am is Sarramon's plausible estimate, but Lemonnier-Delafosse, in his memoirs, puts it at about 3 pm – one of the easier statements to disregard.56
Marmont climbs the Greater Arapile: soon after 8 am in Sarramon's view, but Oman thinks it was not until about 11 am.57
The aborted allied attack: Marmont and John Mills both put this at 11 am, and Aitchison rather supports them by saying that it had been cancelled and the troops had resumed their positions by noon. But the normally reliable Tomkinson says that it occurred at about noon and his word convinces Oman – this also receives some slight support from a statement in Ellesmere's Personal Reminiscences of the Duke of Wellington.58
The French advance onto the Monte de Azan: at 1 pm according to Burgoyne, about 2 pm according to Wellington's dispatch, although Marmont says he only issued the orders about 2.59
The French bombardment begins: Leith Hay and Mills agree that this was about 3 pm.60
Wellington decides to attack: Sarramon puts this as early as 2.30 pm, Napier makes it soon after 3 pm, Andrew Leith Hay, in an unpublished letter, favours 3.30pm; but in his manuscript narrative Leith Hay says about 4 pm, while John Mills puts it as late as 5 pm.61
Wellington personally gives his orders to Pakenham to attack: at about 2 pm according to Brigade-Major Campbell, at 3 pm according to Pakenham himself (although in another letter he says that this was when his attack began); 3 pm or a little later is also favoured by Sarramon and Moggridge, but Oman makes it around 3.45 and D'Urban puts it after 4.30.62
Many more examples could be given, while the most important time of all – when Marmont was wounded – has already been discussed (but see also below, pp. 79–80).
The uncertainty on questions of time, and gaps in other evidence, make it impossible to reconstruct fully the movements of each army on the morning and early afternoon of the 22nd; nonetheless, plausible solutions to some problems can be advanced. The movement of the Third Division across the Tormes has caused some difficulty, with Fortescue making it ford the river at Cabrerizos, while Oman has it cross by the town bridge of Salamanca. Each cites supporting evidence for his view, suggesting that neither is completely wrong. Young and Lawford put forward the obvious compromise, although it is not clear whether there is any basis for their suggestion that another part of the division used the ford at Santa Marta, or for their idea that Pakenham thus divided his division in three on his own initiative. Two sources – Brigade-Major Campbell and Brigadier D'Urban – say that their units crossed the river about noon, which makes it seem likely that they reached Aldea Tejada about 2 o'clock, as both Oman and Young and Lawford suggest.63
When Victor Alten's brigade of light cavalry, now under the command of Lieutenant-Colonel Arentschildt, was moved to the right wing of the army, part of the 14th Light Dragoons remained on the left. Beamish states that this detachment amounted to two squadrons, but Wellington's dispatch clearly states that two squadrons under Lieutenant-Colonel Hervey were on the right, implying that only one had been detached. Rather oddly, Arentschildt makes no mention of the detachment in his report to Cotton (although his account implies that a substantial force of the 14th was with him), while Wellington makes none of Arentschildt or the 1st KGL Hussars – possibly Cotton's wound prevented Arentschildt's report reaching Wellington before he wrote his dispatch. There is also great uncertainty concerning aspects of Arentschildt's subsequent skirmishing with Curto's cavalry: where it occurred, when it began, how long it lasted and how it related to other manoeuvres are all quite unclear.64
Bonnet's occupation of the Greater Arapile poses slightly different problems. All accounts agree that the hill itself was held by the 120th Ligne, but it is not clear whether all three battalions deployed on the summit, or some of the men remained on the reverse slope. The number of guns dragged to the summit of the hill is also uncertain. Sarramon states that twenty guns were so deployed, but this appears to be based on a mis
reading of a passage in Marmont's memoirs which actually refers to the establishment of a battery on the ridge of El Sierro well to the rear.65 Dyneley claims that the French had only four guns on the Greater Arapile and that his own fire silenced them. This may be mere boasting, but the 3/27th, which shared the Lesser Arapile with him and which would have been the most obvious target for French guns opposite, lost only eight casualties all day, and Major Macdonald, Dyneley's commanding officer, later cited this as an occasion on which shrapnel proved its efficacy. Captain Wachholtz, commanding the company of Brunswick Oels light infantry attached to the Fourth Division not far from the Greater Arapile, saw few men on its summit, and states that the French had brought up only a few light cannon, although his view was probably obstructed by the rocky outcrop at the western end of the ridge. Finally, Captain Synge in describing Pack's abortive attack mentions French artillery, but not in a way that suggests he was attacking a large battery. So on the whole it seems likely that the French battery on the Greater Arapile was less powerful than has been commonly thought, even if it was not as insignificant as Dyneley claims.66