On 11 July, Fletcher reported his progress to Wellington: ‘I am sorry to say that having found much stone or rock in the ditches of the redoubts, their completion will necessarily be delayed somewhat beyond the time I had mentioned.’19 Three days later, Fletcher and Burgoyne set of for San Sebastian where the siege was just starting. Despite a belief that the garrison was desperately short of food, Pamplona held out until 31 October.
The First Siege of San Sebastian
After San Sebastian was invested on 11 July 1813, Wellington, Major Charles Smith RE20 and Dickson rode around the fortress, and at the suggestion of Smith,21 the same basic plan of attack was proposed as had previously been used by the Duke of Berwick in 1719. The plan was to breach the wall on the eastern side where it was fully visible, due to the sea going right up to the base of the wall at high tide and preventing any other form of defence in front of it. At low tide it was possible for troops to cross the tidal estuary of the river Urumea and storm any breach. It would also be necessary to take some of the outworks on the land side to reduce the fire that could be brought to bear on any attack across the estuary and also to give access to the foot of the eastern wall. With this aim in mind, the convent of San Bartolomeo was to be captured and trenches thrown forward to allow the defences to be silenced and for enfilade fire on the proposed breaches.
By the time Fletcher arrived at San Sebastian the plan of attack had been agreed. It is not clear if Fletcher agreed with the plan. Burgoyne certainly did not, but the matter was decided before he had a chance to influence the decision. Graham writing to Wellington reported that Fletcher was concerned about the protection of the troops during an assault and recommended a second breach. Graham later, also made the comment ‘It is evident, however, that Major Smith thought too lightly of the strength of the place.’22
Timeline for the First Siege of San Sebastian
11 July 1813
Batteries started against San Bartolomeo on night of 11/12 July
13 July 1813
Work started on batteries against eastern wall
14 July 1813
Fire opened on convent of San Bartolomeo
17 July 1813
Convent of San Bartolomeo taken
20 July 1813
Fire opened on fortress
21 July 1813
Governor rejects summons
23 July 1813
Main breach practicable, second breach started
24 July 1813
Assault postponed due to fires in town
25 July 1813
Assault fails
26 July 1813
Siege guns removed but blockade maintained on fortress
As at the siege of Ciudad Rodrigo, Wellington had effective siege guns available. A new siege train had been sent out to the Peninsula for use in northern Spain and with the six 18-pounders that travelled with the army and six 24-pounders supplied by the Royal Navy, from HMS Surveillante, Wellington had a siege train of forty guns made up of:
• Twenty 24-pounders including six naval pieces.
• Six 18-pounders.
• Four 68-pounder carronades (short-barrelled and not accurate but useful for spraying the breaches with canister).
• Six 8in howitzers.
• Four 10in mortars.
Whilst at first glance this looked like a substantial siege train, it was not for a fortress on the scale of San Sebastian. There were only twenty-six guns to batter the walls, of which six were the smaller 18-pounders which had significantly less battering effect. Of the larger 24-pounders, the six naval guns had shorter barrels, which reduced the accuracy that was vital for bringing down walls. Dickson noted in his journal that ‘Fletcher thought this scanty for such a fortress and it certainly is so, at least in heavy guns’.23
The engineering personnel commanded by Fletcher included seventeen officers and over 300 rank and file from the Royal Sappers and Miners.24 This was the first siege at which there were a significant number of them present. Although John Jones’ published Journal does not record the employment of assistant engineers (officers volunteering from the army to assist), both Burgoyne and Fletcher mention that a number were used on the left attack.25 In addition there was a party of fifty Portuguese artificers and engineers. Overall command of the siege was given to Sir Thomas Graham, Wellington staying with the army to monitor the activities of the French forces under Marshal Soult.
The initial attack was made against the convent of San Bartolomeo. Two batteries were constructed and they opened fire on 14 July. The following day, a force of Portuguese infantry was sent forward, but they encountered strong resistance and retired. The guns continued firing on the convent for two more days, and it was taken on the 17th, though not without considerable loss thanks to an undisciplined charge by British infantry against the main French positions. Two new batteries were started near the convent and the plan was to dig a parallel right across the isthmus.
The batteries against the eastern wall opened on the morning of 20 July and once the French realised where the main point of attack was going to be, they began establishing defences behind the wall being breached. The same night, the attackers started work on the main parallel across the isthmus, but due to the poor weather the majority of the Portuguese troops allocated for the work did not turn up and only a third of the planned work was completed. The following night, whilst completing this parallel, a large drain was found which had supplied water into the town until the supply was cut off. This was explored by Lieutenant Reid RE, who found it went up to the western side of the hornwork and it was decided to place a mine at the end of the drain with the intention of breaching the hornwork.
Burgoyne was sent by Graham with the first summons on 21 July, which was refused.26 Burgoyne regularly was used in this sort of role, a task that you would expect would be given to permanent members of a general’s staff. Clearly he possessed some skills that were judged useful in these situations. It may have been his language skills but I think it more likely that it was an opportunity to get a close-up, professional view of the defences.
The breach in the eastern wall appeared practicable on the 22nd, but the French were making great efforts to clear away the debris despite the Allies continually sweeping the breach with grapeshot and shells. On the following day, the breach was declared practicable and the guns were directed to make a second breach in the wall at a location that locals had suggested was particularly weak. This second breach was ready that night, although the continuous shelling had started numerous fires in the houses behind the two breaches. Graham ordered the assault for the 24th, but in the morning it was cancelled as it was thought the fires that were still raging would impede the troops.27 This delay gave the French an extra twenty-four hours to improve the defences, although they were working under a continuous bombardment from the attackers.
The plan of attack was for troops from the 5th Division to assault the two breaches, starting from the right (eastern) end of the parallel across the isthmus and skirting the foot of the wall until they reached the breaches, the siege batteries providing heavy covering fire from across the bay. The plan was dependent upon the time of low time and daybreak, which were both expected to be around 5 a.m. The signal for the start of the assault would be the blowing of the mine in the drain by the hornwork.
On the morning of the 25th, the mine was blown before daylight and the assault commenced. Filing out of the parallel was very slow and the first troops arrived at the breach only in small numbers. Although initially successful, there were not sufficient men present and they were quickly driven back, having been stopped by a twenty-foot drop from the breach into the town itself. In doing so, they became mixed up with the group who were tasked with assaulting the secondary breach and all retired in some disorder. The assault had failed completely before there was enough daylight for the artillery across the bay to provide any support, casualties amounting to 570 killed and wounded. There were five engineer casualties during the ass
ault. Lieutenant Colonel Richard Fletcher, Lieutenants Harry Jones and Reid were wounded, Captain Lewis lost a leg and Lieutenant Machell was killed. Another officer, Lieutenant Hammond Tapp, had been severely wounded earlier in the siege on 13 July.
When Wellington heard about the failure of the assault, he rode over from his headquarters determined to continue the siege. However, he accepted that it would have to be postponed temporarily as powder and shot were running low and new supplies needed to be delivered. In the meantime, as Soult was still threatening to attack, Wellington ordered most of the siege guns to be removed and returned to the boats where they would be safe until further ammunition was available. He ordered a tight blockade to be kept in place.
Analysis of the First Siege
Both contemporary and more recent writers have criticised the performance of the engineers in a number of areas. Fortescue28 leads the attack with an unjustifiable apportioning of blame for the failures at San Sebastian, while Oman29 only holds the engineers partially to blame but identifies them as the primary culprit. Their assessments in both cases appear to be led by the opinions of one particular army officer who clearly had a dislike for the ‘scientific soldiers’. The analysis below will look at the criticisms and compare them with to the available facts.
Oman, and more recently Myatt,30 criticised Major Smith’s proposal to follow the Duke of Berwick’s plan of attack of 1719. Oman wrote that Graham, Wellington, Fletcher, Dickson (commanding the Royal Artillery) and Frazer (commanding the siege batteries) all agreed with the plan and they ‘forgot’ that the Duke of Berwick did not have to assault the fortress.31 It is inconceivable, even excluding the other officers, that Wellington ‘forgot’ about the possibility of another costly assault. The strength of the fortress was directed against the land approaches for obvious reasons and once again the view was that there would be insufficient time to formally approach from the land side. Fletcher’s view on the proposed attack on the eastern side was that ‘it would certainly save much time [my italics] … compared with a regular siege of the very powerful defences crossing the isthmus’. An attack on the land front would be a ‘work of great difficulty’ requiring a larger battering train and thirty to thirty-five days’ effort.32 Burgoyne initially supported the proposed plan, although with the benefit of hindsight he thought that finding the drain tipped the balance in favour of an attack across the isthmus. He did, however, acknowledge that this would have taken more time. Oman and Fortescue both wrote that when Wellington arrived after the assault he was insistent that the siege would continue and required the engineers to come up with a plan for a formal attack from the land side. Burgoyne and Frazer indicate that an alternative plan of attack was discussed at the meeting with Wellington on 25 July. Jones, in his diary entry for that day, noted that ‘after some consideration, it was decided to persevere in the same plan of attack’.33 Fletcher also wrote to Wellington on the 27th noting that Wellington’s opinion was for an extended attack using the original plan.34 Lack of ammunition prevented any progress in the short term. By the time the new supplies had arrived, the plan, as Jones noted, remained the same as before, with an attack on the east-facing sea walls. Any thoughts of using a different plan were clearly put aside very quickly.
Oman’s narrative stated that when the mine was blown, the hornwork was to be assaulted by Portuguese troops from the parallel on the isthmus. He continued that the engineers were unsure how much damage would be caused by the mine and because of this no concrete proposals were made to make use of the explosion. He noted that for the attack on the 25th, ‘a little more attention, but not nearly enough, was given’, but overall described the engineers’ plans as ‘half-hearted’.35 Burgoyne clearly understood that the mine was to be used ‘as a signal only and with the chance of alarming them [the French defenders]’.36 Jones makes no mention of an assault on the hornwork. Dickson’s view before the assault was that blowing the mine would ‘create such an alarm as may make them evacuate … and so produce a favourable diversion’, a view shared by Lieutenant Harry Jones.37 After the assault, Dickson noted, ‘A party of Caçadores [Portuguese Light Infantry] availing themselves of the consternation produced amongst the enemy … made … their way into the ditch … but the defenders … commenced a fire … which obliged them to make … their way back’.38 It would appear that Oman based his suppositions on the comment above from Dickson, which does not give any real indication that it was a pre-meditated action. There does not appear to be any evidence to back up his claim that an attack on the hornwork was planned and that it was badly organised by the engineers. It should also be noted that it was not the engineers’ responsibility to organise the troops for any attack, but that of the commander of the troops, so any blame should have been directed at Graham, not the engineers.
There are a number of criticisms of the delay between the first breach being practicable and the assault, thereby giving time for the defenders to reinforce the damaged areas. These criticisms are not helped by some confusion amongst the Ordnance officers themselves. Frazer complained in his letter of 23 July that ‘after [making] this excellent breach, they hesitate about using it … I am now ordered to make another breach … by which time the original breach will be entrenched’. His view was clearly that the failure of the assault was caused by ‘delay and indecision’.39 According to John Jones, the general plan as had been used in previous sieges was to open a second breach at the last minute to stretch the defenders. Frazer did not appear to be aware of this, perhaps because this was the first siege at which he was present. Oman and Fortescue both criticise the two-day delay between the first breach being ready and the assault. Fortescue in particular seized on Burgoyne’s remarks after the first siege where Burgoyne commented that the ‘whole of the batteries … were constructed on the right bank … giving them immediate insight into the nature of the attack … and the breach was practicable two days before the trenches’.40 Careful review of the dates shows that the trenches were ready on the morning of the 23rd,41 the breach was declared practicable the same morning and the assault was planned for the following morning. The two-day delay is calculated because the assault was then delayed for twenty-four hours due to the fires behind the breach. This delay may have been unfortunate and significant, but it was not due to the trenches not being ready. It is difficult to see how the work on the breaching batteries could have been delayed to hide the point of attack. They were started on the 13th, which was four days before the convent of San Bartolemeo was taken. It is unlikely that they would have all been ready on the 20th, if they were not started until the 17th, and this would then have lengthened the siege. In every siege in the Peninsular War, time was a critical factor. San Sebastian was no different and the decisions taken were to save time. Graham wrote to Wellington on 24 July, pointing out that the artillery had nearly run out of 24-pounder shot. If the assault had not gone in on the morning of the next day, the siege would have had to revert to a blockade, as there was insufficient ammunition to continue.42 One other factor that must be taken into account is the problems with working parties. The working parties absenting themselves on the night of 20 July cost the attackers twenty-four hours. They should have been ready on the morning of the 22nd, which would have been the day before the breaches were declared practicable.43
There were a number of criticisms around the assault on the morning of the 25th itself. The plan required daylight so that artillery support could be given to the assault. Dickson had told Graham that the artillery would be able to suppress the defenders’ fire during the assault. Graham’s official report stated that the attack took place ‘soon after daylight’, and Fletcher stated that the assault was given at daylight … the mine having been previously sprung’.44 However, the artillery officers recorded that the assault had failed before there was sufficient light for them to determine what was happening. Dickson stated ‘the column of attack certainly moved forward too early, either from a mistake … or from over anxiety on the part of the
directors’.45 Frazer was more forthright, writing ‘The assault was … made … stupidly an hour before, instead of after daybreak’.46 It is almost certain that the mine was blown before 5 a.m. as Graham’s letter to Wellington informing him of the failure of the assault was written at 5:30 a.m.47 In his biography of Sir Thomas Graham, Aspinal-Oglander strongly refutes the claim that the attack commenced before daylight, but seems to base his argument on the fact that Graham’s dispatch reported it was in daylight.48 While no account clearly stated who gave the order to start the assault, it is likely that Graham did. Even if he did not, he must, as commanding officer, still take responsibility for the failure.
General Oswald, the commander of the 5th Division, did not plan the actual attack well. Campbell of the 9th Foot was of the opinion that the troops were too extended during the assault and thought that if a compact mass had arrived at the breach they ‘would have bodily forced through all opposition’.49 He may have had a point about the organisation of the troops, but his approach would not have worked against the twenty-foot drop that the attackers were faced with in the main breach. Oswald’s plan, bearing in mind the concerns about the narrow area in which the assault had to take place, organised his troops so that those heading for the nearest breach went first followed by those who needed to pass the first breach to go to the second breach. With the failure of the assault on the first breach, the troops destined for the second breach could not get past the retiring troops and were swept back into the trenches with them.
Wellington’s Engineers: Military Engineering on the Peninsular War 1808-1814 Page 23