On 20 October 1809, Viscount Wellington issued a memorandum to the commanding Engineer in Lisbon, Lt.Colonel Richard Fletcher, instructing him to commence the construction of the defensive positions that became known as the Lines of Torres Vedras. Wellington had ordered the building of the Lines knowing that Napoleon must soon send a large army to invade Portugal and drive the British Expeditionary Force into the sea.

Wellington could not hope to hold back the French along Portugal's long land frontier. As Sir John Moore had observed the previous year, "it is an open frontier, all equally rugged, but all equally to be penetrated. " (1)

It was clear therefore to Wellington that the "great object in Portugal is the possession of Lisbon and the Tagus, and all our measures must be directed to that object." (2)

Wellington assumed that the French would advance into Portugal along two lines, one north and one south of the Tagus and the erection of any defensive works to cover the Portuguese capital would have to be below the lowest point on the river where troops could cross. The other determining factor for the location of the fortifications was the nature of the terrain in the area north of Lisbon.

The original idea of building a line of fortifications across the Lisbon peninsula was conceived by a Portuguese major, Jose Maria das Neves Costa. Towards the end of 1808 Neves Costa had examined the hilly districts north of the capital and the following June he submitted a report, with maps, detailing his ideas to Dom Miguel Forjas, the Portuguese minister of war. This information was passed on to Wellington in the autumn of 1809.

After conducting one all-embracing survey of this area in October 1809, in the company of Colonel Fletcher, Wellington formulated his plans and ordered Fletcher to begin work without delay. Fletcher was directed to construct a line from the mouth of the Castanheira brook to the mouth of the Zizandra, and another, a few miles behind the first, from Alhandra on the Tagus by Bucellas and Cabeza de Montechique towards Mafra. These approximately represent the two lines of defences ultimately built, though when completed the extreme right flank was drawn back from the Castanheira to the Alhandra stream.

Fletcher was left with sole responsibility for the construction of the Lines. He was assisted by Major John Jones, the original historian of the Lines, and, in addition, eleven British officers of the Royal Engineers, two from the King's German Legion, and three from the Portuguese regular army. Work on the Lines was begun almost immediately, and on 31 October Fletcher ordered 19,000 palisades and 10,000 facines to be prepared at Torres Vedras and along the Tagus. The man-power available to Fletcher was virtually unlimited. The work force was composed of both forced labour and volunteer workers. The forced labour came from the Lisbon militia regiments, who were brought up by alternate pairs, and a conscription from the whole of southern Estremadura for a circuit of forty miles around.

The hired volunteers came from the peasantry of the local districts, of whom from 5,000 to 7,000 were generally available and in the following months as many as 30,000 militia and peasants could be found employed on the Lines at any one time. A memorandum dated 1 July 1810 (3) details the following arrangements:

Location	Works	Men
First Line	32	10,040
Second Line	65	14,600
San Julian	11	3,850 (4)

The men were worked in gangs of some 1,000 or 1,500 peasants and militiamen, each under the command of an engineer officer with the assistance of a few English and Portuguese military artificers, and considerable reliance was placed upon the skills of the labourers. "In some districts a subaltern officer of engineers with a few English soldiers, utterly ignorant of the language, directed and controlled the labour of 1,000 peasantry, many of them compelled to work at a distance of forty miles from their homes, while their lands lay neglected.

Nevertheless, during a year of this forced labour not a single instance of insubordination or riot occurred. The great quantity of work performed should, in justice to the Portuguese, be ascribed more to the regular habit of persevering labour in those employed than to the efficiency exercised over them. (5)

Only four major paved roads led to Lisbon below the point where the Tagus, from its increased depth and breadth, becomes impassable to an army. Three of those roads, at nearly parallel points, pass over, or between, heights of considerable strength (Mafra, Montachique and Bucellas) and the fourth, bordering on the Tagus where the land lies much lower, passes under a strong range of heights at Alhandra. The terrain between the roads is hilly and broken and represented a considerable obstacle to an army with all its artillery and baggage.

It was not intended that the Lines should be a continuous linear masonry rampart like Hadrian's Wall or the Great Wall of China. The Lines were to be composed of chains of mutually supporting redoubts covering every dominant tactical feature and enfilading all approach roads and defiles. Rivers and streams were to be dammed to flood the surrounding countryside, roads blocked with abattis, bridges mined ad the intervening hillsides scarped to form an almost uniformly strong barrier through which the French would have to force their way by direct frontal assault before they could reach Lisbon.

Large sections of the ground selected for the Lines were so naturally strong that very little improvement was necessary. Along the first seven miles of the line, from Ribamar on the coast to the pass of Mafra, "a deep, rugged, and in many parts impracticable ravine leaves scarcely a favourable point for a battalion to advance in column." (6), and the destruction of some of the peasant's paths and the siting of just twenty guns in six small enclosed works was regarded as quite sufficient to render this section secure.

Mafra

The pass of Mafra, on the other hand, was strengthened with particular care, for although the main ascent itself is quite steep and therefore easily defended, the position could be outflanked. By the right-hand side of the pass two roads run along the boundaries of a large walled park (the Tapada). The roads, sheltered by the walls of the park, offered an attacker an ideal avenue around the flank of the main position.

To counter the possibility of such a manoeuvre the park was heavily fortified by adding a banquette for infantry to the interior of the walls and by blocking the roads with loopholed entrenchments. Four redoubts (Nos.74 to 77) were thrown up on the most commanding points within the enclosure of the park, and each feature of the ground overlooking the approach to the park was also occupied by a redoubt. Altogether, twenty- one pieces of artillery, including eleven 12- pounder cannon and 1,915 infantry were detailed just for this flank.

The main ascent of the pass was heavily entrenched with artillery and infantry positions and the town of Mafra was formed into a defensive post and covered by a chain of redoubts (Nos.85-87) that blocked the only lateral approaches practicable for artillery. Approximately three miles ahead of the pass, a little to the left of the village of Morgueira, three redoubts (Nos.82 to 84) were erected to cover the minor pass of Cacheca and to act as a forward bastion for the left-hand defences of the line.

From the pass of Mafra to the pass of Montachique the hills, "though not continuous or precipitous, are high, steep and salient". (7)

These hills were occupied by isolated redoubts (Nos.62 to 73) linked together by a lateral road that ran between the two passes. The heights forming the immediate flanks of the pass of Montachique were naturally so strong and favourable for defence that very little labour was expanded on them. Instead, considerable attention was devoted to blocking the road, with twenty-five pieces of artillery secured in six redoubts (Nos.52 to 61) situated on advanced features of the ground. These redoubts were built so close to each other that they formed a "chain of posts collectively stronger than the main features of the pass." (8)

The heights from Montachique to the pass of Bucellas "are of a nature to preclude the necessity of works." (9) Only the road over the ridge of Freixal offered an avenue of attack and this was blocked with three retrenchments mounting eight guns.

The road that runs through the narrow pass of Bucellas travels between two high and steep mountains and was therefore easily defended. The approach to the pass was guarded by redoubts 43 to 47, and on the bridges at the entrance to the pass were mined and the road blocked.

From Bucellas to the Tagus the Serra de Serves, a high and extremely difficult ridge, occupies a front of over two miles with scarcely any break before dropping sharply down to the low ground bordering the Tagus. This low ground stretches for two and a half miles before reaching the river and it required considerable strengthening. A cluster of redoubts (Nos.34 to 39) formed the main defensive barrier in front of Via Longa, situated roughly in the centre of this stretch of' land, with three more redoubts on the spur of the serra forming the left flank, and a redoubt (No.33) situated close to the Tagus, on the right.

Even though these redoubts mounted forty- seven guns with almost 2,560 men in their garrisons, this section of the line was still regarded as by far the weakest and considerable reliance was placed upon a chain of strong heights, forming almost an isolated feature, about five miles in front of the line at Alhandra. Here, to enfilade the main road and flank the low ground, were established redoubts Nos. 1 to 8 of twenty-one guns, including sixteen 12- pounders.

The whole line, twenty-two miles long, comprising forty-nine redoubts with 232 pieces of artillery and requiring 17,500 men to garrison the works, constituted the principal line of defence across the Lisbon peninsula. Some six to nine miles ahead of this line a series of' detached works were also constructed. Originally designed simply to block the approaches to the line and to delay the invaders whilst the Anglo-Portuguese army took up its positions behind the main defences, they developed into a barrier almost as strong as the line in their rear.

Outer Line

The first section of this outer line, from the Tagus at Alhandra as far as the great ravine that overlooks the village of Arruda, ran for five miles along the crest of a steep, but not very high, ridge. This front was elaborately fortified as it blocked the road which constituted the easiest approach to Lisbon from the north, running as it does along the low ground bordering the Tagus.

No less than twenty-three redoubts were placed here, mounting 96 guns and garrisoned with 6,000 men. A mile of land by the Tagus was flooded and at one point some 2,000 yards of hillside had been scarped into a long and continuous precipice.

The second section of the line, from the ravine above Arruda to the left of the steep Monte Agraca, had a front of four and a half miles which included the "most lofty and defensive part of' the backbone range of the Lisbon peninsula." (10)

However, another of the four main paved roads to Lisbon from the north passed over the shoulder of these heights and they had therefore to be heavily fortified. A large redoubt for 1,600 men and twenty~five guns built on the top of the Monte Agraca was the main defensive work with a further six redoubts garrisoned by 1,480 men and thirty guns.

The third section of the outer line was, by contrast, only lightly fortified as it had not featured in Wellington's original plan. This section stretched for eight miles from the left of Monte Agraca to the pass of Runa, overlooking the upper valley of the Zizandre and the village of Sobral. Such defences as there were consisted of only two redoubts which commanded the high road from Sobral to Cabeca de Montechique and the dominating redoubts on the Monte Agraca which overlooked it on the right.

There was a distinct possibility that the French might make a push up the valley along the high-road by the village of Zibreira which was regarded by some as "the most probable point of attack in the whole 29 miles of front for the enemy to select", (11) so when Wellington finally decided to hold the outer line, and not'just use it as a temporary rallying point, he had to cram this point with troops and hurriedly construct new works along it. Sobral, the village at the foot of the heights was held as an outpost but was too far to the north to be treated as an integral part of the main position.

The fourth section was that from the gorge of the Zizandre (or the pass of' Runa, as it is also known) to the sea. It was about twelve miles long but more than six miles of this front was covered by an impassable bog formed by the obstructed Zizandre and another mile was taken up by the formidable entrenched camp of San Vincente above the town of Torres Vedras from which the whole defensive system took its name. This stronghold lay outside the main line beyond the river, covering the bridge and the paved road from Leiria to Lisbon, the only road suitable for wheeled traffic on the western side of the Lines.

The whole front on both sides of Torres Vedras and its great fort was so strong and inaccessible that there was little danger of the French attempting an assault upon it, especially as any force operating in this area would be completely isolated from any other units in front of the eastern and central parts of the Lines.

Communications

A system was devised for communicating orders and intelligence from one end of the Lines to the other. There were five signal stations equipped with semaphores situated at important points along the defences.

The apparatus consisted of "a mast and yard from which balls [inflated pigs' bladders] were suspended; and there being no trained signallers available amongst the troops, Lord Wellington applied for and obtained the services of a party of seamen under their own officers, from the fleet in the Tagus." (12) The signal stations were at:

1. The redoubt No.39 near the Atlantic
2. The great redoubt of Torres Vedras
3. The Monte de Socorro above Wellington's headquarters at Pero Negro
4. The summit of the Monte Agraca
5. The hill behind Alhandra on the Tagus

After some practice it was found that a message could be sent from one end to the other of the twenty-nine miles in seven minutes, and from Wellington's headquarters to either end of the Lines in four minutes. There was a similar line of four semaphores on the second series of defences.

Garrisons

The Lines were garrisoned mainly by the militia and the Ordenanca, the primitive Portuguese home guard. Only those sections of the Lines that were directly threatened by the French advance were actually manned:

1 . In the Alhandra forts:

Militia Regiments of Santarem, Idanha, Castello Branco, Covilhas and Feira and the 12th Regiment of Infantry of the Line. 3829 men

2. In the Bucellas forts:

Militia Regiments of Lisbon, Thomar and Torres Vedras 1907 men

3. In the forts facing Sobral:

Atiradores Nacionas (embodied Ordenanca) 761 men

4. In the Torres Vedras forts:

Militia Regiments of Lisbon (W), Lisbon (E), Setubal and Alcacer do Sul. 2231 men

5. In the Mafra forts:

Militia Regiment of Vizeu. 691 men Militia Artillery 2886 men

Total: 11,092

When the field army took up positions behind the Lines the 1st, 4th, 5th and 6th Divisions were stationed along the Monte Agraca to Runa section. The 2nd Division, Light Division and Hamilton's Portuguese Division were situated along the Alhandra- Arrucla sector with only one division (3rd) stationed around Torres Vedras. In total, including the Spanish Army of Estremadura which had sought refuge behind the Lines, the fortifications were defended by 77,690 men.

All the redoubts of the two defensive lines were, in effect, little more than enclosed battery emplacements situated in positions where artillery fire was required to cover some particular point. The redoubts were placed so as to deny access along a road or to delay the repair of a bridge, or sweep the entry of a pass. Most of the redoubts "were perfectly independent of each other, and were made of a strength of profile to resist an assault, and placed on points where artillery could with great difficulty be brought to cannonade the." (13)

Those redoubts that were built to defend a particular feature of the terrain or point of communication were usually constructed upon the summit of the heights that they occupied, so that each face of the work would have a full command of the ground in its front or of the point it was intended to protect. These elevated situations also gave the garrison increased protection from direct enemy fire. Many redoubts, though, were placed on top of hills with sides so steep that the effectiveness of the defending artillery was considerably reduced and the muskets of the infantry were unable to scour the whole of the hillside below. Because of this, on some of the very steep hills, particularly on the Monte Agraca, additional small redoubts, or fleches, were established in front of the main work. These advanced batteries were made of the same strong profile in their front as the redoubts, and their gorges equally secured, except that the rear parapets were formed as mere screens so as not to give cover against the fire of the main work.

At the same points, where it was considered likely that the troops of the army would act in conjunction with the redoubts occupying the summits of the very elevated knolls, gun emplacements were prepared for the field artillery on the best flanking or enfilading situations much lower down on the face of the hill. With some redoubts, where there was insufficient internal space to permit the mounting of artillery within their walls, the guns were placed on lower advanced levels which were connected at their flanks with the defences of the main works.

Where the purpose of the redoubts was simply to prevent an enemy occupying a certain spot, they were positioned on an inclined plane on the reverse slope of the height so that only its front face rose above the crest of the hill, giving the work more protection than if it sat exposed on the summit.

The actual shape and size of each readout was determined by the nature and importance of the position which it was to defend and the amount of available ground space. As a result there were a great variety of different constructions. Many of the first redoubts to be built were star-shaped (Fig. 1) so as to permit a flank defence for the surrounding ditch. This design was later rejected as it restricted the interior space of the redoubt and it did not allow for the concentration of a large number of guns on any single front.

The later redoubts were more individual in design, with extensive faces for artillery positions along their fronts (Figs.7 & 8). The great works at Torres Vedras and Monte Agraca were by far the largest of all the redoubts and both had secondary defensive positions, or retrenchments, within their walls. Each of the salient angles of the Torres Vedras redoubt (Fig.3) were formed into independent posts ad the Monte Agraca redoubt (Fig.2) had its most salient points cut off by internal earthen traverses. These interior defences were intended to serve as rallying points and to "prevent the loss of the work by the entry of the assailants at any weak or ill-defended points." (14) Many of the small, circular stone windmills that were found throughout the region were also incorporated into the defences some within the walls of the redoubts as observation posts or secondary strongpoints, others being converted into advanced lunettes (Figs.4 & 5).

Although the profile of the different redoubts varied on every face and flank, depending on whether it was likely to be stormed by infantry or bombarded by artillery, a number of general principles were adopted in the construction of all the redoubts. All ditches had to be at least fifteen fcet wide at the top and ten feet deep. No parapet was more than ten feet thick unless it was likely to be exposed to heavy bombardment (the redoubts on the heights of Almada which stood in "situations open to be violently cannonaded" (15) were built with parapets fourteen feet thick).

Some redoubts that were built on high knolls where artillery could not "by any possibility" be brought to bear upon them, were made of stone only two feet thick in order to gain more interior space. In many of the very elevated situations the banquettes were raised to within four feet of the crest of the parapet to allow the defending infantry to lean over the parapet when engaging the enemy. The angle at which the exterior slopes were cut was entirely dependent on the firmness of the soil, but after the first winter it was found that at any angle greater than 45 degrees the soil was simply washed away by heavy rain.

By 1811 most of the exterior slopes of the redoubts were retained with dry stone walls. The interiors of the parapets were retained with facines and sandbags, but many of the sandbags rotted and burst after the first winter.

The first line of defence for each redoubt was a row of abattis. These were formed from whole trees, trimmed to a point and with their small branches removed so that the front of the abattis "afforded neither cover nor concealment to an assailant, although it presented a barrier of spears, five six and seven feet in height." (16)

The abattis were usually placed from twenty to thirty yards in front of the work, each stem and large branch being firmly staked into the ground. Behind the abattis were dug trous-deloup, consisting of between eight and ten rows of pits some two feet to two feet six inches deep, well staked at the bottom and in the intervening ground.

The number of troops required to garrison the redoubts and retrenchments was set at a figure of two men per yard of frontage, deducting for the spaces occupied by the artillery as it was estimated that each man required three feet of room to be able to wield his musket effectively, this figure ensured that each parapet was adequately manned and still left a strong reserve in each work to replace those that were killed, or to charge any attackers that broke through the main defences.

The scarped hillsides were made by cutting the front slopes of ranges of heights near their summit as perpendicularly as the soil or rock Would permit. Fig.6 depicts a section of scarpe nearly two miles in length, formed along the summit of the front of the position at Alhandra in August and September 1810. Much of the upper twenty or thirty feet of that range was found to be a ledge of precipitous rock only covered with a few feet of earth. The rock face was easily made completely insurmountable by blasting with gunpowder.

French Invasion

The French invasion, when it finally came, had been delayed so long that the autumn rains had begun to fall as the invaders approached the Lines. The rains, as they do each year, continued for many days. Much of the low lying ground became boggy swamp, unpaved mountain paths disintegrated into quagmires and streams which were "nearly dry are now so deep as to be impassable." (17)

This restricted the French line of advance to the main paved roads which encouraged Wellington to make a determined effort to hold back the invaders at the outer line. Improvements to the defences of the outer line had been put in hand as early as July. On the extreme right by the river Tagus, the musketry trenches (Nos. 1 and 2) that ran across the between the river and the heights of Alhandra, were converted into strong lines, flanked by powerful batteries that were established on the side of the mountain. These batteries were completely covered from all ground in their front, could not be cannonaded, or even seen by an enemy, till almost on the glacis of the defences across the low ground, and consequently rendered any effort to force that line utterly hopeless. The Alhandra heights themselves were made "strong to an excess".

Along the face of the heights, near the summit, an almost perpendicular scarp was cut or blasted from the solid rock to a depth of some eighteen feet, every part of which was closely flanked by musketry or enclosed artillery positions. The enclosed works were themselves overlooked by larger and stronger redoubts (Nos. 114 to 120).

The Alhandra position was slightly in advance of the rest of the line and although its right flank rested securely upon the Tagus its left flank was exposed. To protect this open flank, work on four new redoubts was commenced on the heights above Calhandrix. In order to block the valley between the heights of Alhandra and Calhandrix, and to connect these two positions, a strong abattis, with a covered road in its rear, was formed across the valley, the front of which could be swept with artillery fire from the works on the Alhandra. Time was also found for the construction of redoubt No. 125 which connected Alhandra with the Serra de Serves. The ground near the summit of the Serra was scarped in the same way as the Alhandra heights.

On the left flank of the line additional redoubts were thrown up to form a chain along the left bank of the Zizandra. In summer the Zizandra is an "insignificant stream" but after the autumnal rains it becomes a full, flowing river. Obstacles were placed in the bed of the river, at points within range of its flow. In a very short time the river overflowed its banks and more than half the valley became "so complete a bog that ... that portion of the front which in summer had been the weakest, became, during the winter, in some degree secure from attack." (18)

In the central section of the outer line strong redoubts were thrown up on the commanding points around Ribaldiera and Runa (Nos.128-130) and the valley in the rear of the Zibriera was blocked up by a well-flanked abattis, with field batteries established on various flanking points and roads of communication formed to them.

A redoubt, armed with 9-pounders, was also commenced on the ascent of Monte Agraca, on a lower level, and to the right of the main work, to enfilade and block the main road from Sobral. Another redoubt was built above Matacaes to cover the road through the pass of Runa and the heights above Portella and Patameira were scarped and occupied by new redoubts (Nos. 150 and 15 1) for field artillery.

When complete, in 1811, the Lines comprised 152 works, which mounted 534 guns required 34,125 infantry to man its ramparts.

Footnotes

1. Moore to Castlereagh, 25 Nov. 1808
2. Memorandum to Lt.-CofFletcher 20 October 1809
3. An Engineer Officer Under Wellington in the Peninsula, by Lt. Rice,lones p.56
4. The Bay of St..Julian was the embarkation point for the arniv in the event of a Fren(h victorv. The bav was protected by a semi-circle of' defences 5,300 metres in length and included thiteen separate works with ninety- four pieces of artillery, and designed to be held by 5,.i5O men. Across theTagus at Almada was a fourth line of seventeen redoubts. These were built with the intention of preventing the French from bombarding Lisbon and its harbour if they succeeded in delivering an attack south of the Tagus.
5. J.T. Jones, Memoranda Relative to the Lines Thrown up to cover Lisbon, p.79 6. Ibid p.8
7. Ibid p. 11
8. Ibid p. 10
7. lbid p. 11
10. C. Oman, A History of the Peninsular War Vol. 6 p.425
11. Oman Vol. 6 p.421
12. Rice.Jones p.54
13. J. Jones p.73
14. The dimensions of these works were:-

Works	Feet	Inches
Height of interior crest of parapet	7	0
Height of parapet above banquette	4	3
Thickness of parapet	14	0
Breadth of ditch at top	16	0
Depth below surface of the ground	12	0
Crest of glacis below crest of parapet	5	6

15. J. Jones p.83
16. J. Aitchison, An Ensign in the Peninsular War p.86
17. J.Jones,p.23

Quotations from An Engineer Officer Under Wellington in the Peninsula by kind permission of Ken Trotman Ltd.

Torres Vedras Fortifications Photos (slow: 188K)