There is a plethora of studies regarding applications of photogrammetry in the discipline of archaeology. Some of them are focused on the recording of archaeological features and excavation trenches (De Rue et al., 2014), (Doneus et al., 2011), (Benko et al., 2004), others spotlight the recording of buildings and smaller archaeological artefacts (Kersten & Lindstead, 2012). These studies are partially related to this work, although certain publications specifically target larger monuments.

This thesis benefited from studies performed on castles (El-Hakim et al., 2007; 2004), caves (Lerma et al., 2010), (Chandler et al., 2005) and temples (Wiegmann, 2016). It is important to stress that most of the available research is focused either on the data integration, uses of combined techniques or comparison between photogrammetry and laser scanning. It is also expected to favour aerial photography in a large-scale survey project, when possible (Remondino, 2013). Nevertheless, the above studies are still relevant to the framework of this project.

One of the earliest works concerning survey of a complex building, was a photogrammetric and archaeological recording of Windsor Castle (Dallas et al., 1995). In 1992 the monument was partially destroyed by fire. The priority was to assess the damage and salvage the remains of original features. The photographic survey was carried out by English Heritage. The work was conducted according to Specifications for Architectural Photogrammetric Survey of Historic Buildings and Monuments by Dallas (1988). Due to health and safety reason the recording needed to be performed quickly, which was one of the key factors contributing to the choice of photogrammetry. The obtained data proved to be of significant importance and was exploit by a range of specialists involved in the restoration process (Dallas et al., 1995: 237)

Detailed 3d Modelling of Castles (El Hakim et al., 2007) and 3d Reconstruction of Complex Architectures from multiple Data (El Hakim et al., 2005) are other important publications regarding digital documentation of complex heritage sites. These studies aimed to capture detailed and complete geometric images of several castles located in Northern Italy. It explored different approaches to digital 3d modelling, including integration of photogrammetry and laser scanning.

According to the authors, the average time of data acquisition for each castle was no longer than 3 days. The time needed to process the obtained data varied between one and two months per castle.  The final 3d models achieved accuracy measured in the range of millimetres. El Hakim et al (2004) also implemented photogrammetry methods during the modelling of the Benedictine Abbey of Pomposa near Ferrara, Italy.

Another study regarding applications of architectural photogrammetry in modelling and visualisation was focused on the survey of north German castles, by Kersten, Padro and Lindstaedt (2004). Kersten (2006) used photogrammetry for large scale buildings in his comparison studies to terrestrial laser scanning. He proved this technology to be 17 times cheaper than laser scanning. Digital and analytical photogrammetric recording was successfully performed on St Domingo de Silos Church in Alcala La Real, Spain (Mata et al., 2004). The survey was commissioned as a part of a church restoration project and was applied both outside and inside of the building. The 3d model created was of very high metric quality and was sufficient for heritage building cataloguing. It was also used during the restoration process. Similar work was carried out at the Acropolis of Athens by Moullou and Mavramati (2007).

Chapter 1:   Photogrammetry

1.1                 Historical background

The evolution of photogrammetry as we know it today, is often divided into 4 phases, and is strictly associated with the technological advances made in photography, aviation, computers and electronics (Schenk, 2005).

1.1.1                     Before photogrammetry

Photogrammetry is as old as the beginning of the modern photography, but the study of its mathematical and procedural concept can be traced back to ancient times. The optical projection was of interest to Aristotle in about 350 BC.  At the end of 15th century, Leonardo Da Vinci built the theoretical foundations of modern photogrammetry, by describing the principles of optical perceptivity (Ghosh, 1988). Photogrammetry was a concern of Albrecht Durer, Girard Desargues, Johannes Kepler, Joseph Nicephone Niépce, Johan Heinrich Lambert and many others. Durer for example created an instrument which helped produce true perspective drawings (Ghosh, 1988).

1.1.2                    Phase 1. Early developments

The revolution started in 1837, when Jacques Mande Daguerre developed a technique called Daguerreotype. A new method produced the first, practical photograph. It was also advocated by Dominique François Jean Arago. Remarkable achievements in terrestrial photogrammetry were performed in 1849 by Aime Laussedat who used it for map-making (Ghosh, 1988).

The revolution matured during the second half of the 19^th century. It has been marked by a rapid development of aerial photography. The first images were taken in 1855, by balloonist Gaspard Felix Tournachon, known as Nadar. It was quickly discovered that this method could be used for military reconnaissance. Kite aerial photography was explored in 1882 by E.D Archibald. 30 years later Julius Neubranner presented a breast-mounted camera, design to be carried by pigeons (CFPT, 2017). Photogrammetry benefited from the inventions of Ignazio Porro. His design of photogoniometer addressed the problem of lens distortion. His later introduction of three asymmetrical lenses remarkably improved the quality of photographs. Among his creations were also the tacheometer and the panoramic camera (CFPT, 2017).

The term “photogrammetry” was first used in 1893 by Albrecht Meydenbauer, the founder of the Royal Prussian Photogrammetric Institute. It took its name from the Greek words phôs / photos, light, grammameaning something written, and metronmeaning to measure (ISPRS, 2017).

In 1867 Maydenbauer developed his own camera equipped with wide angle lens, which was employed in his architectural surveys (CFPT, 2017).

1.1.3                    Phase 2. Analogue photogrammetry

The second phase took place between 1900 and 1960 and is often referred to as analogue photogrammetry. It is characterised by two main inventions: stereogrammetry and the aeroplane. Employed by the army, it was clearly a time of rapid development for the technology.

An historic breakthrough occurred when a stereoscopic-plotting device was constructed in Canada by Edmund Deville. His instrument called Stereo-Planigraph was used to survey mountains. The Austrian inventor Theodor Scheimpflug introduced the theory of the double projector and concept of radial triangulation. His development of multi lens camera was the first one to succeed in practical mapping.  The first publications regarding the principles of double- image and the methodology of orientation were published by Sebastian Fisterwalder. Work on the first photogrammetric instrument, the stereocomparator, were started simultaneously by Carl Pulfrich and Henry Gorge Fourcade. It was later manufactured by Zeiss (CFPT, 2017).

These discoveries lead the way to the invention of stereoautograph by Eduard von Orel. Other significant contributors to the development of analogue rectification and stereoplotting were; Max Gasser, Umberto Nistri and Reinhard Hugershoff. Progression towards analytical photogrammetry was done by Otto von Gruber and Earl Church. The twentieth century was also marked by the development of manufacturing techniques and new brands of photogrammetric instrument, including Fairchild Aerial Camera Corporation founded by Sherman Mills Fairchild (CFPT, 2017).

1.1.4                    Phase 3. Analytical photogrammetry

The way to analytical photogrammetry was paved by work of Finsterwalder, Gruber and Church, but really begun with computerisation. Early work on a semi-analytical method for analytical control came from Ralf O. Anderson (Doyle, 1964). In 1951, Everett Merritt produced a series of numerical methods regarding space resection, camera calibration, and interior and exterior orientation. Merritt (1958) later collectively describes it in his book Analytical Photogrammetry. Other great works of this time includes the development of multi station analytical photogrammetry by Helmut Schmid, analytical control extension by Paul Herget and the theory of photo-triangulation by Duane Brown (Doyle, 1964).

1.1.5                    Phase 4. Digital photogrammetry

Digital photogrammetry is relatively new and in contrast to old analogue photography is based on digital images.  The pioneers in this discipline include Gilbert Louis Hobrough and Uki Helava. Hobrough contributed in building electronic dodging printers, airborne profile printers and the Raytheon-Wild B8 stereomat. The stereomat was designed to “correlate high-resolution reconnaissance photography with high-precision survey photography in order to enable more precise measurement of changeable ground conditions “(CFPT, 2017: 345).

1.2                 Photogrammetry today

1.2.1                    Introduction

Probably the most common definition of photogrammetry is the one introduced by American Society for Photogrammetry and Remote Sensing. It states that Photogrammetry is “the art, science, and technology of obtaining reliable information about physical objects and the environment through processes of recording, measuring and interpreting photographic images and patterns of recorded radiant electromagnetic energy and other phenomena” (ISPRS, 2017). Depending on the method of obtaining the images this technique can be classified as either “aerial” or “close range” (Krauss, 2007). Both types are widely employed by archaeologists. Aerial photogrammetry, as the name suggest, employs the photographs taken from elevated position, typically aeroplane, unmanned aerial vehicles, balloons or kites. It is used to create digital elevation models (DEMs) and maps (Wolf, 1974), (Fryer et al., 2007). Close range photogrammetry is based on images obtained from the ground level, where a hand held or tripod mounted camera is relatively close to the photographed object (Boberg & Legerqvist,1997), (Luhmann et al., 2006).

Photogrammetry can be described as metric or interpretative. Metric photogrammetry is based on measurements derived from images, with applications in the field of mapmaking or creating orthophotos. Interpretive photogrammetry is focused on recognition, analysis and identification of objects (Wolf, 1974).

One of the most important, recent developments in photogrammetry, is a technique called Structure from motion (SfM).  This method is used to reproduce 3d geometry from the series of overlapping, offset photographs (Appetecchia et al., 2012). The fundamental advantage of this technique is its ability to automatically decode the camera position, geometry and the orientation of the scene (Westoby et al., 2012). The software used in this study was based on the SfM technique therefore all processes leading to creation of the 3d model could be performed with minimum interaction from the user (Agisoft LLC, 2017).

1.2.2                    Alternatives to photogrammetry

There are two, main, 3d modelling techniques successfully employed by archaeologists; laser scanning and photogrammetry. Both techniques are closely related although both use different methods in obtaining the data.

Laser scanning, as the name suggests, is a remote sensing technology, which uses beams of light to measure distance between scanner and illuminated object (Wehr & Lohr, 1999). There are many laser scanners available on the market, often operating on different principles and in different conditions. Scanners mounted on the airborne devices, referred to as Lidar (Light Detection and Ranging) are used to produce digital elevation models (DEM) (Boehler & Marbs, 2004). The obtained data is of a very high resolution, typically measured in centimetres (Slawik & Zaplata, 2011), (Bewley et al., 2005), (Kraus& Pfeifer, 1998). The scanning is not ‘’visibility’’ dependant therefore can be performed in complete darkness (Baltsavias, 1999). One of the biggest advantages of this technology is its ability to produce data with filtered out vegetation obstructing the features (Devereux et al., 2005; 2008).  The Lidar system is currently at its prime. It has contributed to the discovery of forgotten cities lost in the jungle (Chase et al., 2015) and was used to revise the archaeology present at the World Heritage Site of Stonehenge (Abbot & Anderson-Whymark, 2012)

Laser scanners can be used on the ground when hand held or mounted on a tripod. This type of scanner is referred as TLS (Terrestrial Laser Scanning) (English Heritage, 2011), (Engstrom & Johanson, 2009). This approach is especially suited for a building or cave recording (Lerma et al., 2010). TLS has typically range of 70-80 metres, and can produce extremely dense clouds during a very short period (Koch & Kaehler, 2009). 3d models built using TLS are characterised by precision accuracy often not achievable by other techniques (Baltsavias, 1999).

Laser scanning is then the obvious choice for sites where speed and accuracy are a priority. That unfortunately comes with a very high price tag. The initial cost of the laser instrument can be as high as hundreds of thousands of pounds, and requires a qualified person to process the data. Laser scanning can also be affected by some slight errors regarding distant measurements and interpretation of edges (Koch and Kaehler, 2009). TLS was successfully employed to study lithics (Lin et al., 2010) and pottery (Karasik and Smolinski, 2008).

It seems that Laser Scanning and Photogrammetry share common ground and both techniques suffer from different limitations. It goes without saying that the only way ‘’not to compromise’’ is to integrate both techniques. Such approaches have already been executed in the field of archaeological building recording (Alshawabkeh, 2006), (Lerma et al., 2010), (Plets et al., 2012), preservation (De Reu et al., 2013) and reconstruction (Fatuzzo et al., 2011). The accessibility of the laser scanning technique to the average archaeologist is very limited. There is no such thing as ‘’low budget laser scanning”. Nevertheless, in some circumstances, this method can be described as the most affordable and efficient, especially when working with vast, complex areas.

1.2.3                     The applications of photogrammetry in archaeology

The photogrammetric methods with its ability to record fine geometric details was quickly explored by the discipline of archaeology. This technique was employed in the field, as a mean of recording of archaeological excavations and features (De Reu, 2014), (Doneus et al., 2011), (Benco et al., 2004). Tsiafaki and Michailidou (2015: 38) described 3d technologies employed to record the progress of excavations as a factor “limiting the destructive nature of excavation”. They also stress that 3d modelling in conjunction with GIS places the excavated archaeology into a much bigger picture (Tsiafaki & Michailidou, 2015: 39-40). The value of 3d models has been appreciated in the recording and study of artefacts (Kersten & Lindstead, 2012), (Karasik & Smolinski, 2008), (Lin et al., 2010). It was performed in conjunction with panoramic images (Luhman et al., 2006) and macrophotography (Gajski et al., 2016). It has proved faster and more accurate than traditional methods of drawing (Doneus et al., 2011), (De Reu, 2014) and has the advantage of being produced in a digital format. Photogrammetry survey can also be performed underwater, and is widely used in maritime archaeology (Drap, 2012), (Martorelli et al., 2014).

There is a demand for highly accurate 3d models in the field of cultural heritage management, preservation and conservation (Ogleby, 1995). First of all, it would improve the standard of the records. There is a constant threat to the world heritage sites caused by either military conflicts, natural disasters or simply human negligence. Accurate 3d representations of archaeological sites or objects provide the opportunity to assess and often prevent further damage. The most common factors affecting heritage objects are weather, vegetation and destructive animal activity. It seems logical that standard methods of documentation, including drawings, 2 dimensional photographs and total station surveys will not provide enough details necessary to restore and reconstruct archaeological site, buildings or artefacts.

3d modelling is also used to create replicas of the heritage sites and objects. It was successfully implemented in places badly affected by excessive tourism, including sarcophagus of Seti I (Factum Arte,2017), and tomb of Tutankhamun (Factum Foundation, 2017).

The importance of 3d modelling using photogrammetry, outlining technical details and procedures employed in cultural heritage has been covered by Stylianidis and Remonido (2016). Several case studies regarding heritage protection were published by the Masovian Voivodeship Heritage Protection Officer. This work covers large scale monuments, various size artefact and archaeological excavations (Wisniewski & Ostrowski, 2016). Photogrammetry and the technique of SfM proved to be useful as a monitoring tool in the hands of Heritage Inspectors (Wisniewski, 2014).

3 d modelling is also a main tool in visualisations. It aims to engage the public and local community. The need for digitization in this environment is not new and can be traced back to the 1970s (Gionizzi et al., 2013). It took another 20 years for 3d modelling to be utilized by digital museum exhibitions. It finally leads to online museum displays, and virtual reality presentations (Patel et al., 2003), (Vlahakis et al., 2003). Nowadays online presentations on social media seem to be the way to engage people in the most efficient way.

Ideally the approach of archaeology to the outside audience should mature further. It should not be limited to the most significant finds or monuments. The published archaeological reports are more likely to be found boring or difficult to understand by non-archaeologist. The solution to this can be a wide spread of 3d models. If “one picture is worth a thousand words” then “one 3d representation is worth a thousand pictures”.

Digital data is already widely shared among researchers but is not widely available. In that respect the application of 3d modelling should participate in democratization of archaeological knowledge, enhancing public participation in archaeological interpretations (Reilly & Rattz, 2005).  So called crowdsourcing could be implemented in photographic image collection for further 3d model generation (Smith, 2014). Photogrammetric method is a non-invasive technique, which can be performed by literally everybody with a basic knowledge of photography. It could be recommended to engage the public in their own photographic surveys. The popularisation of archaeological knowledge is necessary to raise awareness of the importance of archaeology. That could be achieved by 3d modelling using low budget photogrammetry.

The list of 3d modelling applications is very long and it seems that the only restraint for this technique of recording and documentation is lack of knowledge regarding software and handling of large scale 3d data.

1.2.4                      The accuracy of photogrammetric recording using SfM

The accuracy of low cost photogrammetric survey has been subject to many studies and was proven to be comparable to these performed by terrestrial laser scanning (Wiegman, 2016). The “assessment” approach was used during the excavation of a multi- period hillfort in Schwarzenbach, Austria. The technique of SfM was employed on site and evaluated for the accuracy of created surface models. The results were comparable to TLS, with no more than 4cm inaccuracies at the 95% confidence intervals (Doneus et al., 2011: 81). The results were impressive, considering that Doneus used a very early version of the Photoscan 0.8.1, released in January 2011.

Only few years later De Reu et al. (2014) published their work on image based 3d reconstructions in archaeological excavation practice. The case study was a wetland area associated with the Cistercian Abbey of Bundello, in Belgium. The project was focused on generating metric 3D models of excavated trenches. The software used was again a PhotoScan, version 0.9.0 released in 2012.

The level of accuracy compared to Doneus study, improved significantly. Depending on the coverage of the studied area the error was no greater than 1.5 cm (De Reu et al., 2014).

A similar study was conducted in Spain by Robleda Prieto and Perez Ramos (2015) in which they assessed low cost photogrammetry used to record the front of the Church of Santa Maria Azogue. The image processing in that case was done using relatively a new version of PhotoScan, released in 2015. The accuracy of their model using updated software was in the range of 8.7 mm.

The evidence strongly suggest that the accuracy of low cost photogrammetry can be more precise than TLS, and without doubt suffers less error than manual drawing techniques. There are correlations between the improvement in accuracy and the development of the PhotoScan software. That in theory should have heralded the dawn of a new era for archaeological recording.

1.2.5                    The advantages and limitations of photogrammetry and SfM

It is essential to say that the quality of the 3d model is associated with the value of the 2d images used to create the output. It is the surveyor’s choice to use appropriate image resolution depending on the purpose of the work. In theory every object can be subjected to successful photogrammetric survey if it is recorded by overlapping images. Obviously, there are some strict rules, every amateur of photogrammetry must follow but these can be found in manuals, work specifications, and guides to good practice available online (ADS, 2017), (Agisoft LLC, 2017).

The main advantage of the photographic method is its precision. As mentioned above the accuracy of this technique is comparable to TLS or even greater (Wiegmann, 2016). The survey itself is non-intrusive, and there is no need for physical contact with the recorded monument or object.

This technique can be very cheap. The process of image acquisition can be performed using non – commercial equipment, in theory good quality images can be obtained using modern smartphones. The progress in technology means high powered hardware, slicker and more reliable software. As computers become more powerful, the time necessary to process the images can be reduced significantly, compared to early stages of digital photogrammetry. There is also more software to choose from, many of them available free of charge. These computer programs are often very simple to use, and process models almost automatically. Photogrammetry is also one of the most efficient methods of recording. Data acquisition can be performed relatively quickly and the output can be shared and manipulated by other researches.

There are some drawbacks associated with this technique. To perform successful photogrammetric projects, the obtained images need to be overlapped at the minimum rate of 60 to 80 % (Agisoft LLC, 2017). Depending on the resolution of the photos, and the distance from the object, the survey can produce large quantities of photographs. Furthermore, it leads to longer processing times and creation of oversized datasets. Complex 3d models are not easy to handle by underpowered computers, and are problematic in terms of storage for archiving.

Another important factor is a problem associated with the quality of input images. Assuming that the photographs were obtained using high quality equipment, there is always a chance that the image will be affected by flare from the sun, under or overexposure, shadows in the photos, or back/front focusing.  Changing light conditions, absence of complex shapes and textures, may affect the ability of software to produce accurate models (Agisoft LLC, 2017). It is also impossible to assess the final output of photogrammetry survey on the site. This uncertainty means potential need for re-surveying, which can be impossible when recording progression of archaeological excavations.

To obtain the highest level of accuracy and quality of 3d models it is recommended to use high specification equipment. That considerably increase the cost of the project. The full licence for more sophisticated software, such as PhotoScan used in this study, is available for a fee. Therefore, it could potentially discourage a non-specialist user from performing photogrammetric survey.

Chapter 2:   Southampton’s Town Wall and Defences

2.1                 Archaeological and Historical Background

2.1.1                    Roman Period

The history of Southampton is at least two millennia long, during this period the location of the historic town moved several times.

Southampton existed as a port during the Roman period, with a settlement located at Bittern Manor, now a suburb of the city. It is usually recognised as ancient Clausentum. The Roman town was enclosed by a ditch and bank, later replaced with a stone wall built primarily of flint (Kilby, 1997).

The town encountered rapid changes in 7th century AD. The Roman settlement was abandoned and a new town was established in the region of present-day St Mary’s. The new settlement, called Hamwic was defended by the Itchen River shore and a ditch on the north-west side. In mid-10th century, the centre of this town shifted towards the area of later medieval occupation (Russel & Smith, 2009).

2.1.2                    The Late Saxon Period

The area enclosed by the medieval town wall contained a number of Late Saxon ditches. The defences extended beyond the later Norman defences, reaching the line of the area known today as Above Bar Street. The presence of Saxon settlement was also recorded east of the medieval town wall, at the Supreme Warehouse site; SOU 395, SOU413 (Smith, 1994:2), Canal Walk site; SOU 1482 (Cottrell & Elliott, 2011), and was confirmed absent at the Oceans Boulevard site, Britton Street; SOU 1316 (Smith, 2005), SOU 1366 (Smith, 2009). The Saxon defences were exposed during the excavations at the Lower High Street.  The ditches often do not coincide with extent of Late Saxon settlement, suggesting that these features were in use at different periods (Russel & Smith, 2009).

2.1.3                    The Norman Period

The Norman conquest brought changes to the Southampton’s defences. A new castle was erected at the north-west corner of the town. The importance of Southampton suggests that the structure would have been built in the late 11th century (Kilby, 1997). The castle was also mentioned in the agreement between King Stephen and Prince Henry, dated 1153 (Davies, 1883). The archaeological excavation of SOU 29, SOU116, SOU123 and SOU124, at the Southampton Castle concluded that the structure could have been erected between early 12th to mid-12th century (Oxley, 1986).

It is believed that Southampton also received new fortifications in the form of a ditch, earth bank and a wooden palisade (Rance, 1986). According to Platt (1973), the construction did not start before 1202, which would be roughly at the same time or slightly later than the construction of the Bargate.

The lack of proper defences for nearly a century is debatable, therefore some researchers move the date of the construction towards circa 1100. The archaeological evidence seemed to confirm the existence of earlier structures below the 13th century walls. The excavation at the York Building site, revealed an earthen rampart and ditch, hidden beneath the later 13th century stone wall (Russel &Smith 2009). The historic documents state that in 1202 Southampton received a £100 tax rebate from King John towards the “walling of the town”, a further £200 was allocated for unspecified work on Southampton Castle (Davis, 1883). It is unclear to what extend the town fulfilled the king’s order. It is known that for the next few years Southampton become port to the royal galleys, indicating the vulnerability of the town to attack from the sea (Rance, 1986:38).

The town ditch was also mentioned in documents dated 1220 onwards. From 1260 two murages were granted by Henry III, allowing toll collection, for the repair and construction of the town wall. Another, 3-year murage was granted in 1321, about that time the Bargate would have been flanked by new towers (Davies, 1883). The documents of that time describe the east wall as still ‘new’, and give information about taxes collected to finish the walls and the quay.

It is certain that the 13th century town struggled with the expenses of new defences. In 1246 the people of Southampton were fined for” withdrawing several duties, belonging to the castle” (Rance, 1986: 38). Archaeological excavation of SOU 272, revealed that Arundel Tower and much of the eastern wall were constructed with insufficient footings (Smith, 1993b).

2.1.4                    The beginning of the hundred Year War and the destruction of the town

At the time of the French raid, only the northern and eastern sides of the town were surrounded by the wall. It is likely that wealthy merchants were oppose to any plans of building walls on their private quays (Russel & Smith, 2009). The French attacked on October 4th, 1338. Fifty French Galleys landed on the south beaches of the town, encountering little resistance.  Southampton was sacked, much of the town including the castle was burnt (Davis, 1883).

After the invasion, the king ordered a stone wall built facing the sea. The work started in 1339 and further murages were granted to aid the construction of new defences (Page, 1973). To save money, large area of the walls incorporated earlier buildings. Southampton suffered a shortage of people capable of defending the town. Not without significance was the outbreak of pandemics. The Black Death was brought to Southampton in 1348 (Neal, 2014).

Extensive modernisation took place during the reign of Richard II, that included refurbishment of Southampton Castle and the alternation to the angled front of the Bargate (Davies, 1883). The town also added cannons to its defences. The openings in the southern wall represent some of the earliest gun ports in the United Kingdom (O’Neil, 1960).

2.1.5                    Late Medieval period

The history of the town defences in late medieval period is largely covered by the Terrier of 1454 published by Burgess (1976), annotated in to the early 16th century. The terrier provides information about the people responsible for the maintenance of certain parts of the town wall. It also names most of the town defences, e.g. towers and gates. On the other hand, the document fails to mention some of the existing features, including the Castle walls and refer to the Friars Gate as a square tower. During this period Southampton was used as a military base from which ships were sent to war against France.  It is certain that some work had been started on Gods House Tower during the reign of Henry V (Neal, 2014). As a matter of fact, recent archaeological excavation at SOU 1754 site, exposed remains of the earlier, stone-made structure, buried underneath the existing tower (Russell, 2017).  In 15th century Catchcold Tower was added to the earlier town wall, located to the north-west. Some alteration was also done to Southampton Castle (Page, 1973).

2.1.6                    Tudor and Stewart period

The construction of Henry VIII forts at Calshot, Hurst and Netley made Southampton’s defences redundant (Rance, 1986). A garrison was still kept to maintain the guns of Gods House Tower. The castle was still in use by 1599 but shortly after became obsolete. It fell in to ruins in 1618 and by 1650 the stones from the castle were used to repair ‘other fortifications’ (Davies, 1883). During the civil war Southampton was a stronghold for Parliamentarian forces against royalists. At that time, the town ditch was re-cut and arrow slits on the Bargate towers received new openings for guns (Russel & Smith, 2009).

Later plans of the town, prepared by Mazell in 1771 and Wooley in 1791, show a temporary defence, about 100m south of East Street. This feature was likely constructed during the civil war and was exposed by archaeologists at SOU627 site (Garner, 1994).

2.1.7                    The early-modern period

In the 18th century, the town walls lost their defensive function and, for the people of that time, seemed to suppress the growth of Southampton. Fortunately, the town wall was too substantial to be easily demolished. New dwellings were adjoined the wall or were build inside the towers. Despite that the Corporation managed to demolish Eastgate in 1775. An act of Parliament, the “Harbour Act of 1803” lead to prompt demolition of all defences obstructing access to the docks, including the southern walls and Watergate (Russel & Smith, 2009).

A similar fate awaited the bastion between West Quay, the site of the Castle, and large parts of the eastern wall located north of Briton Street. These were demolished between 1898 and 1899. During this time some parts of the defences were altered, including the addition of steps to the western walls, to provide better access to the town. The iconic Arcades were also blocked to stop rough sleepers (Russel & Smith, 2009). At the beginning of 20th century there were plans to demolish walls south of Westgate, exposed after clearance of Marrett house. In 1905 the Corporation insisted on the demolition of the Bargate, as it was obstructing tram traffic. Only public protests stopped the demolition. As the compromise, 30 years later, the sections of the wall adjoining the Bargate on both sides were pulled down instead (HER, 2016).

2.1.8                    The Second World War

Surprisingly Southampton’s Town Walls were not significantly affected by the German attacks. The wall was used once again, housing a machine gun on Catchcold Tower and search lights (Creighton & Higham, 2005).

2.1.9                    Modern times

The need for legal protection of the town emerged in the 1950s. Most of Southampton’s Walls were listed in 1953, 1954 and 1981, but the preservation was and still is limited to structures above ground level (Russel & Smith, 2009). Supposed protection did not stop further alteration of the monuments and in 1961 the Council agreed to demolition of the York Gate. The gate was a latter addition to the wall and was built in around 1769 (HER, 2016). The new opening in the wall provided access to Copper’s Brewery Site (Tighe, 1970) which went into decline shortly after the demolition of the gate.

2.2                 Surviving sections of the wall

Both sides of the Bargate were cleared of adjoining buildings in the 1930s (HER, 2016). The northern line of the wall located east of the Bargate is largely intact, excluding the opening made after the demolition of York Gate. The Line of defences west of the Bargate was demolished up to the service road between Bargate Street and Castle Way. This part of the wall included a tower (Smith, 1993a) which was removed during the clearance of the prison building in the late 19th century. The walls continue from the service road to the Arundel Gate, with a gap in the wall where the modern street, Castle Way was build.

The west side of the wall was parallel with the shore line of the River Test and is almost complete. It includes Catchcold Tower, so called “forty steps” added in the last century to make pedestrian access to the city, Castle Watergate, Castle Vault, the remains of Biddlegate, the Arcades and the Westgate.

Much of the walls located on the south, were demolished in 19th century. Only a few stretches of it survived in the area of Cockoo lane and Bugle street. The Watergate was demolished but the remains of the tower flanking the gate to the west are still present.

Some remnants of the southern wall can be seen beneath Solent House and in the Platform Public House.

The eastern side of the defences survived between Back of the Wall street and a Lower Canal Walk. It can be seen up to Britton Street.  The surviving monuments consist of God’s House Tower, the wall, the Round Tower, Reredorter Tower and the Friary Gate.

2.3                 Significance of the town walls

The defensive function of Southampton Town Wall may well be long gone but the value of this monument to the city is still invaluable.

The walls are of significant archaeological, historical, cultural and educational importance. Despite being extensively demolished in the last few centuries, over 50% of the monument is still standing. From simple earthworks to sophisticated towers with guns, the town defences exhibit the history and technological progress of medieval military science.  The walls also highlight the socio-economical relationships between people and the town of Southampton. Either way, the wall was the symbol of the civic prestige.

It provides the story and the heritage to the city, which seemed to have lost much of its antique character during the years of modern industrialisation.  The town defences became very important to Southampton’s public image, to the point where it was incorporated in to a logo of Southampton City Council.

The Medieval Town is a major tourist attraction generating wealth for the city. According to the national tourism agency (Visit Britain, 2017), in 2015 Southampton was the 16th most visited city in the United Kingdom, with visitor numbers reaching 257 thousand.

2.4                 Vulnerability of the Southampton defences, and lessons to be learn

As mentioned above, the significance of the walls plunged after 17th century, when the town defences became obsolete. One of the biggest and deliberate blows to the town walls took place in 1775, with the demolition of the Eastgate. During the 19th century, the Harbour Act perceived old monuments enclosing Southampton as an obstacle for development. As a result, most of the Town quay defences were demolished. The 20th century brought some awareness among the citizens regarding the importance of the heritage buildings. Public awareness saved the Bargate from destruction but did not stop the demolition of buildings adjoining the gate. Eventually the local government bodies acknowledged the importance of the town walls, listing some of the major defences in the city. Nevertheless, the legal protection excluded the monuments buried in the ground and failed to save structures such as York Gate.

The threat to the city’s heritage buildings is no less significant than before. The monuments suffer from inadequate restoration commenced in 20th century, involving the repointing of town walls using cementitious mortar. This caused the deterioration of the limestone blocks and retained moisture within the wall. The monument is also affected by atmospheric straining, vibration caused by traffic, gypsum crust formation, vegetation growth and open joists in masonry. New developments within the medieval town wall pose new threats to the buried sections of the wall. New buildings may shadow the existing town defences and restrict the public access.

The problem with ancient monuments in Southampton has being recognised by both Historic England and the Local government. Despite that there is no defined, and approved conservation plan for the Town defences (Russel &Smith, 2009). The comprehensive plan of restoration, preservation and public presentation will require time and government funding. The current financial situation of Southampton City Council suggest that the problem of neglected walls may continue.

Nevertheless, the future assessment and renovation of Southampton’s Town Wall Defences will certainly involve work by an archaeologist. The recording could now be performed using low budget photogrammetry, explored in this thesis.

2.5                 Available publications

One of the most comprehensive publications regarding the history and archaeology of medieval Southampton was published by Platt and Colleman Smith (1975 a; b). Their two-volume work is based on archaeological excavations conducted in the medieval town between 1953 and 1969. The history of the town wall is well presented in older publications by Davis (1883), Hearnshar (1910) and O’Neil (1951). Some description of the town wall can be found in “A walk through Southampton” by Englefield (1805). The main historical source for the late medieval period is covered in the Terrier of 1454 published by Burgess (1976).

The Bargate was described by Faulkner in Plat and Coleman-Smith (1975a: 58-62). The wall located west of the Bargate, up to service road, was mentioned in archaeological reports by Shuttleworth (1994), and partially investigated by Fedorowicz (Fedorowicz & Russel, 2016). The underground remains of so called Another Tower, located west of Bargate were also investigated by Smith (1993). It is important to stress that this part of the wall has no legal protection. The wall behind the service road, going through Arundel Tower and Catchcold tower is covered by Saunders (1976) in “the defences of Southampton in the late Middle Ages”. More information can be found in Southampton City Council Archaeology and Heritage Management Section (Russel, 1991), (Scott, 1991) and the excavation report of the Arundel and Catchold Tower excavations (Smith, 1993).

The excavations of Southampton Castle are covered by John Oxley (1986) and a series of archaeological reports published by Southampton Archaeology Unit (Leivers, 2003), (Mead & Russel, 2000), (Russel & Smith, 2001; 2003), (Russel & Dall, 2000). There are also publications regarding archaeological investigations at Castle Vault (Smith & Russel, 1993) and Castle Hall (Smith, 2007b). The published sources regarding the west wall bastion, Westgate, the Arcades and the west wall between King John ‘s Palace and Westgate can be found in Davies (1883), Engfield (1805) and the Victoria History of the Counties of England (VCH, 1908). More recent publication regarding Westgate can be found in Southampton Archaeological Research Committee Annual Report 1978-1979 (Blackman, 1979). The western wall and its defences were also assessed in 2003 by Russel and Smith (2003) in a Desktop Evaluation of the Archaeological Potential at West Quay.

The High Arcades as called by Englefield (1805) and the south – west wall by Cockoo Lane were investigated during the archaeological watching brief at site SOU 1464. The results were published by Southampton City Council Archaeology Unit (Elliott et al., 2009). The rest of the southern wall, leading towards High Street was investigated during the watching brief on the gas pipe trench at Town Quay (Bareham, 1993).

One of the most iconic, but unfortunately now damaged, structure of the town walls was the Watergate. This monument, similarly to the rest of the defences is partially covered by 19th century publications (Davies, 1883), Englefield (1805), and also by Faulkner in Plat and Coleman Smith (1975: 58-62). There are a number of archaeological reports concerning the Watergate and south wall up to God’s House Tower. These includes reports on the survey and evaluation excavation at the Sun Public House (Garner, 1995), and the watching brief at 84 High Street (Garner, 1996).

God’s House Tower and parts of the wall stretching up to Britton Street were investigated on several occasions. It produced a handful of publications.  In 1995, Peckham, Vincent and Dall (1995) published a report on excavation and watching brief at God’s House Tower. In 2017, Southampton Archaeology Unit excavated 18^th century Debtor’s Prison adjoining the tower, and found remains of earlier stone built structure inside the building (Russel, 2017)

The “Back of the walls” was investigated at sites SOU 252, 253,267,327, 395,406, 413, and 433 and the final report was written by Smith (1994b). Some investigation of the Friary Gate took place prior the its repairs and alternations. The result of this work was published by Smith (2007a). Other reports were made on occasion of archaeological works at No 39 Lower Canal Walk SOU 1260, and at the site of the Former Trimwise Health Club SOU 1263 by Priestley and Molloy (n.d).

The north-west wall survived in good condition, except York Gate which was demolished in the early 1960s. There are three towers present in this part of the wall, the first two, east of Bargate, are often referred as first and second tower. The third one, located at the north-east corner is called Polymond Tower. This part of the defences was covered by Wacher (1975: 142-149) in Plat and Coleman-Smith (1975a). There is also an interim report concluding the archaeological site at York Buildings SOU 175 by Kavanagh (n.d).