With the political framework in place, Lincolnshire can be examined in detail. Was the destruction as total as Saklatvala, Gildas and the Anglo Saxon Chronicles inform us? In order to investigate the defences in the Lincolnshire area, five settlements shall be examined, Lincoln, Caistor, Horncastle, Ancaster and Great Casterton. The locations of these towns are important, as the latter four towns provide a ring of protection around Lincoln. All of the towns are unique in Lincolnshire for the fact that they were walled. In order to better investigate the relationship between them, they can be divided into three groups, the first is the coastal defences of Caistor and Horncastle, the second the inland defences of Great Casterton and Ancaster, and finally Lincoln. Before this, however, the 'Litus Saxonicum', the 'Saxon Shore'19 must be considered, as the main national form of defence against Germanic incursions. The forts and signal stations forming this barrier were constructed throughout the late 3rd and early 4th Centuries to guard the south and east coasts of Britain from seaborne raiders. 
This shows the Lincolnshire coast to be devoid of defences, whilst the south east appears very well defended. Although London, as the capital, required protection, Lincoln would have also been worthy of defence, and susceptible to attack. However, the coastline of Lincolnshire consisted mainly of salt marsh, which would have made fortifications difficult. There is possible evidence of a fort at Skegness which has now been lost to the sea. The Tudor 'historian' Leland reported in 1540 that 'Mr Paynelle sayid onto me that he could prove that there was ons an haven and a towne wallid having also a castelle.'20 Whilst this is not absolute proof, this account shows that there may have been a fort or signal station on the Lincolnshire coast. It is at this point that the walled towns of Caistor and Horncastle become relevant. The main rivers of Lincolnshire, namely the Witham, Bain, Slea and Old Ancholme, allowed assailants access to land a great distance from the sea. For this reason alone, Lincolnshire required some form of coastal defence similar to that of the East Anglian forts. The past twenty years have seen much debate on the subject, but it can now surely been accepted as fact that Caistor and Horncastle functioned as coastal defences. The two sites are joined by a prehistoric trackway known as the 'High Street'.21 This occupied a natural ridgeway and would have been ideal for both communication and observation. It has been noted, though, that the lack of Roman roads around the area would make troop movements difficult if a landing were to take place. To counter this, it seems likely that the garrisons stationed at the towns were cavalry, and would be able to move quickly and effectively without metalled surfaces. This is made more likely by the archaeological evidence of cavalry occupation at Burgh castle, one of the east Anglian shore forts. The defensive qualities of both sites are enhanced by their location on major waterways, the strategic importance of which has been mentioned previously. The lack of excavation is a problem commonly encountered by anyone carrying out studies of rural Lincolnshire. The small population for the land area means that developments are few, especially those on sites of archaeological interest. This is especially true in the modern day, when excavations are happening less frequently, and rarely for pure research purposes. This is true for all of the sites discussed in this study, be they towns, villas or cemeteries. Fortunately, enough excavation has been carried out over the last century to enable the main points to be established, it is merely the smaller details that are now open to interpretation. The Roman town of Caistor consisted of a walled enclosure covering around 3.5 hectares, and an extramural settlement. The walled enclosure by no means replaced the earlier unwalled settlement, as the find of a 4th century lead casket proves. It was inscribed with the phrase 'Cunobarrus fecit vivas', translated to mean 'Cunobarrus made this, may you live (happily)'.22 The existence of such a well made and intricately decorated casket indicates that at least one wealthy person lived in Caistor in the 4th century. The walls were constructed no earlier than the late 3rd Century and were an irregular ovoid shape, due to the ground contours that they followed. At some point, the bastions were added to the circuit, the plan showing those that have been discovered. Bastions are a feature common to all of the walled settlements to be discussed, and generally occurred as a later strengthening. Archaeological evidence has proved inconclusive about Caistor's bastions, and although Burnham and Wacher claim they were built in the mid 4th Century23, a table compiled by Field and Hurst in the SLHA24 journal tentatively states that the bastions are contemporary with the wall. The similarities between Caistor and Horncastle are numerous, both in layout and function. As there is greater evidence that Horncastle's bastions are of the same build as the wall, it would seem likely that the same was true for Caistor. The second element of a defensive fortification is a ditch. Caistor's ditch eluded archaeologists for many years, due to a lack of suitable opportunities to excavate. However, an excavation in 1992 at Caistor Grammar School25 identified part of a ditch running parallel to the Roman wall. From the angle of the sides, it was estimated to have been at least 8 metres wide, and therefore a sizeable fortification. Unfortunately there was not enough dating evidence to ascertain whether the ditch was early or late Roman. The fortifications at Caistor were therefore substantial, but still open to interpretation. The presence of a large defensive ditch and bastions placed at intervals around the entire walled circuit suggests that the site expected to come under attack from all sides. Modern military history tells us that the Germanic tribes, whom presumably these defences were built to repel, had no experience of siege warfare and were therefore not equipped to assault a fortification such as Caistor. With this in mind, it seems that the Britons were either overcautious in their approach to the defences, or they expected the attacks to come from an enemy more adept at siege warfare. However, the location of the site strongly suggests coastal defence, and therefore the only plausible assailants would be Germanic. Horncastle is similar to Caistor in a number of ways. It also comprised an unwalled settlement and a walled enclosure, measuring c.54 hectares and c.2 hectares respectively.26 The unwalled enclosure has been shown to be of Iron Age origin, and occupation continued through to the late 4th Century. As with Caistor, it is the walled enclosure which is of greatest interest to this study. It was situated on a naturally defensive position, at the junction of two rivers, the Bain and the Waring. This position appears to have made a defensive ditch both impossible and unnecessary. As has been mentioned in the Caistor discussion, Horncastle's bastions appear to be of the same build as the curtain wall, which is of the late 3rd Century.27 The excavations carried out by Field and Hurst conclude by identifying similarities between Horncastle and the recognised shore forts of Burgh Castle and Portchester, in that the walls are massive, but the interior of the defences has yielded little evidence of occupation, a feature commonly associated with cavalry installations.28 a further intriguing observation on the Horncastle bastions was made by Hawkes and Richmond, who suggested that the circular bastions were suitable for the mounting of artillery.29 This interesting proposition has a bearing on the town of Great Casterton, and the implications of such devices at both sites will be discussed in due course. The photographs below show some of Horncastle's remaining wall and a bastion at Burgh castle, illustrating how the Horncastle bastions may have appeared. 
Great Casterton is located on the Lincolnshire border, just off the line of Ermine Street south of Lincoln and consisted of a fortified town and nearby villa. The villa is of importance to this study and will be discussed later. The defensive wall at Casterton, unlike those of Horncastle and Caistor, was designed to enclose the greater part of the settlement.30 The enclosed area was estimated to be between thirteen and eighteen acres.31 This suggests that the purpose of the wall was to defend the actual settlement rather than as a garrisoned fort located to protect Lincoln, although defence of the Provincial Capital must still be assumed as a function of the site. The earliest defences of the site date to the 2nd century, consisting of a wall, rampart and at least two steep sided rock cut ditches.32 In the mid 4th Century these defences were reorganised. The multiple ditches were replaced with a single ditch around 60' wide, dug further away from the wall. This served two purposes. Firstly it served as a larger obstacle for attackers, and secondly, it allowed bastions to be added to the walls. The diagram below shows both phases of the defences. 
The bastions at Casterton demand special attention. It appears that the stone used to construct them was freshly quarried for the purpose, a task that would require thousands of tons of rock and a great amount of labour. This type of endeavour would have been deliberately planned, as opposed to defences rapidly thrown up in the face of an attack.33 The bastions also housed artillery defences, as has been mentioned in the Horncastle discussion. Corder estimated, using Richmond's work 'The City Walls of Imperial Rome', that two ballistae, large, torsion operated bolt firing machines, would be housed in each bastion. This would ensure that there was no 'dead ground' in which an attacking enemy would be able to gain respite from the weapons.34 Each weapon would have required a crew of four, therefore the question is raised as to whether a military garrison was posted in the town, if a civilian defence force existed instead. Whatever the truth, such defences would have been formidable, and, as mentioned in the discussions of Caistor and Horncastle, they lead to the suggestion that, whereas the earlier walls may have been to control entry and exit, these new fortifications were designed to withstand siege warfare. The date for the construction of the bastions was given earlier as the mid 4th Century. This dating originated from coins discovered in the bases of the bastions. Corder uses these to date the work to between AD337 and AD367.35 Frere agrees, and comments that the bastions 'can hardly have been put up before 360 (and) we can hardly doubt that this reorganisation … was the work of Count Theododius'.36Ancaster occupies a position geographically close to Great Casterton, yet strategically similar to the coastal defences, making it an awkward settlement to categorise. The town is situated on ground of great strategic importance, overlooking the Ancaster gap. The town is also sited on the river Slea, and in this respect may serve a function similar to that of the coastal sites, despite being much further inland. The defences are similar to that of Casterton in that a 2nd Century wall and rampart were later strengthened. At Ancaster this strengthening involved the erecting of bastions and digging of ditches. The rampart has been dated through the discovery of five coins sealed at a low level, and also pottery sherds. These finds indicate a date between AD250 and 28037, contemporary with the initial phase of defences at Great Casterton, and the 'one off' defences of Caistor and Horncastle. The reinforcement of these defences appears to be of a similar date to Great Casterton's. There is little evidence to accurately date the two ditches that surround the enclosure, but the evidence that does exist suggests a 4th Century date. The inner ditch was up to 16.8 metres wide, and the outer ditch up to 9 metres wide. Such a formidable barrier was further enhanced by the addition of the four corner bastions, which have also been tentatively dated to the 4th century.38 The shape of these bastions is unusual, as the diagram section shows. 
It can be seen from the diagram that the towers differ considerably from those of Casterton, Horncastle and Caistor, portrayed in the Burgh Castle photograph. The Ancaster towers follow the line of the defences rather than coming out from them. This may be simply due to a different architect supervising works, but the geographical proximity of Casterton, combined with the similar build dates suggest either a strategic or topographical reason for the difference. The design may mean that the towers were incapable of mounting ballistae, suggesting one of two possibilities:
1. The town at Ancaster was not expected to withstand sieges of the intensity that Horncastle and Great Casterton seem to have been prepared for, or
2. The topography of the site made projecting bastions impossible, but it was felt that any strengthening of the defence was preferable to none. The final settlement to be discussed here is the one which the previous walled towns seem to be designed to protect. Lincoln itself was walled for defence, and as at Great Casterton and Ancaster, these defences developed through time, becoming stronger as a greater need for protection was felt. The stages in Lincoln's defences were numerous, and spanned the whole period of Roman occupation. It is needless to say that the defences were more substantial than those of the smaller towns, and enclosed a larger area. There are various surviving remnants of Roman Lincoln, including some of the city wall. One section of 4th century strengthening is shown below, and can be seen to the west of Newport arch. 
The 4th Century strengthening was extensive, but in some places showed signs of hasty workmanship, though as some sections appear to have been dismantled and then rebuilt, the amount of work carried out must have been immense.39 Despite much excavation being carried out on the defences of Lincoln, the focus of this study is on the settlements other than the provincial capital, and a brief summary such as this will have to suffice.
|
|
