Appendix 10.D Metal Detecting Report Entec UK Limited
Entec UK Limited
KELMARSH WINDFARM SITE: NASEBY BATTLEFIELD Report on a survey undertaken by the Battlefields Trust on behalf of E.ON during October-December 2008 Summary Metal detecting survey was undertaken to establish whether the western part of the proposed windfarm site at Kelmarsh does or does not encompass part of the battlefield of Naseby. The survey has shown a total absence of early modern lead bullets indicating that no significant combat action took place in this area as part of the battle of Naseby. All other metal finds of any date were very few in number. Background The work reported here was undertaken by the Battlefields Trust on behalf of E.ON on land owned by the Kelmarsh Estate which is being considered for the development of a windfarm. The survey was developed in consultation with Daniel Badcock of E.ON and Rob Johns of ENTEC. Liaison regarding access for survey was with Jeremy Watson, Land Agent to the Kelmarsh Estate. The survey was carried out to establish whether the western part of the proposed windfarm site at Kelmarsh does or does not encompass part of the battlefield of Naseby. The interpretation presented in Foard 1995 would indicate that no action took place in this area and this interpretation has been followed by the English Heritage Battlefield Register. However, as Marix Evans of the Naseby Battlefield Trust has indicated in discussion with EON, the interpretation of the deployments and action at Naseby provided by Rogers could be cited as indicating that the battlefield extends into the western edge of the proposed windfarm site. 1 The sampling for battle archaeology has been achieved using systematic metal detecting survey of 26 hectares in the western part of the windfarm site. As work at Edgehill has demonstrated, cavalry action of the Civil War may be expected to leave a slight but distinctive signature in the form primarily of carbine and pistol calibre bullets. Its presence or absence can be considered definitive evidence as to whether significant combat took place in the area. 2 There is no record of previous metal detecting survey in this area, although unreported recreational detecting may have taken place, but this is unlikely to have significantly affected the distribution of bullets. Implementation Detecting was supervised by Dr Glenn Foard, Project Officer of the Battlefields Trust. Four other detectorist have been employed, all with substantial detecting experience, including work during 2005-8 on the Bosworth Battlefield Survey. Finds identification and analysis was undertaken by Dr Steve Ashby, Dept of Archaeology, University of York. 1 Foard, 1995; Rogers, 1968 2 Foard, 2008b, ch 5
detectorist detector Glenn Foard Minelab Explorer II Barry Wright Tesoro Cibola Malcolm Green Fisher 1267 Lee Walkley Laser Hawkeye Peter Hartley Tesoro Cibola Figure 1: Survey area with survey field numbers indicated (red), shown in relation to the full area being considered for the windfarm (blue). 1km national grid. Six fields, centred on NGR SP708789, were surveyed on 28/10/2008 (fields 1, 2 and part of 3) and 17/11/2008 (fields 4, 5, 6 and part of 3). All fields were in ideal condition for detecting and signals were good with reasonable depth. Fields 1-3 had very low stubble (which did not interfere with detecting sweep) and has been direct drilled with rape which was very sparse and less than 5cm high and the soil was damp. Fields 4-6 had been ploughed, disked and drilled with a sparse cereal crop less than 5cm high and the soil was wet. Weather conditions were generally good but with difficulty caused by heavy snowfall during the last stage of work on field 3 on 28/10/2008 which led to slightly faster detecting speed and to one transect accidentally diverging from the flagged alignment. The survey employed an intensity of 10m spaced transects which is sufficient to locate evidence of cavalry action. This is based on the results of the Edgehill survey. 3 While infantry action can yield a density of up to 8 bullets per hectare at 10m transects, cavalry action yielded a maximum of 1.5 bullets per ha in the core of the clash and elsewhere the density was of the order of 0.5 bullets per ha. Thus surveying an area of 26ha at 10m transects some 13 bullets might be expected. This is an 3 Foard, 2008b
adequate sample to distinguishing of combat evidence from the background noise which can exist from sporting activity between the 16 th and 19 th centuries. Ferrous artefacts were discriminated out and those any incidentally recovered were not retained as none were of significance. All 18th century and later finds were discarded without record except where of intrinsic interest. The metal finds were processed to clean, add a label, bag and input into a database. All finds that were certainly post battle, unless of particular significance in their own right, were discarded either at this stage or by the finds specialist. The retained finds are stored in a polyethylene box with silica gel to maintain low humidity and with a humidity indicator strip to enable monitoring. All finds have been subject to spot dating. No finds likely to be of 17 th century date were recovered and so no detailed reporting has been prepared on the artefacts. Survey coverage The map of survey coverage, shown in Figure 2, presents every track point recorded by the GPS units, at 15 second intervals, during the survey. Where concentrations occur these represent time spent at the ends of transects or short breaks in detecting. Small concentrations of points along a transect normally represent time spent digging. The tracks deviate from the central 10m line largely due to drift in the GPS signal which almost always gives an accuracy of better than 10m and normally better than 5m. Small gaps in survey coverage at field margins generally relate to inaccessible land under high grass or scrub or waterlogged areas. A more detailed analysis of survey intensity for each detectorist relative to finds recovery has not been prepared as no significant artefact distribution was recovered during the survey. Figure 2: Survey coverage for the six numbered fields: GPS trackpoints at 15 second intervals Detecting speed graphs are provided to give a crude indication of the average reconnaissance speed while scanning, which here ranges between circa 12-18 metres
per minute. The higher speeds on the graph represent time spent setting up grids etc while lower speeds represent where time was spent digging. NASEBY EON SURVEY: DETECTING SPEED 28/10/2008 Number of records 320 280 240 200 160 120 80 40 0 BWright GFoard LWalkley MGreen PHartley 0 6 12 18 24 30 metres per minute Number of records 280 240 200 160 120 80 40 NASEBY EON SURVEY: DETECTING SPEED 17/11/2008 BWright GFoard LWalkley MGreen PHartley 0 0 6 12 18 24 30 metres per minute Survey results Figure 7 illustrates the distribution of artefacts identified in the survey area. There was a total absence of early modern lead bullets, indicating that no significant combat action took place in the survey area as part of the battle of Naseby. Other metal finds of any date were very few in number. The recovery of a Roman brooch fragment (find 03) and a possible heavily corroded Roman coin (find 02) may suggest there is a Roman site nearby, although this is unlikely to be in the survey area. The other, later finds were of limited intrinsic interest. Indeed there were a remarkably low number of non ferrous signals of any type, the largest number being from modern.303 bullets and cartridge cases especially in field 6, though none of these were recorded or retained. The negative survey results are not due to limitations in the survey as conditions were in all respects good. Casual observation of the three previously ploughed fields (4-6) failed to reveal any significant scatter of ceramics of any period.
One non-metal artefact scatter was noted at NGR SP70597876 (field 6). This was a low density concentration of worked flint in a small area of gravelly soil on a spur. As this was noted through casual observation while detecting it cannot be taken as a guide to the true density of any flint scatter and no conclusions of significance can therefore be drawn. Systematic fieldwalking could be used to determine the extent and significance of this artefact scatter. Figure 3: Iron wedge (find 11) Figure 4: Fragment of a Roman brooch (find 03)
Figure 5: Obverse face of the terret (find 09) Figure 6: Lead weight (find 12)
Figure 7: Artefact distribution giving find numbers which correlate with finds catalogue. G Foard The Battlefields Trust 30 th January 2009
Appendix: Methodology for survey of unstratified battle archaeology The survey methodology employed was largely that applied in the Edgehill battlefield study (2004-7), which was the first battlefield-wide archaeological survey of a major battlefield in England. 4 The most extensive and important archaeological evidence of a battle is the scatter of unstratified artefacts left in the topsoil by the action itself. A wide range of artefact types may survive in small quantities. The major category, both in volume and significance, is the projectiles with tens of thousands of bullets deposited by the firearms. Thus it is the bullets that provide an extensive recoverable pattern related to the intensity of the fire-fights in particular areas and phases of the battle. It is unlikely that significant archaeological patterning could be derived from evidence of the long distance artillery bombardment with iron roundshot that opened many battles of the period. The rate of fire of the larger field artillery was so slow that roundshot were deposited in very small numbers and their recovery by detecting is likely to be very limited. Weapon related items, sword parts and armour, appear to survive on the field in such small quantities that they may rarely contribute significantly to the understanding of the broad pattern of the action, which is what is required for the present project. The one exception is the lead caps to the powder boxes on the bandoliers carried by some musketeers. Though present only in small numbers they are distinctive military items and, unlike most bullets, they may indicate where the musketeer actually stood or fell or where his body was pillaged or stripped for burial. The problem with non-weapon related artefacts is further complicated because most items brought onto the battlefield will have been of types that were not distinctive to the military. As a result individual unstratified items taken in isolation cannot be directly related to the battle. This can only be deduced from the quantity of such artefacts compared to the expected background noise left by non military activities. For the metal detecting survey transects are laid out 10m intervals aligned parallel to the most convenient hedge, each marked by a minimum of four flagged what to ensure that the detectorist can always accurately maintain their position on the transect. The detectors are used in discrimination mode to exclude ferrous signals because there are vast numbers of ferrous artefacts in the topsoil on most English battlefields, rendering efficient all metal detecting impracticable for large scale survey. The detectorist retrieves the artefacts that they locate. All recording is by hand held navigation grade GPS with the Ordnance Survey algorithm used for conversion from WGS84 to the Ordnance Survey datum ensuring better than 5m accuracy, which is adequate for battlefield-wide survey. 5 The intensity of survey varies by the speed of forward motion during detecting. Each member of the team is encouraged to work at a similar pace to minimise bias, but this will also vary according to ground conditions. Each detectorist therefore carries a GPS unit attached to the stem of his detector, which is continually on in tracking mode logging at 15 second intervals the location of the detector throughout the day. This enables calculation of exact amounts of time spent in survey work and exactly 4 Foard, 2008b 5 The methodology is fully detailed in Foard, 2008b
where and distinguishing the time spent prospecting from that spent in digging or other activities. This enables the consistency of the survey to be fully monitored enabling a higher degree of confidence in the varying artefact densities across the site. Obviously modern non ferrous items are noted to enable a count of the amount of background noise being identified by each detectorist in each field. All potentially significant finds are separately bagged when found and the GPS used to record a waypoint. The finds bag is then annotated with the detectorist s initials and the waypoint number so that it can later be correlated with the GPS data downloaded into the GIS. When the work on that field for that day is completed a note is made on the day record sheet as to the amount of non ferrous rubbish that has been recovered by each detectorist. Brief notes are also made as to the land use, soil conditions of the field, weather and other information relevant for the understanding of the effectiveness of the detecting on that field on that day. The metal detector used by each detectorist is also recorded. The GPS data is downloaded into MapInfo. A polygon is recorded for each area surveyed in each field on a single day and the conditions and related data appended. Waypoint data is correlated with the finds, once washed and re-bagged, a finds number allocated and the data entered onto the GIS and the find number, grid reference etc onto the bag. In undertaking this work the Battlefields Trust guidance on battlefield survey, including risk assessment was applied in the metal detecting survey. Bibliography Foard, G. 1995 Naseby: The Decisive Campaign, Pryor Publications, Whitsable. Foard, G. 2008a 'Conflict in the pre-industrial Landscape of England: A Resource Assessment' University of Leeds / English Heritage. Foard, G. 2008b 'Integrating the physical and documentary evidence for battles and their context: A Case Study from 17th Century England' In PhD in School of HistoryUniversity of East Anglia, Norwich. Rogers, H. C. B. 1968 Battles and Generals of the Civil Wars 1642-1651.