LAKE DOWN, NEAR WILSFORD, WILTSHIRE.

Report on geophysical survey of three pond barrows, March 1996.

INTRODUCTION

Magnetometer and resistivity surveys were undertaken over the locations of three pond barrows on Lake Down, near Wilsford, Wiltshire, in March 1996. The surveys were instigated at the request of Mr Edward Flaxman, a retired civil engineer currently investigating the possibility, originally mooted by Ashbee (Ashbee et al 1989, 7, 134, 138), that pond barrows may conceal the locations of buried shafts such as that discovered beneath the site of pond barrow Wilsford 33A.

In an attempt to test this theory, Mr Flaxman arranged for ground penetrating radar (GPR) traverses to be recorded at several sites, including the site of the Wilsford Shaft and two of the four pond barrows in the Lake Down Group. This work was undertaken by Mr John Trust in September 1995 (Flaxman nd). Pairs of orthogonal radar profiles were recorded in each case with a GSSI SIR2 kit using 500MHz and 100MHz antennae. The 500MHz profile over the Wilsford Shaft appears to have successfully located the sides of the upper part of the weathering cone (see Fig 2). However, the shaft itself, the top of which is buried at a depth of approximately 6m, proved to be beyond the range of the equipment used. As the other sites are all located on the same chalk geology, this suggested that, if a shaft was located beneath any of the barrows under investigation, the technique stood a reasonable chance of locating at least part of the weathering cone, if not the main body of the shaft itself. At Lake Down, however, the transects did not produce the same reflection pattern: instead a shallow interface at the barrow centres was indicated.

The surveys reported upon here were intended to help provide more information regarding the near surface characteristics of these barrows and thereby aid the interpretation of the GPR profiles. They also provided an opportunity to assess how conducive the site conditions are for further investigation using more penetrative geophysical configurations. Barrows WS 77, 77A and 78 were each surveyed using both resistivity and magnetometry although to date only WS 77 and 78 have been included in the GPR coverage.

It is worth noting that Grinsell (1957) refers to the excavation of the pond barrows on Lake Down by the Reverend Edward Duke (1790 - 1850) of Lake House, Wilsford. This account was originally published in the Wiltshire Archaeological Magazine for 1907-8 (WAM xxxv, 585) which records that four pond barrows on Lake Down (see Fig1 ) were excavated (namely WS 76A, 77, 77A and 78). It appears that within each barrow an area roughly in the centre was opened and in three cases "nothing found". In the fourth a circular cist containing burnt bone was discovered within 1.5 feet of the surface. Unfortunately it is not stated clearly to which of the four barrows this discovery refers.

The area surveyed, centred on SU 11 39, lies over Upper Chalk (Institute of Geological Sciences 1950).

METHOD

D a Grid of 30m squares was laid out at the site by Mr Flaxman (see Fig 3) which encompassed three of the barrows (ie WS 77, 77a and 78 - see Fig 1). Each of these grid squares was then surveyed using a Geoscan FM36 fluxgate gradiometer. Measurements were recorded at 0.25m intervals along traverses 0.5m apart and the data was periodically down-loaded to a microcomputer in the field. Greyscale and graphical trace plots of this data appear on Figures 4 and 5.

Each square was then resurveyed with a Geoscan RM15 resistivity meter using the Twin Electrode configuration. Measurements were recorded at 1.0m intervals along traverses spaced 1.0m apart. The resulting data is illustrated in this report in the form of greyscale images on Figure 5. The latter includes a plot of the raw data as well as plots of the data after statistical treatment using both a Wallis contrast enhancing filter and a high-pass Gaussian filter (Scollar et al 1990).

During the earlier GPR survey Mr Flaxman had obtained a rudimentary measure of the surface topography of barrows WS 77 and 78 by stretching a tape horizontally across each barrow. Measurements to the ground surface were then recorded at regular intervals along this tape. This procedure was carried out along the traverses used for the radar profiling which, unfortunately, occupied a different alignment to that used for the magnetometer and resistivity surveys. However, the corresponding transects have been extracted from the relevant data-sets to allow a visual comparison with the surface topography to be made. These are presented in Figure 6.

RESULTS

Magnetometer Survey

It is clear from the plots (Fig 5, 1-3) that the magnetometer survey has successfully detected the three barrows, although the response is complex. The process of interpretation is not made easier by the very subdued magnetic response - the frequency distribution histogram on Figure 5.1 shows that the great majority of the readings recorded lie between 0.7 nT, which is close to the maximum sensitivity of the instrument.

From the surface, all three barrows appear to be similar (see Fig 1) with WS 77A apparently a smaller scale version of barrows 77 and 78. The data, however, suggests that this may indeed be a superficial view of the barrows and that significant differences in their make-up exist.

WS 78

The magnetometer response to WS 78 reveals this to be the most uniformly circular of the three, and a sequence of subtle concentric circular magnetic anomalies, both positive and negative, has been mapped. Comparison of this data with both the topographic and resistivity data (see below and Fig 6) suggests that the top of the barrow bank has been detected as a positive magnetic anomaly. This appears to be flanked inside and out by a narrow negative anomaly. There is also the suggestion of further circular positive anomalies both to the inside and outside of these latter. The detection of a bank as a positive anomaly is unusual although a possible explanation is that the bank has a magnetic (turf) core. Alternatively, magnetically enhanced topsoil may, for some reason, be concentrated on the top of the bank, as is the case in the detection of ridge and furrow (see for example Cole 1992).

Within WS 78 a discrete positive anomaly has been detected, just to the east of centre and between two dipole responses, which may represent a pit. Further weight to this interpretation is provided by the resistance survey which has detected a low resistance anomaly at the same location (see below).

WS 77

The sequence of anomalies detected over WS 77 appears to be the same as that over 78. The innermost positive anomaly, however, is not only more prominent than at WS 78 but is shaped more in the form of a horseshoe than a circle, with a possible opening to the north. The presence of such a feature is not evident on the surface. It may be significant that the positive response over the bank is distinctly elevated at two discrete positions in this same area.

The only obvious anomaly located within this barrow is a dipole response, just to the east of its centre, due to a ferrous object. Whilst this may simply be extraneous surface litter, it is possible that it is associated with Duke's excavations (see above).

WS 77A

The response to this barrow is perhaps the least complete of the three. An outermost negative anomaly is only partially visible as an arc running around the southern half of the barrow. The most clearly detected anomaly is the sub-circular positive over the centre of the barrow(cf WS 77). The circuit of this anomaly is incomplete and varies in magnetic intensity with apparent interruptions to both the east and the west. Comparison with the results of the resistivity survey reveals that this feature surrounds a discrete zone of high resistance (see below).

In the area between the barrows, the magnetometer survey has located a linear anomaly running approximately north-south for about 20m just to the east of WS 77. The course of the former field boundary visible on the 1975 Ordnance Survey 1:2500 map (see Fig 2) has not been detected.

On the southern edge of grid square 4 the magnetometer has recorded a characteristically strong response due to a modern ferrous pipe. A deep plough furrow at the edge of the arable field to the west has also been detected as a negative anomaly running north-south through grid squares 1 and 4.

Resistivity Survey

Unfortunately, due to the limited amount of time available in the field, it was only possible to conduct this survey at a lower resolution (1.0m x 1.0m) than was planned (0.5m x 0.5m). An informative plan of the barrows has nevertheless been obtained.

As with the magnetometer survey, the response to the barrows is broadly similar in each case. The barrow banks have been identified as a broad rings of high resistance surrounding an area of predominantly lower resistance. Interestingly, within the latter, all of the barrows show a central zone of higher readings. This appears to correspond well with the GPR profiles which, after topographical correction, indicate a shallow interface at the centre of each barrow (see the profile over WS 78 on figure 2). Given the conditions at the site, the uppermost interface in the GPR profile is likely to be that corresponding with the underlying solid chalk (John Trust pers comm). Taken together, therefore, these results suggest that the surface of the solid chalk is closest to the ground surface at the centre of the barrow. Furthermore, the GPR profile suggests that to either side of centre the depth to the chalk interface increases. This may indicate an accumulation of topsoil against the inside barrow bank, or conceivably an inner ditch. This would in turn offer an explanation both for the reduced resistance and also the positive magnetometer readings in this area (see above). This central area of high resistance is particularly pronounced within WS 77A - unfortunately no GPR profiles have, as yet, been recorded over this barrow.

In addition, the resistivity survey has located other anomalies worthy of comment. Within WS 78, a distinct low resistance anomaly has been detected, just to the east of its centre, which correlates well with an anomaly detected by the magnetometer survey (see above) and which is likely, therefore, to represent a pit. Just to the north of this is a further, though less obvious, low resistance anomaly which corresponds directly with a dipole response in the magnetometer survey. Again, this is perhaps indicative of a former excavation. The unusual horseshoe-shape of the magnetic anomaly encountered within WS 78 is clearly replicated as a low resistivity anomaly.

CONCLUSION

The site conditions proved conducive to geophysical survey and both resistivity and magnetometry have succeeded in revealing more information relating to the structure of the three barrows than is visible from the surface. Significantly, the results indicate that despite their outwardly similar topographical appearance, differences in their subsurface character exist. The magnetometer and resistivity data show good correlation, especially for the barrow interiors. Here comparison of the resistivity data and the GPR profiles suggests that within each barrow there is a central core area where the underlying chalk is apparently closest to the surface (ie resulting in a high resistance and a shallow interface). Surrounding this area, the depth of overburden then apparently increases towards the barrow bank. On this evidence, therefore, it would seem unlikely that a shaft exists beneath any of the barrows. Some disturbance in the magnetic data was detected in the interiors of barrows WS 77 and 78 which may be associated with 19th Century excavations.

The survey results are positive and suggest that any further work should include a more deeply penetrating resistivity configuration. This would profitably be combined with a detailed micro-topography survey. The GPR profiling could also usefully be broadened by carrying out a series of parallel, closely spaced traverses. Microgravity measurement may also offer some potential for shaft location although the contrast in density between the rubble-fill of a shaft and the surrounding chalk may not be sufficient to create a detectable gravity anomaly.

References

Ashbee, P et al 1989 Wilsford Shaft: excavations 1960-62, English Heritage Archaeological Report No 11, Historic Buildings and Monuments Commission for England.

Atkinson, R J C, Brailsford, J W, and Wakefield, H G, 1951 A pond barrow at Winterbourne Steepleton, Dorset, Archaeological Journal, 108 , 1-24.

Cole, M A 1992 Nuneham Courtenay, Oxon. Report on geophysical survey, April 1992, Ancient Monuments Laboratory Report Series, 57/92.

Flaxman, E W nd A radar survey of pond barrows, unpublished report.

Grinsell, L V 1941 The Bronze Age round barrows of Wessex, Proceedings of the Prehistoric Society, 7 .

Grinsell, L V 1957 Archaeological Gazetteer, Victoria County History of Wiltshire, Vol 1, Oxford.

Institute of Geological Sciences 1950 1" map Geological survey of Great Britain, Sheet 298, Salisbury - Drift.

Scollar, I et al 1990 Topics in Remote Sensing 2: Archaeological Prospecting and Remote Sensing, Cambridge.

List of figures

Figure 1 (This photograph not available on Web version of report) The Lake Down Barrow group from the air looking south-west

Figure 2 GPR profiles over the Wilsford Shaft (WS 33A) and WS 78

Figure 3 Location plan of geophysical survey (1:2500)

Figure 4 Greyscale of magnetometer survey overlain on location plan (1:2500)

Figure 5 Greyscales of magnetometer and resistivity data (1:750)

Figure 6 Profiles of topographic, magnetic and resistivity data along GPR traverses

Return To:
On-Line Reports Index Database Home AML Home