Introduction

Following an archaeological survey conducted by the Field Archaeology Unit, University College London (FAU) between Seaford Head and Beachy Head, E. Sussex, a suspected Bronze Age round barrow on Baily's Hill was identified as being under imminent threat through coastal erosion (Holgate 1986) <="crow01.tif">[view of Baily's Hill from Birling Gap 7.9kB tif]<>. Given the proximity of the barrow to the cliff edge and current rates of erosion it seems likely that the barrow will be destroyed within the next few years. A programme of archaeological evaluation has therefore been proposed by the County Archaeologist to record the monument prior to its eventual destruction through a project design commissioned by English Heritage (Greatorex 1998). The project is to be jointly funded by Sussex County Council, English Heritage and The National Trust, who own the land on which the barrow stands.

The monument (SAM No. East Sussex 342), subsequently referred to as "Barrow A", is believed to represent a bowl barrow dating to the period circa 2500 to 1500 BC, although a number of other similar mounds excavated in Sussex have revealed Saxon burials in primary contexts (Drewett et al 1988). A second topographic feature 75m NE of Barrow A has also been identified as a bowl barrow and is included with in the same SAM designation. The aim of the geophysical survey was to place Barrow A in the context of the surrounding landscape and more specifically, to confirm the presence of an encircling ditch around Barrow A and provide evidence for any additional archaeological activity associated with the monument.

The site (centred on TV 54 96) is located on a shallow layer of well drained calcareous silty soil of the Andover 1 Association (Soil Survey of England and Wales 1983) developed over a substrate of undivided Upper and Middle Chalk (Institute of Geological Sciences 1971). It is included within the Sussex Downs Area of Outstanding Natural Beauty and is designated as a Site of Special Scientific Interest (SSSI) due to the downland flora and associated insects that it supports. The monument survives as an approximately oval raised mound with a diameter 15.5m EW and 13.5m NS standing to a recorded height of 0.5m situated 7.5m from the current cliff edge.

Method

Due to the success of magnetic survey over similar monuments and geology (eg Caburn Bottom, Glynde, E. Sussex; AML archive), this technique was adopted to cover an area of approximately 2.5ha surrounding Barrow A. A 30m grid was established over the site (Figure 1) and two permanent survey markers inserted along the baseline. Unfortunately, due to instrument failure it was not possible to establish the precise NGR coordinates of the survey grid for this report. Therefore Figures 1 and 2 provide only an approximate location and field measurements should be made with reference to the permanent survey markers established at the site.

Data was collected from each 30m grid square using a Geoscan FM36 fluxgate gradiometer along N-S traverses following the standard method outlined in note 2 of Annex 1. In addition, an earth resistance survey was conducted over a more limited area but encompassing both scheduled monuments identified at the site (Figure 1); this technique has ed successful in locating the encircling ditches of a bowl barrow at Bullock Down, Beachy Head, E. Sussex, (Hackmann 1976). The earth resistance survey was conducted at a 1m x 1m sample interval over an area of approximately 0.6ha with a Geoscan RM15 resistivity meter utilising a twin-electrode array and mobile probe spacing of 0.5m (note 1 of Annex 1).

More detailed resistivity survey was conducted over an off-set 30m x 30m square centred on Barrow A (see Plan B for location). This latter survey utilised a Geoscan MPX 15 multiplexer and adjustable PA5 electrode array to simultaneously collect 0.5m and 1.0m mobile probe separation data. The greater separation of the mobile probe electrodes forces the applied current to penetrate further into the ground and can often detect anomalies arising from more deeply buried features (Scollar 1990, 321-4, Linford 1993). In this case, the sample interval was 0.5m (NS) x 1.0m (EW) for the 0.5m mobile electrode array and 1.0m x 1.0m for the deeper penetrating 1.0m mobile probe spacing array.

<="crow02.tif">[Resistivity survey in progress over Barrow A 10.9kB tif]<>

Plan A shows a greytone image and X-Y traceplot of the magnetometer data after statistical processing of each survey line to provide a zero-centred mean. This process eliminates offsets between adjacent survey lines that may occur due to the directional sensitivity of fluxgate gradiometers when data is collected from alternate "zig-zag" traverses and considerably improves the presentation and interpretation of the resulting data. Plan B1 and B2 show a traceplot and greytone image of the raw resistivity respectively. An enhanced greytone image of the data is presented in Plan B3 after processing with a high-pass Gaussian filter (radius = 5m). Results from the detailed resistivity survey over Barrow A are presented in Plans B4 and B5 together with residual near surface anomalies enhanced by subtracting the deeper penetrating data set from the shallow readings (Plan B6). The latter data was collected with the remote electrode pair separated to a distance at which their contribution to the recorded reading became negligible. Under these conditions measurements recorded with the twin-electrode array multiplied by a factor of 2%r (where r = mobile probe separation) express the apparent resistivity of the volume of ground immediately below the mobile electrodes in units of %m.

A series of ~100g topsoil samples were also collected at 15m intervals along the survey base line (Figure 1) to assess the variation in magnetic susceptibility over the site (Figure 3).

Results

Significant anomalies discussed in the following text are numbered and can be identified by reference to Plan C.

Magnetic data

The most obvious magnetic anomaly [1 ] is found over the location of the second monument immediately NE of Barrow A. This takes the form of a circular, ditch-type response of diameter 12m that appears to contain two central pit-type anomalies. Such a response would be expected from a bowl barrow with an in situ encircling ditch and confirms the suitability of magnetic survey for the location of such monuments at the site. Similar responses, ( [2] , [3] and [4] ) were recorded in squares 1-4 to either side of the walkers' path which appears to have created a detectable anomaly [5] along both edges through the erosion of topsoil along its course. Whilst [2] , [3] and [4] can not be identified as the response to barrows the area did contain a number of topographic features with similar dimensions to both Barrow A and the tumulus at [1] .

Additional positive anomalies [6] and [7] visible in this area may also represent significant archaeological activity, perhaps a pair of large pits.

A series of linear anomalies cross the entire survey area and appear to represent two phases of a relict field or enclosure system. The two systems can be distinguished both by their orientation and their relative response with regard to the zero mean of the combined data set. The first pattern [8] forms a single rectilinear enclosure with a largely negative response and can be attributed to a series of extant linear impressions observed in the field. The negative response would appear to arise from the lack of magnetic topsoil (see Topsoil susceptibility below) in the open ditches and is contrary to the positive anomalies normally produced by the magnetically enhanced fill expected from buried ditches.

The second pattern of ditches [9] does exhibit a positive response which appears to run as a "zig-zag" NS across the site. However, it is possible that the magnetometer has only detected part of a more extensive enclosure system suggested by a number of incomplete linear anomalies (eg in grid squares 12 and 28) which, if extended would form a pattern of rectilinear enclosures over the crown of the hill. It seems reasonable to assume that [9] predates [8] on the grounds that it is still partially extant and passes through [1] suggesting that the ditch system was constructed long after the monument ceased to be respected as part of the surrounding landscape and may, perhaps, be quite recent in origin.

It is of interest to note the variation in magnitude of response demonstrated by [9] which is greatest through squares 1 and 3 and may, perhaps, be indicative of increased anthropogenic activity in this area enhancing the magnetic susceptibility of the topsoil in-filling cut features. To the south a single section of parallel ditches is found in squares 7 and 11 and a distinct pit-type response [10] is visible in square 15.

Additional linear anomalies [11] , [12] and [13] may well also represent parts of a relict field system. However, they appear to be on a different alignment again to patterns identified above and are too incomplete to allow a more confident interpretation. Two further areas of amorphous magnetic disturbance [14] and [15] are found in square 12 and may represent archaeological activity although a natural origin cannot be ruled out. It is worth noting the presence of substantial periglacial features recorded along the cliff top at Birling Gap (eg Ballantyne and Harris 1994; pp102) which may give rise to magnetic anomalies.

The data in the vicinity of Barrow A (squares 25 and 26) is partially obscured by the presence of quite intense ferrous noise [16] which is most evident in plan A1 and has hampered the identification of subtle magnetic anomalies in this area. However, the location of the mound itself is not directly effected and fails to produce any readily identifiable magnetic anomaly. An area of enhanced magnetic response [17 ] exists to the SE although this is closely associated with two areas of ferrous disturbance. Whilst this may represent significant activity the magnitude of this response (~9nT) is far greater than the much weaker archaeological anomalies identified throughout the rest of the survey and is therefore suggestive of a more recent origin (see Topsoil susceptibility below). A weak linear anomaly [18] and a pit-type response [19] are found immediately N of [17] but neither anomaly appears related to either Barrow A or the field systems identified above. Further weak linear anomalies [20] to the E in squares 26-28 run on an EW orientation and again fail to respect any other anomalies in the data set defying a confident interpretation.

Resistivity data

Area survey (squares 14, 17, 18, 21, 22, 25 and 26)

It is of interest to note the low background resistance of the site in the vicinity of the cliff edge (Plan B2) contrary to theoretical expectations for the approach to similar high resistance lateral boundaries such as excavation trenches (eg Scollar 1990; pp321-4). In this case, it would appear that the dimensions of the boundary with respect to the electrode spacing and position of the survey have resulted in a negligible contribution to the recorded resistance from the cliff to air interface. The low resistance recorded reflects a genuine increase in the conductivity of the chalk close to the cliff edge, possibly caused by saline effects from seawater blown against the cliff face.

Beyond the background response the data contains few significant anomalies but does, in part, corroborate the results of the magnetic survey. Two high resistance anomalies [21] are found in the vicinity of the extant tumulus at [1] but there is no low resistance response to the encircling ditch as might be expected. To the south a series of ditch-type low resistance anomalies [22] are recorded (

Square 21 contains a direct correlation between an area of intense magnetic disturbance and a high resistance anomaly [23] . Given the location of the site it seems probable that this represents a modern ferro-concrete structure possibly related to war-time activity on the cliff top.

Detailed survey over Barrow A (squares 25 and 26)

Initial analysis of the detailed resistance survey over Barrow A reveals no definite indication of a ditch-type low resistance anomaly encircling the extant mound with the exception of an indistinct response in the deeper penetrating data (Plan B5). Furthermore, comparison of Plans B4 and B5 shows little variation of resistance with depth and suggests that the majority of causative features are close to the ground surface. Subtraction of the deeper penetrating data set from the shallow should emphasise such near-surface variations and in this case (plan B6) reveals an apparent high resistance anomaly [24] within the area encompassed by Barrow A. Whilst this anomaly may tentatively represent an accumulation of rubble forming the basis of the raised mound it could also be provoked by the topography of the mound itself.

Beneath such a mound the current lines will be free to spread out resulting in a diminished density and a subsequent low resistance response for a quadripole array with a electrode spacing less than or equal to the dimensions of the topographic feature (Scollar 1990, pp349; Fox et al 1980). The twin-electrode array deployed will approximate to the response of a quadripole and would produce a positive residual anomaly only if the magnitude of the topographic response is greater in the deeper penetrating data set which in this case seems unlikely.

A circular low resistance anomaly [25] is evident to the SW of the Barrow A and corresponds with the location of the strongly magnetic response [17] . This anomaly produces a negative residual in Plan B6 which again suggests a near surface origin for the underlying causative feature.

Topsoil magnetic susceptibility

Figure 3A shows the variation of topsoil susceptibility along a NS baseline across the site measured with a dual frequency Bartington MS2 meter and 10cc laboratory coil from subsamples of the soil collected in the field. Initial measurements were made on the samples within 24 hours of collection and then repeated after air-drying at room temperature until no further loss in mass was observed. Minor variations in both Chi and ChiFD% were observed with the latter demonstrating the most consistent alteration to a lower value for the majority of samples after air drying. This indicates either a low-temperature oxidation of ferrimagnetic mineral (such as greigite which is highly unlikely in this context) or increased grain-grain interactions between superparamagnetic particles caused by the reduction in sample volume.

Due to the thin layer of soil present at the site the samples contained varying proportions of mineral soil, organic matter and fragments of the underlying chalk. Whilst this admixture will have no effect on the values of ChiFD% the intercomparison of susceptibility along the traverse requires the normalisation of the volume susceptibility to the mass of mineral soil present. This was determined by sieving through a 1mm mesh to remove inclusions of chalk followed by heating at 375oC for 16 hours to remove the majority of organic matter present. Values of Chimin presented in figure 3A have been normalised to mass specific units through the experimentally determined mineral soil mass. The results demonstrate the wide range of susceptibility values encountered over the site with high values recorded in the vicinity of Barrow A. Whilst this peak is not necessarily significant the two samples collected over the monument do exhibit anomalous values of ChiFD% <6% in comparison to values >10% demonstrated by the rest of the data (Figure 3B).

The magnitude of ChiFD% is related to the grain-size population of the magnetic particles present in the sample (Thompson and Oldfield 1986; pp54-56) with values of ChiFD% exceeding 10% equating to the majority of particles present being ferrimagnets in the superparamagnetic range. This suggests that topsoil in the vicinity of Barrow A contains either (i) a high concentration of larger single/multi domain particles or (ii) increased grain-grain interactions between the superparamagnetic particles present. To investigate this further additional mineral magnetic measurements were performed on two samples 60S and 225S both before (wet) and after (dry) air-drying. These measurements included:

Partial Anhysteretic Remanent Magnetisation (pARM) determined by applying a 0.05mT steady field through incremental 5mT windows during AF demagnetisation in a peak 200mT field. ARM for the sample was subsequently measured after applying the steady field throughout the entire AF demagnetisation (Figure 3C and 3D).

Isothermal Remanent Magnetisation (IRM) acquired through exposure to pulse magnetic fields to a maximum of 1000mT (Figure 3E and 3F).

AF demagnetisation of IRM(1000mT) subsequent demagnetisation of IRM(1000mT) with incremental peak AF fields from 1mT - 200mT (Figure 3E and 3F).

Results of the pARM measurements demonstrate the low coercivity of the magnetic minerals present in the samples. Approximately 90% of the ARM is acquired by particles with a coercivity <40mT indicating either the presence of large multi-domain grains or an assembly of very fine interacting superparamagnetic grains. Sample 225S demonstrates a consistently lower pARM response than sample 60S suggesting an apparent shift of the grain-size population towards larger SD particles. The results of the IRM acquisition curves (Figure 3E and 3F) confirm this shift in the coercivity envelope and show that the majority of IRM is acquired in fields <100mT. Demagnetisation of IRM acquired in a 1000mT field (Figure 3E and 3F) indicates a much greater similarity between the two samples (both wet and dry) than was found for the IRM acquisition curves. This again suggests the greater influence of grain-grain interactions in sample 225S as the additional magnetic fields generated between particles will act to oppose the creation of IRM and aid its demagnetisation.

Both samples 60S and 225S appear to have a very similar grain-size distribution which may be attributed to the dominance of a low coercivity mineral such as magnetite or maghaemite. From consideration of the high values of ChiFD% it would appear that the majority of particles are superparamagnetic with considerable grain-grain interactions accounting for the observed remanence behaviour after exposure to laboratory fields. Whilst much larger multidomain particles may also demonstrate a similar low coercivity and offset between acquisition and demagnetisation curves, due to internal demagnetising fields, no frequency dependence would be expected. Sample 225S demonstrates a much greater grain-grain interaction than 60S that may indicate either a variation in the pedogenic enhancement of topsoil towards the cliff edge or an anthropogenic modification, such as the localised action of fire.

Conclusion

Magnetic survey at this site has revealed a number of significant anomalies apparently related to a group of prehistoric burial mounds. Additional linear anomalies suggest the presence of a two-phase ditched enclosure system which appears unrelated to the funerary activity at the site. However, results in the vicinity of the threatened monument, Barrow A, have been hampered by the presence of unexpected magnetic disturbance related, perhaps, to war-time activity. In particular, the presence of a magnetic anomaly related to a palisade or ditch encircling Barrow A has not been detected although a similar anomaly does surround the location of the other recorded tumulus covered by the survey area.

Resistivity survey failed to detect a ditch-type anomaly around either monument but did reveal the presence of a high-resistance anomaly associated with the mound of Barrow A. Thus if Barrow A does indeed represent a funerary monument it would appear to take the form of a raised mound without the presence of any obvious encircling ditch.

Anomalous topsoil susceptibility results in the vicinity of Barrow A suggest the pedogenic enhancement of magnetic minerals has been modified through the action of fire at some point. Whilst this may be due to a natural fire the location of a coincident enhanced magnetic anomaly [17] and a shallow low resistance response [25] adjacent to the monument on a prominent cliff-top suggests the site of a deliberate beacon or fire.

Subsequent excavation of Barrow A should be extended to confirm the nature of the magnetic disturbance surrounding the monument and also investigate the adjacent anomaly to the SE tentatively identified above as a burnt feature of indeterminate age. Wider evaluation of the surrounding area is also recommended with particular regard to the recording of topographic anomalies N of Barrow A.

References

Ballantyne, C. K. and Harris, C., 1994, The Periglaciation of Great Britain, Cambridge University Press.

Drewett, P., Rudling, D. and Gardiner, M., 1988, A regional History of England: The South East to AD 1000, Longman.

Fox, R. C., Hohmann, G. W., Killpack, T. J. and Rijo, L., 1980, Topographic effects in resistivity and induced-polarization surveys, Geophysics, 45, pp94-108.

Greatorex, C., 1998, A project design and costing for the archaeological excavation of a round barrow located near Crowlink, East Sussex, University College London, Field Archaeology Unit, Project No. 696 , - unpublished.

Hackmann, J. T., 1976, Bullock Down, Sussex, Ancient Monuments Laboratory Report Series, (old) 2305 . - unpublished

Holgate, R. 1986, Prehistoric sites threatened by coastal erosion between Seaford Head and Beachy Head, East Sussex, Sussex Archaeol. Collect., 124 , pp243-44.

Institute of Geological Sciences, 1971, Geological Survey of Great Britain, Sheet 334, Eastbourne - Drift.

Linford, N. T., 1993, Geophysical survey at Reigate Priory, Surrey, Ancient
Monuments Laboratory Report Series, 44/93 . - unpublished

Scollar, I., Tabbagh, A., Hesse, A. and Herzog, I. (eds.), 1990, Archaeological Prospecting and Remote Sensing, Cambridge.

Soil Survey of England and Wales, 1983, Soils of England and Wales, Sheet 6, South East England.

Thompson, R. and Oldfield, F., 1986, Environmental Magnetism, Allen and Unwin.

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