EXCAVATIONS AT THE BECKHAMPTON ENCLOSURE, AVENUE AND COVE, AVEBURY 



255 



to the south-east of F.53 (Figure 5). It is argued 

 that these, along with F.87, form part of an original 

 terminal to the Avenue. They are set 40m apart on 

 a line that bisects F87 at its mid point. They were 

 termed 'Beacock Holes' after the student excavator 

 of the first of these features. 



Both features were distinct from the other 

 excavated stone sockets by virtue of their pure chalk 

 rubble fills, making them difficult to detect on the 

 surface (F.83 only appearing after a period of 

 weathering). Both were very similar in morphology 

 and fill; c.3.0 x 2.0m in extent and 0.5-0. 7m deep, 

 with moderately steep sides and flattish bases. 

 Deeper, sloping linear recesses on the south sides 

 of both bases are possibly leverage points to facilitate 

 the erection or removal of the stones, and slight 

 hollows along the northern edge of F.54 could 

 represent settings for anti-friction stakes. The base 

 of F.54, and to a lesser extent that of F.83, were 

 extremely compacted and smoothed, consistent 

 with compression from having held large stones. 

 However, the stones could not have stood long, and 

 seem to have been deliberately removed soon after 

 erection, the pits then being backfilled with clean 

 chalk rubble. 



Stakeholes 



Once the surface of the chalk had had time to 

 weather, numerous stakeholes were observed, 

 particularly in the northern half of the area. Where 

 possible, these were base-planned, thorough 

 investigation being confined to a 1 x 1 0m area in 

 the west corner of the trench. Over 200 stakeholes 

 were revealed in this one area, a number being 

 sealed by the fill of the ridge-and-furrow, indicating 

 a pre-late medieval, and most probably prehistoric, 

 date for most (cf. similar concentrations of such on 

 later Neolithic sites at Coneybury (Richards 1990, 

 138) and Down Farm (Green 2000, 73)). Though 

 no discernible structural patterns could be 

 recognised within these, their distribution is non- 

 random and several sets appear to describe short 

 arcs. They could easily represent a palimpsest of 

 temporary dwellings, fence lines and compounds. 



DISCUSSION 



The enclosure 



Sufficient of the enclosure has now been excavated 

 to be certain of its character. In all of the sections 



examined the ditch is narrow, shallow, flat-bottomed 

 and dug as a series of intersecting pits. Its segmented 

 form was perfectly displayed in the long lengths of 

 ditch exposed during the 2000 season. Small 

 causeways, some too narrow to have acted as points 

 of entry into the monument, were discovered in 

 three locations (Trenches 22, 23 and 24). With 

 minor variations, the sequence of filling is identical 

 in all the excavated sections: primary chalk rubble, 

 followed by the formation of a thin secondary silt 

 and intermittent soil, and then a uniform backfill 

 deposit of chalk rubble (almost certainly re- 

 deposited bank material pushed from the inner 

 side). Whilst localised scoops may have been dug 

 to receive 'decommissioning' deposits prior to the 

 episode of levelling, there is no evidence for re- 

 cutting. Preliminary analysis of molluscan samples 

 taken in 1999 suggests conditions of grazed 

 grassland throughout the life of the enclosure. A 

 certain amount of depositional activity followed 

 immediately on from the digging of the ditch, 

 involving placed spreads of butchered animal bone 

 (cattle and pig), soil, and small amounts of Grooved 

 Ware and worked flint. This is particularly 'event 

 like' (limited to a specific horizon) and largely 

 limited to the area closest to the main eastern 

 entrance. Despite investigation, no indications have 

 been found of contemporary activity within the 

 interior of the enclosure - it is remarkably 'clean'. 



Radiocarbon determinations obtained on bone 

 recovered during the 1999 excavations suggest the 

 enclosure ditch was dug, began to silt and was finally 

 backfilled and levelled within a short period of time. 

 The span of dates runs between 2885-2200 cal BC, 

 though Bayesian calibration of these narrows the 

 range to 2650/2500-2510/2300 cal BC. These mid- 

 3rd millennium BC dates are supported by sherds 

 of Grooved Ware since recovered from the base of 

 the ditch. 



There is no doubt that this is an unusual 

 monument, and one that was rather 'out of time'. 

 Its form is highly reminiscent of 4th millennium 

 BC causewayed enclosures, though it also has 

 affinities (in terms of scale and restricted access, 

 though not geometric regularity) with the earlier 

 Stonehenge 1 (Cleal er al. 1995) and Flagstones 

 enclosures (Smith et al. 1997). Yet, it must be more 

 or less contemporary with the Avebury henge (Pitts 

 and Whittle 1992). Its anachronistic form may have 

 been quite deliberate, making direct reference to 

 earlier traditions, and thus standing to some degree 

 in contrast or opposition to the novel values and 

 bodies of sacred knowledge presented by a 



