228 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE 
composition of the livestock. Pigs, with their high 
reproduction rate and limited usefulness as adults, 
are an ideal feasting animal. They also prefer 
woodlands and are excellent at clearing regenerating 
woodland. The problems of interpretation have 
frequently been discussed (Grigson 1982; Richards 
and Thomas 1984; Maltby 1990), and the degree to 
which their prevalence reflects selection for feasting 
or the amount of local woodland is unresolved (see 
Bradley 1984, chapter 3). The presence of beaver 
remains may also indicate that the nearby Avon 
valley was more wooded than today. 
Although the sample is very small, the high 
number of pig bones, the selection of joints, and 
lack of sheep bones all fit well with material from 
Durrington Walls where the midden in particular 
contained a mass of pig bones (Wainwright and 
Longworth 1971). The association with Grooved 
Ware and flint artefacts is significant; it is highly 
probable that the material from this group of pits 
reflects the ritual activity at Durrington Walls, and 
indeed in the wider Stonehenge environs. 
PART 2: AVON VALLEY 
FLOODPLAIN SEDIMENTS: 
THE PRE-ROMAN 
VEGETATIONAL HISTORY 
by Robert G. Scaife 
The northern floodplain of the River Avon, 
approximately 300m east of Durrington Walls, was 
surveyed and augered to provide a detailed cross- 
profile of the valley alluvium (Figures 2 and 6). 
Samples for pollen analysis were obtained from the 
deepest sequence of peat and organic sediments. 
The location was of special interest because of the 
possibility of pollen preservation in alluvial 
sediments and peats in proximity to Durrington 
Walls (Wainwright 1971). This might enable 
correlation with Dimbleby’s ‘on-site’ pollen analysis 
of the henge (in Wainwright and Longworth 1971, 
332-4) and Evans’s environmental changes as 
shown by molluscan analyses (Evans 1971, 329-37). 
The full report is available in archive. 
STRATIGRAPHY 
by Michael F. Allen and Robert G. Scaife 
Ten boreholes were examined on the northern side 
of the River Avon floodplain. The surface of the 
floodplain, which is now largely under pasture, 
supports a gleyed soil caused by a fluctuating 
ground water table. In places, this also comprises 
highly oxidised peat. The lithostratigraphy ranges 
from grey alluvial silts with varying thicknesses 
and degrees of organic content to humified fen peat 
with only small quantities of inorganics. A 
maximum depth of 1.68m of monocot peat and 
organic silt was recorded in borehole 6 (Figure 6) 
from which detailed pollen analysis was obtained. 
The character of these sediments is given below. 
These rest on sands, which appeared in the field to 
be glauconitic and derived from the Upper 
Greensand, and in places on_ gravel of 
undetermined age (e.g. borehole 1). 
0 -0.30m Oxidised chocolate brown, humified peat 
with silt. Monocotyledonous remains were 
evident. Occasional chips of flint present. 
0.30—0.35m Wetter, darker brown peat and grey silts 
containing monocotyledonous remains and 
occasional flint chips. 
0.35 -—1.47m Chocolate brown peat. Well humified but 
with identifiable monocotyledonous 
remains. Charcoal present at 1.12m and 
Roman pottery at 1.15 m. Sharp, well 
defined junction with he underlying very 
dark brown to black highly humifed peat 
(DURR: 4 DURR: 3 DURR: 2). 
1.47—1.60m Very dark brown to black humified peat 
with little visible structure (DURR: 1). 
1.60-—1.74m Grey-green glauconoitic fine to medium 
sand. Weathered or transported Upper 
Greensand. 
The gravel is likely to be river terrace gravel or 
gravel sheets laid down during the late Devensian 
or early Flandrian (c. 10,000 BP). Calcareous marls 
and silts form a thin deposit over the gravel and 
glauconotic sand, and varying depths of floodplain 
and local channel peat form most of the floodplain 
profile (Figure 6). A horizon of fine charcoal 
fragments was recorded at 1.12 m in borehole 6. In 
this same sequence, a single sherd of undiagnostic 
Roman pottery was recovered from a depth of 1.14 
m indicating that most of the floodplain sequence is 
probably of post Romano-British date. 
There was a noticeable change between the 
lower dark brown/black peats and overlying lighter 
peat and alluvial silts noted in a number of the 
profiles. This is one of a number of possible hiatuses 
in the alluvial stratigraphy which have been detected 
in the pollen/biostratigraphy. Immediately adjacent 
