238 THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE 
coarse, of first millennium BC date. These are very 
small and have no diagnostic attributes and so 
could date from the Late Bronze Age through to the 
latest pre-Roman Iron Age. The upper fill of this 
ditch contained rim sherds from a carinated bowl, 
dated elsewhere to the fifth century BC (Cunliffe 
1984, fig. 6.54-6.55). 
The upper fill of ditch 40 also produced two 
undated sandy fabric coarseware sherds and a tiny 
fragment of first-second century samian, along with 
avery worn, illegible Roman coin. Ditch 3 (Site 15), 
the Earl’s Farm Down linear, produced two sherds 
in a coarse sandy ware, similar to a fabric from ditch 
40, from its upper fill and lynchet 42 (Site 19) 
produced a sherd of fine, white New Forest colour- 
coated ware and one of fine, micaceous Oxfordshire 
ware, both of third to fourth century date. One body 
sherd of Black Burnished ware was recovered from 
a layer on Site 18. 
Other finds consist of 230 generally 
undiagnostic struck flints, amongst which are 19 
cores and core fragments, 4 scrapers and 9 edge 
retouched flakes, and a very small amount (37g) of 
burnt flint. 
ANIMAL BONE 
by Sheila Hamilton-Dyer 
Very littlke bone was recovered and only 19 
fragments were scanned from three probably Late 
Bronze Age ditches (ditches 3, 7 and 40). These 
comprise 2 horse, 3 sheep/goat, | pig bone together 
with 13 unidentified fragments from hoofed 
animals. Ditch 40 (context 39) contained part of the 
jaw of a small female horse. The sieved samples 
contained mostly amphibian and small mammal 
remains; the bones and teeth of a shrew, mouse and 
voles were present. Samples from the lower fill of 
ditch 3 (context 19/20) and the upper fill of ditch 40 
(context 38) contained common eel (Anguilla 
anguilla) vertebrae, the only fish recovered from the 
assemblage. 
LAND MOLLUSCA 
by Michael 7. Allen and S.F. Wyles 
The two large linear ditches excavated on Earl’s 
Down Farm (ditches 3 and 40; figure 9) were 
sampled for Mollusca by the excavator. Although 
there was no dating evidence from the lower fills, it 
is thought that both ditches date from the later 
Bronze Age. The aims of the analysis were to 
determine the environment and land use of the area 
into which the ditches were cut and existed, and to 
attempt to determine their function. A further aim 
was to see if the environment and land use 
determined by mollusc analysis was compatible 
with the assumed late Bronze Age date of the 
ditches. 
Standard methods of molluscan analysis were 
employed as outlined by Evans (1972, 44-5). 
Mollusc nomenclature follows Walden (1976). The 
results are given in Tables 3 and 4, and for ditch 3 as 
a histogram of relative abundance with Pupilla 
muscorum being calculated over and above the 
remaining assemblage (see below). 
Results 
The assemblages from both ditches typically 
comprised open country species dominated by 
Pupilla muscorum, which is consistently high (over 
80% of one sample). Pupilla often occurs in large 
numbers and is known to have been abundant in 
the area in the Late Neolithic/Early Bronze Age at 
Durrington Walls (Evans 1971; 1972, 148), Earl’s 
Farm Down (Kerney 1967), and the barrows on 
King Barrow Ridge (Allen and Wyles 1994). PR 
muscorum favours areas bare of vegetation such as 
patches of broken ground induced by sheep grazing 
on grassy chalk slopes but also the ditch micro- 
environments themselves where patchy vegetation 
and small actively eroding areas of bare chalky soil 
may occur. 
The super-abundance of this one species creates 
problems in interpretation as it distorts and masks 
the relative and absolute trends within the 
remaining assemblage (Thomas 1985, 134). In 
order to lessen the obscuring effect of a super- 
abundant species histograms can be plotted in 
absolute numbers but in this case the numbers of 
shells were both too high and too variable. For these 
reasons Pupilla was calculated as_ relative 
percentages over and above the remaining 
assemblage and thus the relative trends of other 
species could be observed. The diagram produced 
in this way for ditch 3 (Table 3) typifies the 
sequence and is published in (Figure 10), while that 
for ditch 40 (Table 4) is available in archive. 
Both ditches have similar assemblages so are 
discussed together. The assemblages, excluding 
Pupilla, are typical of very open environments and 
are dominated by Vallonia costata (up to 80%). The 
predominance of V costata over its cogener Vv 
excentrica is indicative of short-turfed grazed 
downland. Trichia hispida and Helicella itala have a 
constant but low presence throughout the 
