INVESTIGATION OF THE WHITESHEET DOWN ENVIRONS 1989-90 161 
Table 3. Whitesheet Hill: Composition and incidence of flint 
1 2 3 4 6 7 8 Burnt Broken Burnt 
Unworked 
Primary ditch - - 9 1 20 1 3 15 221 384g 
Ditch recut 4 3 6 - 27 - 13 10 329 371g 
Internal features 116 248 161 4298 12599 = 1389 16 131 6805 7706 50545g 
1 = Irregular Waste 4 = Chips 7 = Flakes from Ground 
2 = Cores 5 = Flakes Implements 
3 = Core Rejuvenation Flakes 6 = Blades 8 = Retouched Forms 
Flint-working 
Salient features of the material include the 
relatively low level of blade production (Table 3) 
and the high proportion of cores which are 
unclassifiable or fragmentary (125 out of 251, ie., 
50%). The latter is due to the frequency of burning 
(Figure 12) and to the frequency with which cores 
split along thermal fractures while being worked. 
Samples of chips, flakes and blades were 
selected for attribute analysis from contexts 
providing adequate numbers of complete, 
unbroken and unburnt artefacts with the aim of 
detecting technological variation between 
industries from: 1; the primary fill and the recut of 
the causewayed enclosure ditch (n=341), and 2; 
features located in different parts of the interior 
(n=900). A similar methodology was applied to 
1303, to examine the assemblages from what 
appeared to be earlier and later cuts in a single 
sequence of features. The results for the features in 
the interior showed little variation, either 
horizontal or vertical. 
There is only one complete blade core (Figure 
10, L4), although others are represented among the 
fragments and L9 may have been crested in 
preparation for blade production. Sixty single 
platform and 42 multiplatform flake cores were 
recovered as well as five keeled, non-discoidal 
forms. Cores were normally worked from platforms 
prepared by flaking off the rounded end of a noduie 
although thermal fractures also served as platforms. 
Plain, non-cortical butts are consequently 
~ dominant (archive). Many cores, like L7, seem to 
have been used to produce a few flakes and then 
abandoned and 17 examples may be considered to 
be no more than tested nodules. The extensive 
flaking of L5 is rare. Platforms were, however, 
sometimes rejuvenated by the removal of rather 
irregular core tablets such as L10 or flakes like L11 
struck along the junction of platform and core face. 
Core faces often exhibit a mass of hinge fractures, 
with problems in flaking caused by cherty 
inclusions in the flint and by the thick, dense lower 
layer of cortex. Flakes such as L12 may have 
resulted from the deliberate removal of an 
intractable core face. The mean weight of cores is 
84g for cores from the ditch and 111g for cores from 
other contexts. 
A single regular, subspherical flint 
hammerstone was recovered from the primary fill 
of the enclosure ditch. Some 10%-15% of flakes 
from the primary fill of the ditch and some 30-40% 
of flakes from the interior features may have been 
soft hammer struck (archive), on the evidence of 
features such as diffuse bulbs, unfocused points of 
percussion, indistinct conchoidal fractures and 
occasionally lipped butts (Ohnuma and Bergman 
1982, 169). Such features may, however, be pro- 
duced by soft stone hammers including cortical 
flint ones. Areas of battering on the cortex of some 
cores and tested nodules, among them L6, indicate 
that they did indeed serve as hammers. Cortical 
flint pebbles seem to have been similarly used at the 
Etton causewayed enclosure in Cambridgeshire 
(Middleton 1989, 45). 
Blades and blade-like flakes are concentrated in 
the middle size-range of removals, although some 
bladelets are present. They are more frequent 
among the material from the features of the interior 
than among that from the primary ditch fill, and are 
marked by relatively frequent platform edge 
abrasion and low extents of dorsal cortex (archive). 
The rarity of blades is not simply a reflection of 
the recovery of numerous, predominantly squat, 
small flakes by sieving, since visually distinguished 
blades form only 4% of the manually retrieved 
material from the primary ditch fill but 7% of the 
sieved material from the interior features (Table 3). 
In metrical terms, when removals less than 20mm 
long are excluded, as is usual in the presentation of 
breadth: length ratios, only 7% of the removals from 
the primary ditch fill and 13% of those from the 
