Soil evaluation - The mainly Iron Age, Roman and medieval site of Carlton Colville, Lowestoft, Suffolk was visited on the 26th of February 2002. Two areas of the site were discussed with Jezz
Meredith (Suffolk County Council Archaeological Service, Field Team). These are the
prehistoric and Roman up slope area of blown sands (Area 1 ?) and a lower flooded
medieval area by a wet ditch (Area 4). Soils were examined from across the site, with
samples being taken from the uppermost slope where there was a 'spread' containing
numerous flints, pot and burned stones etc., around an area of post-hole buildings, a
hearth and ditch features.
These are mapped as the typical argillic brown sands formed in glaciofluvial drift over
chalky till (Newport 3 soil association; Hodge et al., 1983)). The typical argillic brown
sands are composed of some 300 mm ofvery dark grey (10YR3/1) medium sandy
ploughsoil (Ap horizon), which occurs over a brownish yellow (10YR6/8) sandy subsoil.
Below this level, the 'spread' occurs in a ~200-300 mm thick layer ofbrown to strong
brown (7.5YR4/4-5/6) sand, which contains artefacts. Stony brown (7.5YR4/4) post
pad/posthole fills occur in brown (7.5YR4/4) sandy soil. Locally, a hearth containing
rubified red (1 OR4/8) burned soil fragments is formed of dark reddish brown sand
(5YR3/2) and is ~160 mm thick, over yellowish brown (10YR5/6) sand.
Downslope, an area of light greyish brown ( 1 OYR6/2) sand was examined. This
soil is also very faintly mottled, and some human bones are present. A c. 1 m deep
machine cut trench found that this grey sand became more moist with depth and occurs
over a heterogeneous deposit of grey clay and chalk clasts - chalky till. Here then is an
area of stagnogley soils as recorded as part of the Newport 3 soil association. WhEre
soils are dominated by this chalky till substrate to the north of the site, typical stagnogley
soils (Beccles soil association; Hodge et al. , 1983) have been mapped.
Examination of the lower site, by the wet ditch, found it to be composed of grey
and strongly mottled sands over river gravels. Spreads of yellowish clay and stony yellowish clay on which stone pads had been placed were discussed with Jezz Meredith.
This yellowish clay is likely locally imported subsoil from typical stagnogley soils
formed in chalky till (Beccles soil association). It has probably been employed as a form
of' cob' floor/foundation. A similar material, for example, was used to make floors at
Medieval Cressing Temple, Essex (Macphail, 1995). This part of the site needs to be ·
scrutinised for possible medieval buildings/activities, which are possibly evidenced by
spreads ofthis yellowish clay.
Discussion and samples:
The sandy soils at Carlton Colville have likely undergone phases of aeolian activity, with
loss from deflation (Corbett, 1973), and sedimentation in hollows through ploughwash,
blowing and possibly trampling. Sandy soils have been probably been lost from the Iron
Age and Roman areas, by both deflation and ploughing, with a sandy colluvium infilling
low, wet ground. The chalky clay till substrate here both impedes drainage and likely
encourages the preservation ofbone, because perched soil water is probably calcareous.
This till also influences the flow of groundwater. This whole area may have been wet
through prehistory into present day times. The grey, fine mottled colour of the area
testifies to this. It's possible that this area was utilised for discard, animal wallowing, and
as a burial ground(?), but this area has not yet been fully excavated.
Up slope, a possible hollow area was formed by deflation, while a ridge is an area
where prehistoric soils (hearth, post-holes) are still extant. The hollow has been infilled
with brown soil containing artefacts (the 'spread'). This may imply that soils forming
during the life of the settlement through aeolian activity accumulated midden waste. In
order to test this, 1 thin section monolith sample was taken from the hearth, alongside a
bulk sample (Table 1). In addition, bulk samples were collected from two areas of the
'spread', a post hole fill and adjacent soil, and a sample of the yellowish brown subsoil.
This 1 monolith sample when studied through soil micromorphology (Courty et al., 1989)
and the 6 bulk samples, when tested for their organic content (LOI), phosphate (P) and
magnetic susceptibility (Crowther and Barker, 1995; Macphail et al., 2000), will permit a
clearer view about these soils and the nature of the site. Obviously, soils have been lost
by erosion, but sufficient examples were collected to allow some comparison with other Iron Age rural settlements such as at Salford, Bedfordshire and Potterne, Wiltshire, where
midden-like soil spreads have been studied previously (Macphail, 1997; 2000).