89
snow cover, despite the low temperatures, otherwise the wedges would be blocked and ingress
of the sand prevented. Within the wedges laminae, broadly parallel with the sides (Plate 10), are
considered to indicate repeated, possibly annual, contraction allowing more sand in each year
(Rose et al., 1985). Thus the sand wedges are very important climate indicators. Sustained
temperatures below -22°C prevailed in an arid climate with precipitation levels of less than 100
mm per year (French, 1976).
The recognition of the sand wedges in the Chelmsford area was only the second record in
Britain, the first being in Cheshire. The latter are much younger in age, forming during the
Devensian cold stage, probably between 120,000 and 20,000 years ago. There is now clear
evidence from two cold stages, the Anglian and Devensian, and fragmentary evidence from
others, that there is an initial arid phase (tundra desert) as the climate deteriorates, followed by a
short-lived glacial advance. The information from the Devensian indicates that the arid phase
occupies much of the cold stage. The oxygen isotope record suggests that this might be a
pattern applicable to all cold stages (Fig. 1) and therefore a long term factor to consider when
the present interglacial comes to an end. anthropogenic global warming permitting.
Humid tundra phase
At Broomfield and Newney Green, the wedges are deformed into rounded forms and at Newney
Green the wedges arc accompanied by periglacial involutions (Plate 11). Involutions arc
common in many Essex quarries. The rounding of the wedges may be due to instability brought
about by saturated ground conditions, in which the sharp wedge form is slowly lost as the
sediments react to the saturation, Involutions are rounded or irregular forms also associated
with unstable ground conditions, whereby heavier or more competent sediments sag into lower,
saturated sediments that cannot bear their weight. At Newney Green, the overlying sediment
was cleared for quarrying purposes but fortuitously revealed the involutions forming a
polygonal pattern (Plate 12) of a type that has been replicated in laboratory experiments into the
formation of load structures. At an early stage, these structures were thought to be associated
with the amiual freezing and thawing of the uppermost metre or so of the ground that occurs in
periglacial areas, but later it was realised that it was the thaw stage that was important for the
process described above to occur. As the process involves saturated conditions, it was
associated with humid conditions. More recent work at Newney Green and at Barham. Suffolk
(Murton et al, 1995) has looked to analogies in arctic Canada and indicated that it was only the
major thaw phase al the end of the periglacial period that was significant, not the annual. Also,
the process does not necessarily involve humid conditions as sufficient frozen groundwater
could build up over time for the forms to occur on thawing. However, humid conditions are not
ruled out, especially as many areas of involutions have no associated aeolian sediments or other
indicators of aridity. The subsequent advance of the glaciers would also suggest a change to
more humid conditions, to provide the precipitation for them to develop.
Glacial phase
It was during the glacial phase that the major changes to the landscape occurred. As the ice
advanced into Essex, it not only deposited sands and gravels (glacial outwash) and till (boulder
clay), the Lowestoft Formation, to a minor extent it also deformed underlying sediments, but
much more significantly it diverted the Thames and created deep sub-glacial channels.
Essex Naturalist (New Series) 16 (1999)