210 THE ESSEX NATURALIST
separation between the higher of the ancient roads and the base
of the 19th century road foundation where the ground had
obviously been made up to a level. It is of particular interest to
note that the modern road follows the alignment of its pre-
decessors most precisely, having exactly the same axis, and the
ancient ditches are spaced evenly from the centre of the modern
road. No material dating evidence was obtained from the
narrow trench, and it would have been surprising if it had been
forthcoming considering the slight nature of the excavation in
proportion to the whole length of the road. However, there can
be little doubt of the significance of the section obtained, (Fig. 1).
Roadbed I is the Roman road laid directly on the natural
gravel, the metal may have been obtained by grading the material
obtained from the ditches and by gravel digging at intervals
along the line. The metalling is clearly the same type of gravel
as that obtained from the Flood Plain Terrace. The ditches of
this road were heavily silted and the silting is continued across
the whole surface. This suggests a period of disuse or of lack of
maintenance, which latter may be more likely since it is difficult
to imagine such an important transit line going out of use what-
ever the condition of the country. Above this a new level of
roadbed was laid down—Roadbed II. This extends slightly over
the earlier ditches and its drainage ditch was cut into them. This
is almost certainly the mediaeval road which appears to have
remained in use until the 19th century. The considerable thick-
ness of Roadbed II must be attributed to repair and remetalling
over a long period. It is perhaps improper to suggest that there
is a relationship between the 0.40 m. thickness of the supposed
Roman road, representing nearly 400 years of use and upkeep
and the 0.70 m. of the upper ancient road which should represent
some 700 or 800 years of similar treatment.
It can be assumed with some certainty that since the ancient
alignment is on the same axis as the existing road at this point,
the direction of the ancient road may be inferred from the course
of the modern road. This fact is of some importance since it sheds
light on the crossing of the Lea.
Before considering the river crossing and possible routes into
London, one final point remains. The silting levels extend widely
to the sides of the mediaeval roadbed; and the area to the north
of the modern road, even to-day, gives trouble with surface drain-
age. It would seem that although culverts and proper through
drainage were provided for the Roman road, this was not done,
or the culverts were not maintained, in mediaeval times. As a
result the natural drainage was disrupted and, to a certain extent,
still is; surface water tends to collect in an old glacial period
buried channel immediately north of the road and the resulting
waterlogging is enhanced by the presence of London Clay within
a few feet of the surface.
Smith (1910) records the dredging of masonry blocks, sup-
posed to form part of a causeway, from the River Lea opposite