212 THE ESSEX NATURALIST.
The line, dividing this district from the remaining two areas,
follows the Mardyke marshlands, the crest of the hilly land
near Billericay to Danbury, and swings in a crescent-shaped
curve along the margin of the Boulder Clay to Colchester and
Tendring. The remaining areas—more clearly patterned—lie
to the east of the county, and are divided by the Danbury mass.
The fact that strong physical features form the divisions is
very significant.
Naturally any routeway nexus shows a strong tendency to
be related to the slopes of the land. Sharpe appears to have
ignored this obvious possibility in drawing his centuriation map,
and, in this respect, seems deserving of Haverfield's adverse
criticism.10 Many of the former's suggested vestiges seemed to
be roads—probably of Saxon date—following the trend of the
landscape. Not being accurately laid out they do not lie
truly parallel to one another, and therefore support Haverfield's
remark that "the straight line and the right angle formed the
great difference between Roman and other roads." Roads
only approximating to the rectangular pattern have been
ignored in the centuriation map of Essex. It is noticeable from
an examination of the map that, whilst the approximately
straight type of road appears in all parts of the British Isles,
the accurate rectangular pattern of routeways is only to be found
within the Civil Area of Roman Britain.11
Considering the suggested Essex centuriation, it is apparent
that many of the known Roman roads of the O.S. map fall into
place within the scheme, appearing in certain cases to be the
"limites"—roads dividing adjacent centuriated districts.
Furthermore, many of the Roman roads pre-dated settlement,
and, for this reason, would have no part in later centuriation
schemes—Stane Street perhaps coming within this category.
Portions of Miller Christy's suggested Roman roads, deduced
from local fieldwork, also figure in the scheme. Again, the Roman
settlements and villas, indicated on the O.S. map of Roman
Essex, fit into the network with surprising ease—the villas
appearing set back a little from the road in the customary
10 English Historical Review, 1918.
11 This examination included many areas beyond the Civil District and regions of soils similar
to those of Essex, and comparable in their physical character. The Vale of York provided an
ideal comparison, but failed to furnish evidence of centuriation.