surface strewn with Roman refuse, such as tiles, pottery, and oyster shells. This fixes the date of the layer above as post-Roman; but the low position of the Roman land-surface, now at about mean-tide level, is due in great part to shrinkage since the marsh was embanked and drained—it is unconnected with any general post-Roman subsidence of the land.
Beneath the Roman layer occurs more marsh-clay and silt (2), resting on a thin peat (4) which according to Mr Whitaker is sometimes absent. Then follows another bed of marsh-clay (5), shown by the engineer as four feet thick, but which in places thickens to six or seven feet. Below this is a thick mass (6) of reedy peat (the "main peat" of Mr Spurrell), which is described as consisting mainly of Phragmites and Sparganium, with layers of moss and fronds of fern. The other plants observed in this peat were the elder, white-birch, alder and oak. Associated with them were found several species of freshwater snails and a few land forms; but the only animal or plant showing any trace of the influence of salt water was Hydrobia ventrosa, a shell that requires slightly brackish water.
The main peat rests on another bed of estuarine silt (7 and 8), which seems to vary considerably in thickness, from 5 to 12 feet. It is not quite clear from the descriptions whether the "thin woody peat" of Mr Whitaker and the "sand with decayed wood"
