158 THE ESSEX NATURALIST.
As at site DII above many of the fragments of red pottery—
both fine and coarse—dug up at holes A and B were found to be
coated with a thin glaze, resistant to hydrochloric acid.
The coating on the thin ware of finer quality was a dull white,
and the biscuit was doubtless coated with white clay slip. The
glaze on the crude ware was pale green, and closely resembled the
green slag associated with it in the burnt earth. The presence of
this green glaze as a hard, continuous and even coating on the
surface of the ware suggests that the pots, before firing, were
coated with an alkaline siliceous11 slip to render them less
permeable to liquids. Some of the slag found in the deposit was
less glassy and may have been spilt or discarded slip that had
subsequently come into contact with fire; on the other hand, the
glassy slag is similar to that found in furnaces elsewhere of the
same period and its presence is attributed to the action of heat on
the furnace lining. Lead was absent both from the slag and the
glaze.
Experiment shows that if the clean surface of a piece of crude
ware is smeared with sod. carbonate, covered with a second piece
and then heated to redness in a fire a pale green glaze is produced
very similar in colour and properties to the slag. Solid salt under
the same conditions of firing is found to have no action on the
pottery.
So far nothing resembling the sides or dome of a furnace has
been found at any of the Canvey sites. It is possible that the
larger sized pieces of pottery were fired in open furnaces, the
smaller pieces in roughly built kilns, the domes of which were
demolished after each operation. If this were the case the
burnt clay thus produced might well account for a considerable
proportion of the waste material left on the site.
The presence of flints scattered on the surface at Leigh Beck
Marsh, and also found by Mr. C. Stamp below the surface,
associated with bones and fragments of Romano-British pottery,
suggests that some at least of the houses occupied by the workers
may have been built with flint walls and roofed with tiles similar
in construction to those disclosed by recent excavations near the
Roman theatre at Verulamium. Fragments of tile in quantity
11 See Analysis by J. H. B. Jenkins, Exploration Committee's Report, 1906-7, p. 23