"cesses, he considers, might form two successive "stages of one operation, as at present practised "with the Catalan forge." The Wilderspool discoveries prove that this is what actually took place. YEAR 1900.

They

Inside the fortified area the foundations of only four walled enclosures which could have belonged to ordinary dwelling-houses have been traced; but several clay floors, consisting of three-inch layers of stiff, well-puddled boulder-clay, burnt on the surface, or sprinkled with crushed tiles to harden. them (opus testaceum), have been uncovered. all are in alignment with the via on the west and north sides, and adjoining the rampart on the east side. The presence of only slight traces of enclosing walls or partitions, and of quite a number of iron nails and fragments of stone roofing tiles in their vicinity, indicates that their superstructures were mainly of wood. As the subsoil over the entire site is a pure glacial sand bed, the presence of clay is always a sure indication of local industry, and the material for the clay floors has in every instance been previously prepared and brought in by the neighbouring potters. This is proved by numerous enclosed fragments of a kind of earthenware which was manufactured in a potter's kiln discovered, in July, 1900, at Stockton Heath, a quarter of a mile direct south from the fortification.

Potter's Kiln.-The kiln referred to was come upon accidentally in digging a sand pit, but was carefully explored and measured before being destroyed. The structure was entirely built up of massive boulder-clay in two separate parts, with no direct communication between them, namely, (1) a heating furnace and kiln floor above it, and (2) a globular oven or drying chamber in the rear of the former.

In shape the furnace resembled the top of the skull of some animal having a low forehead, and three apertures underneath opening into a hollow dome-shaped interior, the plan being square in front and semicircular behind. Its appearance when first uncovered is shown in the accompanying photographic plate. (Plate VII.)

The two piers supporting the roof, and dividing the inside into three separate fire-holes, were I foot 8 inches long and 4 inches thick. The extreme length of the interior was 3 feet 4 inches, and therefore the hollow chamber at the back, where the fire from the three fire holes was concentrated, measured 1 foot 8 inches from front to back. The external dimensions were:-Vertical height, 2 feet to 2 feet 4 inches; width, 5 feet; original length, 6 feet, the front portion having been broken down for a distance of 2 feet, as shown by the red colour of the adjoining sand bed. The thickness of the encircling walls varied from 2 inches to 9 inches.

The heat was conveyed through the roof into the kiln, where the vessels were arranged for baking, by two holes 2 inches in diameter, placed a foot apart at 6 inches from the ends of the piers, and there was a flue 4 inches in diameter in the rear of the right hand stoke hole.

Nothing remained of the circular wall of the kiln except a single block of hammer-dressed sandstone, measuring I foot by 10 inches by 3 inches.

Six feet from the front of the fire-holes, and three feet from the present surface, there was a dense mass of fragments of broken and distorted vessels of soft red unglazed ware, of well-known shapes, intended only for useful purposes; patella, patina, ollæ, urcei, lagenæ, ampullæ, urna, and mortaria or pelves. Many of the larger fragments have been. removed to the Warrington Museum, where there is a large collection of whole and restored vessels

from the locality, evidently of the same kind of ware. Among the latter are four mortaria, stamped across the rim with the names of two of the local potters, BRUCI (for Brucus) and BRICOS, in rude and reversed characters.

Clay Floor (I).—The largest of the above-mentioned clay floors was situated at the west end of INSULA II, on the north side of the fortification, 8 to 10 feet from the inside edge of the via, with a frontage along the latter of 42 feet and an average width of 30 feet. There were between it and the via several patches of a cobble-stone pavement (measuring in one instance 15 feet by 3 feet), which appeared to have covered the greater portion of the interval. On the east side it was bounded by the tumbled foundations of a dwelling-house, uncovered in 1898, and described in my preceding report.9

Wall. On the west side there was the foundation of an enclosing wall, 5 to 6 feet wide, formed by a bedding of gravel overlaid by boulder-clay and sandstone rubble, extending at least 26 feet at right angles to the via.

Pavement. On the surface of the floor and along the inside of the last-mentioned wall there was a level pavement, apparently of secondary construction, 22 feet in length and 18 feet in width, formed of unshaped blocks of sandstone, which may have covered an enclosed court-yard, or a street running north and south. The vertical sections given (figs. 3 and 4, plate III, No. 3) show that the stones of this pavement lay at a depth of only 6 to 8 inches below. the present surface. A wall of rude and indefinite construction and no depth of foundation appears to have skirted the paved area along the east side.

Though structural remains were wanting along the south side of the clay floor, the soil was blackened with charcoal for a depth of from 1 to 4 feet,

9 Trans. Hist. Soc. of Lanc. and Ches., vol. xiv, n.s., p. 19.