In one case there was a layer of yellow clay of very unequal thickness between two beds of debris, the lower one of which rested on a floor with tesserae. The ancient broken walls appear to have been sometimes roughly cut down to a uniform level, so as to serve as the foundations for a temporary building; and Mr. Joyce suspects that some of these buildings were wattled sheds, plastered with clay, which would account for the above-mentioned layer of clay.
Turning now to the points which more immediately concern us. Worm- castings were observed on the floors of several of the rooms, in one of which the tesselation was unusually perfect. The tesserae here consisted of little cubes of hard sandstone of about 1 inch, several of which were loose or projected slightly above the general level. One or occasionally two open worm-burrows were found beneath all the loose tesserae. Worms have also penetrated the old walls of these ruins. A wall, which had just been exposed to view during the excavations then in progress, was examined; it was built of large flints, and was 18 inches in thickness. It appeared sound, but when the soil was removed from beneath, the mortar in the lower part was found to be so much decayed that the flints fell apart from their own weight. Here, in the middle of the wall, at a depth of 29 inches beneath the old floor and of 49.5 inches beneath the surface of the field, a living worm was found, and the mortar was penetrated by several burrows.
A second wall was exposed to view for the first time, and an open burrow was seen on its broken summit. By separating the flints this burrow was traced far down in the interior of the wall; but as some of the flints cohered firmly, the whole mass was disturbed in pulling down the wall, and the burrow could not be traced to the bottom. The foundations of a third wall, which appeared quite sound, lay at a depth of 4 feet beneath one of the floors, and of course at a considerably greater depth beneath the level of the ground. A large flint was wrenched out of the wall at about a foot from the base, and this required much force, as the mortar was sound; but behind the flint in the middle of the wall, the mortar was friable, and here there were worm-burrows. Mr. Joyce and my sons were surprised at the blackness of the mortar in this and in several other cases, and at the presence of mould in the interior of the walls. Some may have been placed there by the old builders instead of mortar; but we should remember that worms line their burrows with black humus. Moreover open spaces would almost certainly have been occasionally left between the large irregular flints; and these spaces, we may feel sure, would be filled up by the worms with their castings, as soon as they were able to penetrate the wall. Rain-water, oozing down the burrows would also carry fine dark-coloured particles into every crevice. Mr. Joyce was at first very sceptical about the amount of work which I attributed to worms; but he ends his notes with reference to the last-mentioned wall by saying, "This case caused me more surprise and brought more conviction to me than any other. I should have said, and did say, that it was quite impossible such a wall could have been penetrated by earth-worms."
In almost all the rooms the pavement has sunk considerably, especially towards the middle; and this is shown in the three following sections. The measurements were made by stretching a string tightly and horizontally over the floor. The section, Fig. 13, was taken from north to south across a room, 18 feet 4 inches in length, with a nearly perfect pavement, next to the "Red Wooden Hut." In the northern half, the subsidence amounted to 5.75 inches beneath the level of the floor as it now stands close to the walls; and it was greater in the northern than in the southern half; but, according to Mr. Joyce, the entire pavement has obviously subsided. In several places, the tesserae appeared as if drawn a little away from the walls; whilst in other places they were still in close contact with them.
In Fig. 14, we see a section across the paved floor of the southern corridor or ambulatory of a quadrangle, in an excavation made near "The Spring." The floor is 7 feet 9 inches wide, and the broken- down walls now project only 0.75 of an inch above its level. The field, which was in pasture, here sloped from north to south, at an angle of 30 degrees, 40 seconds. The nature of the ground at some little distance on each side of the corridor is shown in the section. It consisted of earth full of stones and other debris, capped with dark vegetable mould which was thicker on the lower or southern than on the northern side. The pavement was nearly level along lines parallel to the side-walls, but had sunk in the middle as much as 7.75 inches.
A small room at no great distance from that represented in Fig. 13, had been enlarged by the Roman occupier on the southern side, by an addition of 5 feet 4 inches in breadth. For this purpose the southern wall of the house had been pulled down, but the foundations of the old wall had been left buried at a little depth beneath the pavement of the enlarged room.