This trench afterwards proved to have been dug within an adjoining room (11 ft. by 11 ft. 6 in. in size), the existence of which was not even suspected whilst I was present.
On the side of the trench farthest from the buried wall (W), the mould varied from 9 to 14 inches in thickness; it rested on a mass (B) 23 inches thick of blackish earth, including many large stones. Beneath this was a thin bed of very black mould (C), then a layer of earth full of fragments of mortar (D), and then another thin bed (about 3 inches thick) (E) of very black mould, which rested on the undisturbed subsoil (F) of firm, yellowish, argillaceous sand. The 23-inch bed (B) was probably made ground, as this would have brought up the floor of the room to a level with that of the atrium. The two thin beds of black mould at the bottom of the trench evidently marked two former land-surfaces. Outside the walls of the northern room, many bones, ashes, oyster-shells, broken pottery and an entire pot were subsequently found at a depth of 16 inches beneath the surface.
The second trench was dug on the western or lower side of the villa: the mould was here only 6.5 inches in thickness, and it rested on a mass of fine earth full of stones, broken tiles and fragments of mortar, 34 inches in thickness, beneath which was the undisturbed sand. Most of this earth had probably been washed down from the upper part of the field, and the fragments of stones, tiles, &c., must have come from the immediately adjoining ruins.
It appears at first sight a surprising fact that this field of light sandy soil should have been cultivated and ploughed during many years, and that not a vestige of these buildings should have been discovered. No one even suspected that the remains of a Roman villa lay hidden close beneath the surface. But the fact is less surprising when it is known that the field, as the bailiff believed, had never been ploughed to a greater depth than 4 inches. It is certain that when the land was first ploughed, the pavement and the surrounding broken walls must have been covered by at least 4 inches of soil, for otherwise the rotten concrete floor would have been scored by the ploughshare, the tesserae torn up, and the tops of the old walls knocked down.
When the concrete and tesserae were first cleared over a space of 14 by 9 ft., the floor which was coated with trodden-down earth exhibited no signs of having been penetrated by worms; and although the overlying fine mould closely resembled that which in many places has certainly been accumulated by worms, yet it seemed hardly possible that this mould could have been brought up by worms from beneath the apparently sound floor. It seemed also extremely improbable that the thick walls, surrounding the room and still united to the concrete, had been undermined by worms, and had thus been caused to sink, being afterwards covered up by their castings. I therefore at first concluded that all the fine mould above the ruins had been washed down from the upper parts of the field; but we shall soon see that this conclusion was certainly erroneous, though much fine earth is known to be washed down from the upper part of the field in its present ploughed state during heavy rains.
Although the concrete floor did not at first appear to have been anywhere penetrated by worms, yet by the next morning little cakes of the trodden-down earth had been lifted up by worms over the mouths of seven burrows, which passed through the softer parts of the naked concrete, or between the interstices of the tesserae. On the third morning twenty-five burrows were counted; and by suddenly lifting up the little cakes of earth, four worms were seen in the act of quickly retreating. Two castings were thrown up during the third night on the floor, and these were of large size. The season was not favourable for the full activity of worms, and the weather had lately been hot and dry, so that most of the worms now lived at a considerable depth. In digging the two trenches many open burrows and some worms were encountered at between 30 and 40 inches beneath the surface; but at a greater depth they became rare. One worm, however, was cut through at 48.5, and another at 51.5 inches beneath the surface. A fresh humus-lined burrow was also met with at a depth of 57 and another at 65.5 inches. At greater depths than this, neither burrows nor worms were seen.
As I wished to learn how many worms lived beneath the floor of the atrium--a space of about 14 by 9 feet--Mr. Farrer was so kind as to make observations for me, during the next seven weeks, by which time the worms in the surrounding country were in full activity, and were working near the surface. It is very improbable that worms should have migrated from the adjoining field into the small space of the atrium, after the superficial mould in which they prefer to live, had been removed. We may therefore conclude that the burrows and the castings which were seen here during the ensuing seven weeks were the work of the former inhabitants of the space.