away by the peasantry, since 1830, and the site only appeared a hollow filled with large flints and a few fragments of tile, and some tesseræ. Still we felt disposed to make further excavations, and immediately decided on putting four able men to work carefully to the immediate south of the hypocaust, conjecturing that if more building existed it would be to a sunny exposure. The next day, the 4th April, the men began early, and by the afternoon of the 5th we had laid bare long lines of flint wall, enclosing four oblong spaces, apparently having formed two small rooms, 17x13 ft. each; and two passages, 6x17 ft. each. The passage to the east was hollow in the whole length of its centre (concave), apparently having fallen in from the sides, but the pavement little disturbed by it. The plan we pursued was, when we hit on a bit of flint wall, to follow it and clear it out completely, and then clear the spaces included by these walls. The rooms and passages had portions of their pavement remaining, composed of small square tesseræ of white or red colour, of rude work, without border or figure. It is possible, from two-thirds of the pavement of the two rooms having been removed, that there may have been in the now vacant parts a better sort of pattern, or finer work; but I doubt this, as there is no symptom of such work at the north side, which is well preserved.
We found neither coins nor pottery, nor any object of interest, excepting the fragment of an urn, some nails, excellent mortar, and the tesseræ, and fragments of the stony slate alluded to by Sir J. Pollen, as probably composing the roof. This substance is not found anywhere in the immediate vicinity. Dr. Mantell states that it is the Freshwater limestone of the Wealden formation. The way the floor was laid showed great care; the soil having been removed till the workmen reached either a solid chalk or stiff clay, and then concrete was laid on, to the height required, to make the level for the tesseræ, thus guaranteeing the inhabitants from all damp exhalations. The stone walls only remained about 8 inches above the pavements; they were set in strong mortar.
On the 6th, in the morning, we still continued our diggings, so as to lay bare a continuation east and west of the long north wall, but found no more rooms or pavement. I was then obliged to leave for London, and I understand the works ceased; but the following day (Sunday, the 7th) troops of boys and peasantry assembled on the spot, and carried away and picked out the tesseræ, which probably induced Sir John Pollen not to continue digging. I have since learned from Sir John that he is aware of numerous remains in his neighbourhood. The beautiful specimen of pavement at Thruxton is distant only about three miles from the spot. Another, which has never been more than partially opened, on Mr. Best's estate, in the parish of Abbots-Anne; and he adds, "I have been told there is one at Clanville," (all these are places in the vicinity,) but this I neversaw." He has also supplied me with a fragment, the lower portion of a red vase, dug up at the west bridge of his own park, evidently Roman; and some horns, and fragments of pottery, from Privet Wood, in the park itself.
The vicinity of all the places which have been mentioned, to the great Roman Road leading from Old Sarum to Silchester, seems sufficiently to account for their locality, and suggests that they are Roman.
Perhaps some person who has studied the Roman roads and stations, accurately, of Wiltshire and Hampshire, may know whether the long nearly straight line of road, to be traced on the Ordnance Map, from Marlborough through Savernake Forest, in a south-east direction, although
