dential purposes, and as fortifications, and many are recorded as having been erected as castles. All this is probably quite correct, but the point that interests us here is, that there are nearly one hundred examples of truncated cones of earth thrown up in England after the Roman times, and not one before. If this is so, the conclusion seems inevitable that Silbury Hill must belong to the latter age. Whether this conclusion can be sustained or not, must depend on what follows from the other monuments we are about to examine. The evidence of the monument itself, which is all we have hitherto had an opportunity of bringing forward, may be sufficient to render it probable, but not to prove the case. Unless other examples can be adduced whose evidence tends the same way, the case cannot be taken as proved, however strong a primâ facie presumption may be established.
Though a little distant, it may be convenient to include the Harden circle in the Avebury group. It is situated in a village of that name seven miles south of Silbury Hill. When Sir R. Colt Hoare surveyed it fifty years ago, the southern half of the vallum had been so completely destroyed, that it could not be traced, and he carried it across the brook, making the whole area about fifty-one acres.[1] My impression is that thisis a mistake, and that the area of the circle was only about half that extent. The rampart was of about the same section as Avebury, and the ditch was inside as there. Within this enclosure were two mounds, situated un- symmetrically, like the circles at Avebury. The greater one was opened with great difficulty, owing to the friable nature of the earth of which it was composed; and Mr. Cunnington was con-
- ↑ 'Ancient Wiltshire,' ii. 5. Unfortunately there is no scale attached to the plan of the Marden Circle, and no dimensions quoted in the text.
