374 The Excavation of a Roman Well near Silbury Hill. 



readied a mass of sarsen stones of various sizes. These stones 

 were loose and easily removed, but under them, to a. depth of 5|ft. 

 we came across one sarsen which took the united strength of seven 

 men to raise. It must have weighed nearly half-a-ton, it was 

 standing on end, and very difficult to rope, as the well at this 

 depth contracted to a diameter of 4ft. 6in. From subsequent 

 "finds" it is a reasonable conjecture that Messrs. W. and H. 

 Cunnington failed to remove this impediment, because from this 

 point downwards relics came up in every bucket, which would 

 never have escaped the fingers of such enthusiasts ! 



The relics consist of necks, bases, handles, and pieces of pottery 

 in a great variety of shapes, colours, and sizes, pieces of Samian 

 ware, various bronze relics, beads, fragments of glass vessels, antler 

 picks; bones of deer, dog, sheep, ox, pig, rat; flint flakes (3), 

 square-headed iron nails, perforated roofing tiles, an iron bucket- 

 handle clip, a moulded freestone corbel and base of a column, 

 oyster and snail shells ; and thirty-three Eoman coins. 



Careful record was kept of each article found, with notes of its 

 depth, position, and association with other relics, but as the date 

 of filling in must necessarily be posterior to that of the most recent 

 coin, a lengthy list of the positions of the various articles brought 

 to light would be superfluous — the coins being scattered indiscrimi- 

 nately throughout, the most ancient (a sestertius of Trajan) being 

 found at 18ft. deep, and the more recent coins of Valentinian I. 

 being found at the bottom of the excavation. 



It took six days to reach the solid undisturbed chalk bottom, 

 which, upon measurement, was found to be 26ft. from the surface. 

 The well was unsteined and waterless, but several holes or chalk 

 tubes which once furnished the supply were noticed. 



The work was hindered throughout by large sarsen stones, which 

 were probably thrown in to get rid of them, and these in falling 

 from a height would naturally wedge in. Several stones had to 

 be broken ere they could be moved — indeed a very difficult pro- 

 cedure considering the cramped space for shoulder and elbow work. 



After consideration I conclude that the well had been filled in 

 while Valentinian I. (364 — 375 A.D.) wore the purple, with j 



