278 On a Roman building at Holbury, near Dean. 
of the pillars (perhaps of timber) which supported lean-to roofs, 
sloping from the outer walls. 
Indications of fire were very frequent, a succession of shallow 
bowl-shaped depressions, marked “ashes” in the plan, and containing 
charcoal and other ashes, extending up the centre of the area from 
the South, and others being found in the positions assigned to them. 
From one, on the Eastern side, a shallow trench, also containing 
ashes, stretched towards the South. There was an artificial hearth, 
3ft. by 2, against the centre of the North wall, formed of hard 
cement laid upon a footing of clay of a deep red colour, quite 
different from the natural clay of the spot, and 18in. in depth. 
Near the West wall, and within the space between the two piers on 
that side, was another smaller fire-place, the hearth formed of a 
pitching of flints, and having a back of the same material. 
Two small blocks of Tisbury stone, 15in. by 9 and 10, resting on 
the clay floor, and fixed in their places by chalk rammed round 
them, one just within the South wall, near its centre, and the other 
about the same distance from the East wall, must have had some 
purpose, as yet un-ascertained. 
I confess to some disappointment at the paucity of facts I have 
obtained, nor can I at present offer any elucidation of those before 
me. I have at all events rescued from oblivion a Roman building 
hitherto unknown, and one differing (as I believe) from any 
previously discovered. Whether it was a receptacle for cattle, or a 
store-house for articles of merchandize, or for provisions—I must 
leave for others to determine. It was a durable structure at all 
events, made strong, perhaps to resist marauders ; and its foundations 
lie now uninjured, with trifling exceptions, beneath the turf which 
has concealed them for fifteen hundred years. I may add that a 
single coin, of the second brass, with the sacred monogram 
upon it, proves the post-Christian occupation of the building. 
An additional item of information subsequently obtained (Oct., 
1870), must be placed on record. The pointed crowbar so usefully 
employed already having been again in requisition in the same field, 
about 60 yards S.W. from the site of the building, after striking 
upon something hard, suddenly left the hand of the workman, and 
.» 
