Notes on a Roman Building, and Interments. 219 
were also a large quantity of the large flue-tiles, used to convey 
heated air from the hypocaust; they are 16in. long, 7in. wide, and 
_ 6in. deep. In the north-east corner was found a fine fibula of 
bronze, here figured,! a quern of millstone grit, pieces of a large 
mortarium studded with coarse grit for bruising corn or flesh, and 
a coin of Constans (small brass) bearing the London mint mark. 
The walls were decorated (in the north-east corner only) with 
‘distemper painting of a dark red colour applied to stucco two to 
_ three inches thick, with which the walls were faced. At the south 
end of the large court were two smaller rooms, each (about) 14 ft. 
_x 12ft.; these seem to have in part escaped the destruction which 
had befallen the rest, and under the wall on the east side were 
many pieces of fine wall painting in perfect condition, the orna- 
“mentation consisting of that usually found in Roman buildings, 
large panels edged with bands of black, red, and yellow, on a deep 
red or orange ground. Some of the pieces are peculiar in having 
a Beashed pattern in imitation of granite, and on some fragments 
aces of foliage are visible. ‘I'he whole was practically as fresh in 
colour as when first applied to the stucco. 
_ On the west side of the small rooms there was a foundation 
running 14ft., forming one side of a small room; unfortunately the 
Test was so confused by constant digging that the size could not 
1! The bronze fibula here figured is of the bow and spring type with T-shaped 
end. It is a fine example, measuring 27in. in length. The broad catch-plate has 
a triangular opening cut in it—a type of make which is not common in Wiltshire. 
General Pitt-Rivers figures two fibule with this opening in the catch, which were 
found in the Romano-British village of Rotherley. None were found in Woodcuts 
village. General Pitt-Rivers suggests that probably the opening may be a 
survival of the original turning back of the nose of the fibula until it became 
fastened to the bow. See “ Racdwghons ¢ in Cranbourne Chase,” vol. II., p. 123. 
Mr. Passmore has lately (1899) obtained a second bronze fibula from the same 
lace. This is formed of one stout piece of bronze wire which is flattened out to 
form the narrow flat bow and catch. The only ornament is a single engraved line 
lown the centre of the bow with short lines carelessly stamped or engraved across 
t. Its length is 2Zin. This type of fibula, in which the bow, spring, and pin, are 
nade all in one piece—the spring being formed of a couple of coils on each side 
f the end of the bow—is thought to follow the earlier type of fibula in use in 
n previous to the Roman Conquest, but they are found not uncommonly with 
man remains. Several are figured in Gen. Pitt-Rivers’ “ Excavations.” [Ep.] 
