Ancient and Historical Monuments in the County of Gloucester Iron Age and Romano-British Monuments in the Gloucestershire Cotswolds. Originally published by Her Majesty's Stationery Office, London, 1976.
This free content was digitised by double rekeying and sponsored by English Heritage. All rights reserved.
The area covered by monuments (1)–(4) is on Forest Marble, except for a strip of Great Oolite along the parish boundary with Winson. The buildings (1) and the fields (2) are clearly connected, but it is notable that the field lynchets contain grass-tempered pottery, of which no example has yet been found in the area of the house. Undated monuments (3) and (4) are almost certainly Romano-British and are possibly related to (1), as might also be the Romano-British settlement in Winson, ¾ mile to N.W.
(1) Roman Villa, Barnsley Park (SP 08100615), currently being excavated by Dr. Graham Webster (plan, p. 10), is sited on almost flat ground with ample water supplies near by. It consists of a building (a) with a winged 'veranda' facing S.E.; it is nearly 100 ft. long and concealed a 'yard', perhaps roofed, with a bath-house in one corner, a room or rooms at either end and other structures including a channelled hypocaust (h) on the N.W. A barn (b), 57 ft. long and 17 ft. wide internally, is set at right angles to (a) Occupation of the site originated in the 2nd century. Both (a) and (b) lie over gullies and post-holes associated with agricultural activity, and dry-stone walls. The circular walls, up to 3 ft. wide, are dated to c. 280–330. The earliest, but undated masonry building, originally an open shed, stands N.E. of (a). The main building and bath-house, associated with domestic occupation, were erected c. 350–60. Wells (W) 1 and 2, were both dug to a depth of about 25 ft. Within about 20 years of its construction building (a) was reduced to primarily agricultural use. The barn (b), partly coeval with the main building phase, continued in use after 380. Straight dry-stone walls packed with mud and debris, including coins, stand in the courtyard and under the barn, and have the same orientation as the buildings and the 4 acres or more of walled closes which surround them; this is also true of most of the associated fields (2). Some walls are older than the upper paving level of the courtyard. Occupation of the site continued into the 5th century.
The masonry of the building of c. 350–60 is generally of high quality. A wall stone was inscribed FIRMINI, and decorative stone roof finials were found in a well. Pilaster bases and column fragments suggest an elegant veranda. The bath-house had tufa voussoirs. The only certainly heated living room (5) had a channelled hypocaust, later blocked, and a concrete floor. Part of another channelled hypocaust was disclosed at (h) in 1971. Room 29 had a concrete floor coved at its junction with the walls. The barn (b) was built of fine herring-bone masonry. The paved courtyard between (a) and (b) was resurfaced in several places with high-quality pitched stone; the latest surface, dated after 380, shows parallel wheel ruts 4 ft. 8 in. apart. There were also carefully constructed stone paths about 5 ft. wide.
Finds: Over 700 coins have been found, ranging from the third quarter of the 3rd century to the end of the 4th century, the peak period being 330–75. Pottery includes samian, the earliest pieces being Trajanic. An intaglio had a Good Shepherd design, probably Christian. Twenty styli were found and large numbers of agricultural and domestic tools, including two very long scythes of 'Great Chesterford' type and a massive carpenter's chisel. Other finds include quern stones, stone troughs, widespread animal bones and, from the wells, worked wood and leather, and identifiable insects.
(2) 'Celtic' Fields (SP 074064–SP 085058), Barnsley Park, disturbed in various ways, but a rare survival in their situation, cover some 120 acres and occupy most of the N. half of the park surrounding the buildings described in (1). They are linked to and, except at E. and S.E., follow the general alignment of 4 or more acres of dry-stone walled closes around the buildings. An area of pasture with an undated enclosure (c), adjacent to a pond, separates the domestic closes from the fields on the N.W. For the most part the land is almost flat, but it rises gently towards the W. where the remains of a substantial bank and ditch (d) stand at the limit of the more clearly defined fields; lesser traces of field banks extend further W. To the N. the ground drops gently to the natural bounds of a narrow valley, potentially marshy, where the headwaters of a stream in Cadmoor are flanked on the N. by Deadlands Copse. A lesser re-entrant bounds the fields on the S.W. A lateral gulley from Cadmoor lies outside the park wall, towards the E., beyond which the ground is ploughed. Ridge-and-furrow covers much of the S. half of the park and intrudes into the field area at SP 084059.
The remains have been surveyed and selectively excavated by Mr. P. J. Fowler. The soil is a rich red loam, about a foot deep above the broken surface of Forest Marble bedrock. Field boundaries examined consist of narrow earthen banks. Their position has been emphasised by ancient plough action on both sides, leaving either low baulks up to 40 ft. across, or lynchets up to 3½ ft. high. There is no sign of original fences or walls. Abraded pottery in the ploughsoil, thought to be the consequence of manuring with domestic refuse, includes grass-tempered ware probably of the 5th century and later than any finds from (1), as well as sherds which could match pottery from that building.
Original sizes of individual fields cannot in general be determined. The overall pattern is akin to that of prehistoric and Romano-British 'Celtic' fields. Earthwork (d) extends across the greater part of the field area in the W. At the N. it survives as an unploughed ditch, 30 ft. to 60 ft. across and 5 ft. deep below its W. lip; there is little sign of a bank. The ditch is interrupted by a narrow gap at the angle, facing the head of a gentle re-entrant. In its S. parts the earthwork appears as a much-ploughed low ditch, about 25 ft. or more across, with traces of a bank on one side or the other. Beyond this, almost levelled field lynchets extend N., one pair spaced 40 yds. apart. Other features may be contemporary with the fields. In the S.E. a flat-bottomed area (x), some 2 ft. deep, is approached by four linear hollows on alignments comparable with those of the fields; a small elongated mound about 2 ft. high lies centrally in the sunken area. About 1,000 ft. S.E. of (x), a low wide bank extending the line of the lane S.W. from Cadmoor is thought by Dr. Webster to be part of a Roman road which thereafter continues S.W., flanking The Grove plantation.
Some 200 yds. N.N.W. of (1) a small undated enclosure (c) is defined by a slight bank, up to 1½ ft. high above an external ditch; it lies in an area without field boundaries at the head of a shallow lateral reentrant from Cadmoor. Other earth-works in the park are later, or probably later, than the fields. They include certain banks and ditches crossing the fields, and many roughly circular low mounds. These most probably have to do with trees and are omitted from the plan.
(3) ? Enclosure (SP 093053), undated, seen as a cropmark near Poultmoor Farm, 1,500 yds. S.E. of (1), lies on the E. side of a stream flowing along a shallow bottom. There is an entrance in the S.E. side and a broad way appears to pass its N. side.
(4) Probable Settlement (SP 097052), undated, seen as a crop-mark ¼ mile S.E. of (3), lies around Furzey Furlong Barn covering perhaps 2 acres (Plate 55). An entrance occurs towards the N. end of the E. side. The earthwork is associated with other ditches of irregular pattern. The nearly level site falls gently westwards.
(5) Probable Settlement (SP 096049), undated, seen as crop-marks on flat ground 200 yds. S.E. of (4), lies just N. of the Welsh Way and has an entrance in the S.E. side. There are other ditches, one intersecting the enclosure.