Compton Abdale

COMPTON ABDALE

(9 miles N.N.E. of Cirencester)

The well-defined, but undated earthworks of a settlement at Cassey Compton (SP 05201465) are excluded from the Inventory. It has been said that they represent a Roman villa, (fn. 1) but from their situation and form (fn. 2) a mediaeval origin is more likely. For position, see plan, p. 25, s.v. Chedworth.

Note: The Roman Villa (1), lying in a narrow combe, cannot be securely linked to monuments (2), (3) and (4), but it is likely that all four are contemporary. Tiles stamped with similar marks were found in the area of the villa and on a Neolithic long barrow (n on plan), ¼ mile to the S., on Sales Lot in Withington.

(1) Roman Villa (SP 04821624), Compton Grove, excavated by C. E. Key in 1931, is now marked by slight terraces and by abundant surface finds within an area covering ²/3 acre, on the level, narrow floor of a combe facing S.E. Uphill from the area a number of scarps cross the rising valley floor. Two streams flow from the junction of Great Oolite with the Fuller's Earth on which the villa was built. Between the two spring-heads, above a 9-ft. scarp, is a roughly circular depression (b), 18 ft. by 16 ft. and over 3 ft. deep. An irregular platform 10 ft. wide lies just W. of this. Scarps about 3 ft. high crossing the combe head further uphill, and others in arable ground to S., could be the remains of 'Celtic' fields.

The excavations of 1931 took place in the area on the S. side of the combe floor, mainly N.W. of a modern swimming-pool (s) and also, probably, S.E. of it, but not extending to feature (a). The latter appears to be a building platform, at least 60 ft. long by 18 ft. wide, lying adjacent to the steep scarp bounding the combe floor. Finds have recently been made on the surface of this platform, as well as in the probable area of excavation, and again at a point 100 ft. to the N., on the far side of the combe floor.

Notes taken during the excavations indicate a building walled in Oolite, without a corridor. The building, 10 ft. wide and probably longer than the 74 ft. investigated had, according to the excavator, five or six rooms, not all shown on his plan; at least two of them were heated; one hypocaust had stone pillars, the other had tile pilae. Only concrete floors were found.

Finds, mostly in Cheltenham Museum, include 3rd-century coins, samian ware of the late 2nd century and pottery of the 3rd and 4th centuries including colourcoated wares, stone and earthenware roof tiles, much coloured wall-plaster including fragments of designs in panels, flue tiles (one now in Chedworth Museum inscribed VL), a horseshoe, barrel lock and 'frying pan' all of iron, glass, a bronze tooth-pick, a quernstone, and coal.

JRS, XXII (1932), 214. MS. notes provided by Mr. C. E. Key and Dr. J. Liversidge, and plan by Cheltenham Museum; accounts in Cheltenham Chronicle and Gloucester Gazette, August 1931.JRS, XLV (1955), 72 (flue tile), No. 21.

(2) Enclosure (SP 04741616), probably Romano-British, W. of Compton Grove and 50 yds. S.W. of (1) (see adjacent map), occupies ground rising above the narrow combe; only parts of three sides survive as a much-worn bank and ditch. The N.W. angle shows a return E., the area formerly enclosed therefore being not less than ⅓ acre. The boundary of Compton Grove cuts the S. side, and the E. end of the enclosure has been destroyed. Where best preserved, at the N. end of the W. side, the bank is 14 ft. wide and stands about 2 ft. above an external ditch of the same width. The interior slopes 7° E.

R.A.F., VAP CPE/UK 1846: 4067–8.

(3) Building Platform (SP 04921630), probably Romano-British, rectangular, about 51 ft. by 32 ft. (map with (1)), is cut diagonally into the brow of the hill, N.E. of (1). Another platform about 10 ft. across, poorly defined, lies immediately to the E. To the N.E. a substantial artificial scarp up to 5 ft. high and at least 50 yds. long approaches, at its S.E. end, a lesser scarp which follows the contour for 30 yds; the two scarps suggest the bounds of an enclosed area.

Oxfordshire colour-coated ware of the late 3rd or 4th century and fragments of opus signinum were found on the building platform.

(4) Probable Settlement (SP 04951610), Bel Tump, Romano-British but of uncertain nature, lies 170 yds. S.S.E. of (1) and is marked by an irregular heap of large rough limestone blocks on the almost flat narrow tip of an E.-facing spur. The ground falls steeply to the E. and rises gently to the W.

Digging in 1931 yielded 15 coins, including one of Allectus and pottery including samian ware.

MS. notes by Mr. C. E. Key and Dr. J. Liversidge.

(5) Cist Burial (SP 05261746), unaccompanied, probably Romano-British, was found in 1938 immediately N.E. of the outbuildings of Springhill. The grave, about 1½ ft. deep, was completely stone-lined, its cover-stones lying some 2 ft. below the surface.

TBGAS, 61 (1939), 118-19.

(6) Scarps (SP 053167), possibly the remains of 'Celtic' fields broken by strip and later ploughing, occur W. of Smallhope Cottages on a S.-facing slope of up to 12°.