^{page 22 note 1} Gratitude is due to Mr. J. Farrow, who began the excavation and then allowed the Commission to continue it; to Mr. Birch, the new owner, for permission to complete the work; and to Mr. N. H. Field who notified the Salisbury office of the original finds. The excavation was carried out by myself with assistance from colleagues on the Commission at intervals during June-August 1959.

I would like to acknowledge my indebtedness for help in preparing this report to my colleagues, particularly H. C. Bowen, F.S.A., and R. A. H. Farrar, F.S.A.; and in addition to Dr. Grace Simpson, F.S.A. (samian ware); Professor S. Piggott, F.S.A., Sir Cyril Fox, P.P.S.A., and Dr. A. E. Werner, F.S.A. (bronze); Dr. D. B. Harden, F.S.A. (glass); Miss M. M. Howard (skeletal remains); Dr. G. W. Dimbleby (charcoal) and L. V. Grinsell, F.S.A. (barrows). R. F. Jessup, F.S.A., kindly read the typescript.

This account was mainly written in 1960, before the publication (1963) of Barrow C, Lamb Down, Codford St. Mary, Wilts., apparently a fourthcentury Roman barrow with a primary cremation; and before the excavation (1962) of the curious timber structures at West Overton, see W.A.M. lviii (1963), 4.35 and 467, and below, p. 49. The significance of this new evidence for the discussion here is not yet clear.

The following abbreviations have been used:

Dorset Barrows Grinsell, L. V., Dorset Barrows (1959)Google Scholar.

Holborough Jessup, R. F., ‘Excavation of a Roman Barrow at Holborough, Snodland’, Arch. Cant. lxviii (1954), 1–61Google Scholar.

Jessup Jessup, R. F., ‘Barrows and Walled Cemeteries in Roman Britain’, J.B.A.A. xxii (1959), 1–32Google Scholar.

Riseholme Thompson, F. H., ‘The Excavation of a Roman Barrow at Riseholme, near Lincoln’, Antiq. Journ. xxxiv (1954), 28–37CrossRefGoogle Scholar.

Wessex Grinsell, L. V., The Archaeology of Wessex (1958)Google Scholar.

Wessex from the Air Crawford, O. G. S. and Keiller, A., Wessex from the Air (1928)Google Scholar.

^{page 23 note 1} NGR: SU/05220733; O.S. 6″ Dorset XXVI NW., and SU OO NE (new provisional ed.); Interim Rpt. Proc. Dorset N.H. and A. Soc. lxxxi (1959), 99Google Scholar.

^{page 23 note 2} The Horton and Woodlands Tithe Map (1841) names the slope on which the barrow lay as ‘shags Field’ and shows it as arable. The yew tree is not indicated but is, nevertheless, more than 120 years old. There was no evidence on the ground or from the excavation that the barrow had been ploughed.

^{page 23 note 3} Antiquity, i (1927), 431Google Scholar, and x (1936), 38; Wessex from the Air, pp. 16–17; Holborough, p. 1.

^{page 23 note 4} The only possible Romano-British find from Woodlands parish itself is suggested by The Salisbury Journ. 20th July 1778, which reported the discovery in a field at Woodlands of ‘an urn, or rather a vessel … somewhat between an Amphora and an Urceus, containing some coins, confessedly Roman, several annulets, a style, and some small brazen statues, supposed to be Dii penates’.

^{page 23 note 5} Wessex from the Air, p. 59; Dorset Barrows, p. 15; and cf. W.A.M. xlvii (1936), 407Google Scholar.

^{page 29 note 1} Cf. temperature of c. 900^o C. given for Anglo-Saxon cremations, Antiquity, xxxiv (1960), 29–37Google Scholar.

^{page 30 note 1} Cf. Holborough, p. 12.

^{page 30 note 2} Meanwhile, however, the interim note about Knob's Crook has been used in a more general survey by Jessup, R. F.: ‘Roman Barrows in Britain’, Collection Latomus, lviii (1962), 853–67Google Scholar.

^{page 30 note 3} Barrow Codford C and the timber circles at West Overton are excluded from this statement (below, p. 50).

^{page 30 note 4} Holborough, pp. 15–18; Riseholme, p. 28.

^{page 31 note 1} Holborough, p. 1 ; Jessup, p. 1.

^{page 31 note 2} Holborough, p. 52.

^{page 31 note 3} Riseholme, p. 33.

^{page 31 note 4} The mausoleum of Augustus, for example, was built in 28 B.C.

^{page 31 note 5} Antiquity, x (1936), 48Google Scholar.

^{page 31 note 6} Ibid., 46.

^{page 31 note 7} Cf. evidence for deliberate and wholesale breakage of objects in Roman barrows in Belgium, Ann. de la Soc. Arch. de Namur, xxiv (1900), 55Google Scholar; and the apparently ceremonial breaking of amphorae at Holborough, p. 12.

^{page 32 note 1} E.g. Jarlshof, cf. Excavations at Jarlshof, Shetland (1956), Appendix II, pp. 206–10.

^{page 32 note 2} Jessup, pp. 2–4.

^{page 32 note 3} J.R.S. xlii (1952), 99Google Scholar; xliii (1953), 124; xliv (1954), 100; xlv (1955), 141; xlvi (1956), 141; xlvii (1957), 222; and J. W. Brailsford, Hod Hill, i (1962) and I. A. Richmond, ii (forthcoming).

^{page 32 note 4} Proc ‘Som. A. and N.H. Soc. ciii (1958–60), 89; Arch. Journ. cxv (1958), 57Google Scholar; Proc. Dorset N.H. and A. Soc. lxxxii (1960), 88Google Scholar; J.R.S. liii (1963). 147Google Scholar.

^{page 32 note 5} Flints. Hist. Soc. Journ. xiii (1952–3), 5–33Google Scholar; Proc. Som. A. and N.H. Soc. ci–cii (1958), 52–88Google Scholar.

^{page 32 note 6} Arch. Newsletter, i, no. 11 (1949), 15.

^{page 33 note 1} Jessup, pp. 2–4.

^{page 33 note 2} Dorset Barrows, p. 23.

^{page 33 note 3} Antiq. Journ. xxxiii (1953), 14–21Google Scholar.

^{page 33 note 4} Arch. Journ. lxxxvii (1930), 304–9Google Scholar. Very small barrows were, of course, also built in the Bronze Age, e.g. Dorset Barrows, p. 11.

^{page 33 note 5} The only other possibly Early Iron Age barrows in Wessex are also small but are surrounded by square embanked enclosures, cf. three in Dorset Barrows, Winterborne Steepleton 44–46, and R. C. H. M. Dorset, ii, forthcoming; cf. also four small barrows on Kings Weston Hill, Bristol and an enclosed one at Leckhampton, Glos. Proc. Univ. Bristol Spel. Soc. ii (1925), 77–78Google Scholar, 238–42; O'Neil, and Grinsell, , Gloucestershire Barrows (1960), p. 21Google Scholar. See generally, Antiq. Journ. xli (1961), 44–62Google Scholar.

^{page 33 note 6} Riseholme, p. 35.

^{page 33 note 7} Jessup, p. 11.

^{page 33 note 8} Cf. also the similar burial deposit at Radnage, Bucks, . Antiq. Journ. iii (1923), 334–7Google Scholar, and cf. Antiquity, i (1927), 431Google Scholar.

^{page 33 note 9} Proc. Dorset N.H. and A. Soc. lxxix (1957), 119–20Google Scholar. In fact, the cremation in the Marlborough bucket is apparently the only Belgic cremation so far found in Wiltshire, cf. V.C.H. Wilts. 1, i (1957), 85Google Scholar.

^{page 33 note 10} Possibly the inclusion of the ‘Celtic’ fragment represents a continuation of the practice of placing fine metalwork with the dead suggested by the mirror handle found with an inhumation at West Bay, Bridport, dated to the latest pre-Roman or early Roman phase, cf. Proc. Dorset N.H. and A. Soc. lxxvi (1954), 90–94Google Scholar and lxxxi (1959), 106.

^{page 33 note 11} Riseholme, p. 35.

^{page 33 note 12} Arch. Journ. lxxxvii (1930), 286Google Scholar; Antiq. Journ. xv (1935), 55Google Scholar.

^{page 34 note 1} Riseholme, p. 35.

^{page 34 note 2} All the finds have been deposited in the Dorset County Museum, Dorchester, except some bronze fragments retained as specimens in the Research Laboratory, British Museum.

^{page 34 note 3} The exceptions were two worked but unburnt flints included in the original, unprovenanced material from cutting I. An association with the burial pit is unlikely, their probable provenance being the topsoil east of the barrow (fig. 3, section CB). One was a spherical ‘hammerstone’ of cherty flint, diam. c. 6 cm., weight 11 oz., with a rough uneven and much battered surface; the other was a roughly worked blade of good quality flint with an awl-like point at one end and possible secondary working along one edge only, 6·5 cm. long, and 3 cm. wide at most. Small flinty pebbles occur naturally in the subsoil of the area. Many such pebbles, most of them reddened and split by heat, were found, especially in the filling of Pit 1 with the lumps of burnt and semi-burnt clay.

^{page 34 note 4} R. P. Wright, M.A., F.S.A., kindly examined and commented on the graffito on this sherd. It is not unknown for a name-plaque to be buried in a Roman barrow, cf. small gold sheet with inscription from Tirlemont tumulus 1, Musées Royaux d' Art et d'Histoire à Bruxelles, Belgique Ancienne, Catalogue Descriptif III, La Période Romaine (1937), 93, fig. 30. Fig. 35 shows inter alia a small bronze bowl with a base apparently similar to the one listed, p. 36, no. 1.

^{page 34 note 1} Oswald and Pryce, Terra Sigillata (1920), p. 126 and pl. xxi, 2 and 3.

^{page 35 note 2} Knorr, Rottweil (1907), pp. 36–37, Taf. xvi, 1–5 and Rottweil (1912), pp. 34–35, Taf. xvii, 1–12.

^{page 35 note 3} Cf. Oswald, , Index of Figure-Types (1936–7)Google Scholar, pl. LXXXVIII, no. 2389.

^{page 35 note 4} Cf. Déchelette, Les Vases céramiques ornés de la Gaule romaines (1904), no. 1004; Oswald, , Index of Figure-Types (1936–7)Google Scholar, pl. LXXXV, no. 2259.

^{page 35 note 5} May, Trans. Cumb. and West. Arch. Soc. xvii (N.S.) (1917), 124 and pl. 2, 16; Oswald, , Index of Potters' Stamps (1931), p. 268Google Scholar.

^{page 39 note 1} Isings, , ‘Roman Glass from Dated Finds’ in Arch. Traiecta, ii (1957), 72Google Scholar and 73, form 55; Faider, , Études namuroises (1952), p. 76Google Scholar.

^{page 39 note 2} Soc. Ant. Report, xvi (1949), Richborough, iv, 158, pl. LXVII, nos. 367–8, and pl. LXVIII, 370; Antiq. Journ. iii (1923), 334Google Scholar, pl. xxxv.

^{page 41 note 1} Dorset Barrows, p. 23.

^{page 41 note 2} NGR: ST 959030, O.S. 6″ Dorset XXV SW., and ST 90 SE (new ed.). Crawford was a consistent advocate of the Roman date of these barrows, vide Antiquity, i (1927), 346–8Google Scholar; Wessex from the Air, p. 59. He was followed by Collingwood, The Archaeology of Roman Britain (1930), p. 149, and, with reservations, by Dunning, and Jessup, , Antiquity, x (1936), 52Google Scholar and fig. 2. More recently Margary, , Roman Roads in Britain, i (1955), p. 99Google Scholar has begged the question of their date, while Richmond, Roman Britain (1955), p. 145, and the Ordnance Survey, Map of Roman Britain (3rd ed., 1956), p. 31 and fig. 5, have accepted it. Grinsell has several times referred to the possibility that the barrows are Roman, but has consistently pointed to the need for more definite evidence, vide Ancient Burial Mounds of England (2nd ed., 1953), pp. 100 and 158, Wessex, p. 233, and Dorset Barrows, pp. 15, 16, 23, and 130.

These barrows are not to be confused with either ‘the Badbury barrow’ well known for its carved stone discussed by ProfessorPiggott, S., Antiq. fount, xix (1939), 291–9CrossRefGoogle Scholar; or ‘the Shapwick barrow’, the inspiration of the Revd.Wool's, C.The Barrow Diggers (1839)Google Scholar. This last work mentions (p. 62) that the three barrows appear to have been opened but their appearance today does not suggest that this was so, except in the case of barrow N where the secondary scarp at the foot of the mound on the west (fig. 6) could be throw-out from a centrally dug hole. On the other hand, it could be the stub of the road bank starting again on the west of the barrow. If it is, then clearly the road bank post-dates the barrow.

^{page 41 note 3} Dorset Barrows, p. 15, discusses the scanty evidence for Roman barrow burial in the county.

^{page 41 note 4} Wessex from the Air, p. 59, fig. 7 and pl. iv. The area plan, fig. 7, is reproduced in Crawford, Archaeology in the Field (1953), fig. 23, and to avoid confusion, Crawford's lettering of the barrows, L, M, and N, is retained here. Dorset Barrows, p. 130, lists them as Shapwick 12–14. Colt Hoare, Roman Era (1821), facing p. 25, and Sumner, Earthworks of Cranbone Chase (1913), plan iv, show a fourth barrow in line immediately to the north-east of the three here discussed. The area is now too over-grown for certainty but there appears to be a slight mound c. 1 foot high, which may or may not be a small barrow. It is not shown by the O.S.; while Grinsell refers to such a barrow in Ancient Burial Mounds of England (1953), p. 158, but not in Dorset Barrows.

^{page 41 note 5} Contra Crawford, , Antiquity, i (1927), 346Google Scholar.

^{page 42 note 1} e.g. Six Hills, Stevenage, . Archaeologia, liii (1892), 260Google Scholar and references; also Antiquity, x (1936), 48–53Google Scholar.

^{page 42 note 2} Cf. barrow at Fawler, Berks, . Antiquity, I (1927), 347Google Scholar, 479; Wessex from the Air, p. 17.

^{page 42 note 3} Wessex from the Air, p. 59.

^{page 42 note 4} Antiquity, i (1927), 346–7Google Scholar.

^{page 42 note 5} Already raised by Grinsell, Wessex, p. 233, and Dorset Barrows, p. 130.

^{page 42 note 6} Permission to carry out an excavation on a Scheduled Monument was granted by the Ministry of Works Inspectorate of Ancient Monuments and by the Kingston Lacy and Corfe Castle Estates, on whose land the barrows lie. The work took two people three days to complete, including filling in the cutting and exact replacement of the turf in its previous position, and was carried out in July 1960. By the end of the year it was difficult to see where the cutting had been.

^{page 45 note 1} Although now generally agreed to be the Vindocladia of the Antonine Itinerary, Badbury Rings, like Old Sarum, has so far produced little convincing evidence of its importance in Roman times. Here, however, this may be due to lack of systematic excavation. Sundry diggings in the area are known to have produced much R-B pottery, and the somewhat abraded earthwork remains on the south side of the Roman road almost certainly indicate a settlement area.

^{page 45 note 2} The occurrence of two possibly Iron Age ‘A’ sherds low down in the barrow ditch silt is interesting, if inconclusive. They could mean that the barrow is of Iron Age date, but this suggestion can hardly be considered seriously at the moment in view of the scanty evidence for barrow burial in Wessex at that time (above, p. 33).

^{page 45 note 3} If this interpretation is correct, the respect shown by the builders of the Roman road for the barrows … because of their size? … contrasts with their treatment of two disc barrows on Oakley Down, nine miles to the north-east, cf. Wessex from the Air, pl. xxxi; Dorset Barrows, pl. 11 b.

^{page 46 note 1} Colt Hoare's remark, Roman Era (1821), p. 36, that Badbury Rings ‘is rendered a very conspicuous land-mark’ by the clump of fir trees, suggests that it had already been planted for some time.

^{page 47 note 1} Wessex, pp. 233–4. I am indebted to L. V. Grinsell, F.S.A., for information used in Appendixes I and II.

^{page 47 note 2} Wessex from the Air, pp. 16–17.

^{page 47 note 3} A Roman barrow claim was inspired for a mound at Winterbourne, north of Bristol, Glos., by Crawford's definition, largely because of a surrounding bank and a profile allegedly identical to that of the Badbury barrows (Proc. Univ. Bristol Spel. Soc. ii (1925), 297). There is considerable doubt not only about both the Roman date—the surrounding bank is in any case a tree—clump enclosure—but also the claim of the mound to be a barrow at all (O'Neil and Grinsell, Gloucestershire Barrows, p. 137, Winterbourne 1, in T.B.G.A.S. lxxix, pt. i (1960)).

^{page 50 note 1} Antiquity, i (1927), 348Google Scholar.

^{page 50 note 2} Cf. Archaeologia, xliii (1871), 301, note a; and Batty Langley, The Landed Gentleman's Companion (1741), pp. 21, 33–34.

^{page 51 note 1} The prompt publication, while this report was in the press, of Dr. I. Smith's final account and discussion of the West Overton ‘timber tombs’ (above p. 49, no. 15) strengthens in some ways the specific suggestion made above (p. 50) that a regional type of Roman burial mound may prove to be, not large and steep-sided, but low and perhaps small. Certainly the suggestion can now be seriously entertained, additionally supported as it is by Dr. Smith's brief re-assessment of three low barrows, almost certainly of Roman date, dug into by Dean Merewether in 1849 near Avebury. W.A.M. lix (1964), 77Google Scholar.