^{page 11 note 1} See his reports in Antiq. Journ. vii, 438–64 and viii, 461–77. The frequent encouragement given by Mr. Leeds to excavate the site further is a very pleasant recollection.

^{page 11 note 2} National grid reference: SU/511983. O.S. 25 in. edn.: Berkshire Sheet X, 3.

^{page 11 note 3} Antiq. Journ. vii, 439–40; loc. cit. viii, 461. The stratification seems to have been similar to that of the inner ditch.

^{page 11 note 4} See The Oxford Region, ed. Martin, A. F. and Steel, R. W. (1954), pp. 76–81.Google Scholar

^{page 11 note 5} Ordnance Survey photograph (R.A.F. No. 3151) given to the Ashmolean Museum in 1930.

^{page 11 note 6} Using Atkinson's method. Field Archaeology (1953). pp. 31–38.

^{page 11 note 7} Warmest thanks are due to members of the Abingdon branch of the Workers' Educational Association who surveyed and excavated in their spare time. Mention should be made of Mr. D. Smith and Mr. D. Williams for the survey, and Mr. H. Green for excavating.

I am grateful to Miss I. F. Smith for help in preparing this paper, having discussed most of it in detail with her; she has most generously given access to her notes and drawings. I am grateful also to Mr. W. E. V. Young and to Colonel Afonso do Paço for showing me the Windmill Hill and Vila Nova de S. Pedro collections respectively, and for valuable discussions I have had with them.

^{page 13 note 1} The ciphers 10/YR/3/4 moist, etc., denote the colour of a moist sample according to the Munsell system. See Munsell Soil Color Charts (1949).

^{page 13 note 2} Antiq. Journ, viii, 461.

^{page 14 note 1} Red soils have occurred in the Oxford region at sites on the Great Oolite or the derived Summer-town-Radley terrace gravel, widely spaced in date: (1) Pit filling sealed by Groove ware at Gassington, Oxon. (Cornwall, P.P.S. xix, 138). (2) Fillings of Bronze Age barrow ditches, e.g. at Radley, Berks. (Atkinson, Oxon. xvii/xviii, 16). (3) Under the bank of the Belgic Grim's Dyke, Ditchley, Oxon. (Harden, Oxon. ii, 90).

^{page 14 note 2} Antiq. Journ. viii, 466.

^{page 14 note 3} An occupation layer occurred at the base of the equivalent layer in the similarly proportioned section N 3 of the inner ditch. Antiq. Journ. viii, 465–6.

^{page 14 note 4} Ibid. viii, 465, fig. 2c.

^{page 14 note 5} The layout of a causewayed camp may have been somewhat like that of a Zulu village. See Krige, , The Social System of the Zulus (1936), 43.Google Scholar

^{page 14 note 6} The term Abingdon ware is used to denote pottery of Western Neolithic type found at Abingdon and nearby, and has no further implication here. The term slip is simply used to denote a surface coating.

^{page 16 note 1} Antiq. Journ. viii, 467. The industry has recently been discussed by the writer. Oxon. xvii/xviii, 10.

^{page 18 note 1} Ibid. viii, 476.

^{page 18 note 2} Piggott, , Neolithic Cultures (1954), pp. 28, 47.Google Scholar

^{page 18 note 3} Loc. cit., p. 8.

^{page 19 note 1} Sherds from the 1926/7 excavations have been fairly widely dispersed to various museums. The collection in the Ashmolean Museum contains some 280 rims, apparently the largest number of any comparable English series. Cf. Windmill Hill (172), Maiden Castle (100), Whitehawk (93), Hembury (90), Trundle (57). Information kindly given by Mr. W. E. V. Young and Miss I. F. Smith.

^{page 19 note 2} Atkinson, , Piggott, , and Sandars, , Excavations at Dorchester (1951), pp. 108–9. The possible exception is sherd no. 51 of corky ware.Google Scholar

^{page 19 note 3} See p. 23, n. 4. The relative proportions of the wares here and at site I, Dorchester, are all the more striking if it is borne in mind that at Abingdon, generally, shell-gritted rims were more than six times as numerous as those of No. 2 ware, body sherds still more so.

^{page 19 note 4} Ant. xxix, 236–7.

^{page 20 note 1} Some like nos. 6 and 14 have an insignificant inner flange.

^{page 20 note 2} The four inverted rims are not accentuated, and come on the borderline for this classification, No. 13 is classed as heavy.

^{page 20 note 3} Antiq. Journ. viii, pl. lxxiv, fig. 1c, and 473, fig. 4.

^{page 20 note 4} Loc. cit. vii, 454, fig. 8b.

^{page 20 note 5} Fig. 3, no. 9, may have been the rim of such a vessel. For shape, see Antiq. Journ. vii, pl. liii, fig. 2c.

^{page 20 note 6} Liddell (Ant. iii, 289) showed that some of these impressions could have been made with a goose-quill and with the bones of a magpie and a hedgehog. It seems possible to the writer that a wooden tool was used in some cases and a flint with serrated edge in others.

^{page 20 note 7} Organized patterns, like on no. 27, as distinct from haphazard marks.

^{page 20 note 8} Antiq. Journ. viii, pl. LXXIV, fig. 2a.

^{page 20 note 9} And loc. cit., pl. LXXIV fig. 2h.

^{page 20 note 10} Antiq. Journ. vii, 450–1. The friable sand gritted comprise the so-called ruddle pots, small cups with fugitive red slip.

^{page 22 note 1} Antiq. Journ. viii, pl. LXXIV, fig. 2b.

^{page 22 note 2} Ibid, vii, pl. LIII, fig. 2a.

^{page 22 note 3} Loc. cit., fig. 1g.

^{page 22 note 4} Antiq. Journ. viii, pl. LXXIV, fig. 1c.

^{page 22 note 5} Loc. cit., fig. 2a.

^{page 22 note 6} Loc. cit., fig. 2f.

^{page 22 note 7} Antiq. Journ. vii, pl. LIII, fig. 2a.

^{page 23 note 1} Antiq. Journ. vii, 450. (ii) Box of sherds (incl. one T-headed) in Ashmolean Museum labelled, ‘N 2 from very bottom’, (iii) Two rims (cf. nos. 11 and 23, i.e. light rims) marked ‘W. end bottom’, in a box of sherds of both wares otherwise unlabelled. These rims are possibly some of the sherds from the ‘earthy layer’ of N 3 (Antiq. Journ. viii, 464).

Other sherds of shell-gritted and No. 2 ware are marked ‘Trench Upper Layer’. One sherd of No. 2 ware is marked ‘Bottom’ but it appears to have come from a pit and not from the ditch.

^{page 23 note 2} Antiq. Journ. vii, 450.

^{page 23 note 3} Inside the neck appear two badly eroded lines of impressions concentric to the rim. Cf. similarly placed impressions of cord on B Beaker sherd from Clifton Hampden, Oxon. (Oxon. xx, forthcoming). Cf. for shape bowl from Eynsham, Oxon. (ibid.). The sherd is orange throughout, and might conceivably be refired, but could hardly be refired Ebbs-fleet ware (cf. shape: Neo. Cult., p. 309, fig. 49, No. 5). An Abingdon ware type rim sherd (No. 33) also from pit D is orange throughout, matches Beaker ware reasonably (cf. Beaker, B sherds from Long Wittenham, Berks. Ashmolean Museum 1936, 330) and not normal Abingdon ware in paste.Google Scholar

^{page 23 note 4} Flanged bases occur on A or late-looking B Beakers from the region, e.g. Cassington (Oxon. xvi, 3, fig. 2a), Eynsham (Oxon. iii, pl. III, D), Cassington grave 1 (Antiq. Journ. xiv, 271). But possible Rinyo-Clacton (Groove) ware (Neo. Cult., 340, fig. 57, No. 2) or Peterborough ware (Oxon. v, pl. I, E) prototypes can also be suggested. Cf. also Lough Gur Class II ware (P.R.IA. lvi c, 334, fig. 16).

The finds from Pit D (Antiq. Journ. viii, 466,474) are worth enumerating in detail: No. 2 ware: At least 28 sherds, incl. 2 heavy rims (cf. nos. 15 and 28), 5 light rims cf. nos. 21, 24, and 29), and 3 sherds (incl. no. 34) of a flat base. (I doubt whether this base came from a vessel with rim, cf. no. 28 as suggested in Antiq. Journ. viii, 474.) Corky mare: 5 sherds, incl. 4 heavy rims (cf. nos. 2 and 15). Shell-gritted ware: 2 sherds, incl. 1 light rim (cf. no. 29). Beaker ware: see note 3. Antler; 1 fragmentary polishing tool. Flint: 1 leaf-shaped arrowhead (Antiq. Journ. viii, pl. LXXH, fig. 1j), 11 flakes, mostly relatively large, and some with marks of use or dentate edges, and 1 two-edged bifacially retouched flake, which Professor Clark has suggested to me may be a stage in the making of a barbed and tanged arrowhead. It may only be coincidental (i) that the leaf-shaped arrowhead is a variant (loc. cit., p. 467), and (ii) that corky ware also occurred at site I, Dorchester.

^{page 23 note 5} Antiq. Journ. vii, 442–4.

^{page 23 note 6} Ibid. viii, 465.

^{page 23 note 7} See p. 23, n. 1. Finds from other pits include both types of ware, and both came from H 4—the only hearth from which sherds were recorded. Judging by Antiq. Journ. vii, 441, fig. 2, this hearth was at the top of the ‘cultural layer’.

^{page 23 note 8} Keiller, Piggott, and Wallis, P.P.S. vii, 60.

^{page 23 note 9} The matrices of the Peterborough ware sherds and of one of the axe fragments (Antiq. Journ. viii, pl. LXXII, fig. 1p) matched layer 1 of the 1954 excavation best in colour. That of another axe fragment was in colour between layers 1 and 2. These comparatively dark tones might correspond to one of the ‘cultural layers’; on the other hand, the ‘upper filling’ was described as having a ‘black tinge’ (Antiq. Journ. viii, 461).

^{page 24 note 1} Loc. cit., p. 17.

^{page 24 note 2} Thomas, Oxon. xx, forthcoming.

^{page 24 note 3} Loc. cit., p. 17.

^{page 24 note 4} Stone and Wallis, P.P.S. xvii, 119.

^{page 24 note 5} Neo. Cult., pp. 70–72; cf. Arch. J. lxxxviii, 82.

^{page 24 note 6} See Neo. Cult., pp. 373–6, for the content of the Early Middle, and Late Neolithic phases.

^{page 24 note 7} Loc. cit., p. 72.

^{page 24 note 8} P.P.S. vii, 69.

^{page 24 note 9} Neo. Cult., p. 311.

^{page 24 note 10} The outer ditch was 8–9 ft. deep. The mass of shell-gritted ware was not apparently found below 6 ft., but one typical rim sherd was found at 6–7 ft. and another (cf. fig. 4, no. 24) at 7–8 ft. The ditch of Wor Barrow, Handley Down, Dorset, was 12–13 ft. deep, and a little steeper. Pitt-Rivers found that it had silted to a depth of 2½ ft. in the centre in four years (cf. Neo. Cult., 23, fig. 4, bottom and Excavations.…, iv (1898), 24, left).

^{page 25 note 1} Ó Ríordáin, P.R.I.A. lvi, c, p. 327.

^{page 25 note 2} Loc. cit., p. 329 fig. 12.

^{page 25 note 3} e.g. loc. cit., p. 317, fig. 8.

^{page 25 note 4} e.g. loc. cit., p. 318, fig. 9 no. 5.

^{page 25 note 5} e.g. loc. cit., no. 4.

^{page 25 note 6} Loc. cit., p. 327.

^{page 25 note 7} e.g. loc. cit., p. 331, fig. 14, no. 11.

^{page 25 note 8} e.g. loc. cit., no. 3.

^{page 25 note 9} e.g. loc. cit., p. 391, fig. 33, no. 10.

^{page 25 note 10} Loc. cit., p. 331, fig. 14, no. 9. But it occurs in an assemblage of sherds with many Abingdon ware features from the Whiteleaf Barrow, Monks Risborough, Bucks. P.P.S. xx, 223, fig. 5, no. 4.

^{page 25 note 11} Loc. cit., nos. 7 and 12.

^{page 25 note 12} Loc. cit., nos. 16 and 18.

^{page 26 note 1} Loc. cit., p. 373.

^{page 26 note 2} Loc. cit., pl. XXIX.

^{page 26 note 3} Neo. Cult., p. 167.

^{page 26 note 4} Audleystown Cairn: Collins, U.J.A. xvii, 23.

^{page 26 note 5} e.g. Evans, Lyles Hill (1953), p. 38, fig. 13, no. 16.

^{page 26 note 6} Loc. cit., fig. 12, no. 10.

^{page 26 note 7} Loc. cit., no. 8.

^{page 26 note 8} Loc. cit., fig. 11, no. 1.

^{page 26 note 9} Loc. cit., no. 3. By open is meant with shoulder diameter less than rim diameter; by closed is meant the reverse. By upright is implied roughly equal diameters.

^{page 26 note 10} e.g. loc. cit., fig. 13, nos. 14, 40, 41.

^{page 26 note 11} Loc. cit., fig. 14, nos. 46–51.

^{page 26 note 12} Loc. cit., no. 53.

^{page 26 note 13} Loc. cit., no. 52.

^{page 26 note 14} Loc. cit., p. 36. Cf. Abingdon: Antiq. Journ. viii, pl. LXXIV, fig. 2f.

^{page 26 note 15} Ibid. vii, 458. They are considered by Childe, however, to be later than the period in question and to ‘represent a parallel to the Almerian Bronze Age’ (Dawn … (1947), 271.)

^{page 26 note 16} Revista Guimarães, lx, 7, 8.

^{page 26 note 17} Loc. cit., p. 454.

^{page 26 note 18} O Instituto, CXV, iii. Cf. Leisner and Leisner, Antas … de Monsaraz (1951), pl. XXXVIII, no. 3, from an indeterminate dolmen. Since the inventory contained a sherd of a flat-based ‘saucer’, it is pre-sumbably to be dated later than the passage-graves with polygonal chambers, and contemporary with the tholoi and with the Palmela culture of the coast (with Bell Beakers). Cf. also from Cascais near Lisbon: do Paço, As grutas do Poço Velho ou de Cascais (1942), pl. XXVII c. (Comm. dos serviços geológicos de Portugal, xxii).

^{page 26 note 19} The finds from this castro represent different periods and have been mixed by cultivation. Stratigraphical indications were, however, found in excavating the wall (Revista Brotéria, liv, 9).

^{page 27 note 1} The sherd published (O Instituto, CXV, fig. 4, no. 42) seems to have come from a closed type of vessel. Cf. for form Leisner, loc. cit., pl. v, no. 6, from a polygonal passage-grave. This type of bowl seems to have been fairly common at any rate in the Monsaraz region; cf. pl. VII, no. 31, XX, no. 9, xxvii, 29, etc.; cf. also from Cascais, loc. cit., pl. XXVII b; also from Algarve and Almeria in passage-graves, Leisner & Leisner, Die Megalithgräber der Iberischen Halbinsel (1943), pl. 81, 3, no. 19, and pl. XII, 2, no.2 (straight-necked). Open unshouldered bowls also seem to have been common (e.g. Antas…, pl. XXVI); Leisner relates some of the more flaring types to Beaker culture bowls (e.g. loc. cit., pl. xx, no. 1. Cf. from Almeria, Megalithgräber, pl. VII, 2, no. 20). Open bowls with shoulder approximately midway up the body or higher like the Irish and British examples seem to have been rare (cf. from passage-grave, Marcella, Algarve, Megalithgräber, pl. LXXVI, no. 47; another from Cascais was flat-based, loc. cit., pl. XXVIII c.

^{page 27 note 2} Discussed by Leisner, Antas, pp. 81–84.

^{page 27 note 3} Particularly the type with ‘Eifel Tower’ profile. Revista Brotéria, liv, 21, fig. 14, e.g. no. 5.

^{page 27 note 4} Neo. Cult., pp. 72–74.

^{page 27 note 5} Curwen, S.A.C. lxx, 50, pl. VIII, no. 8.

^{page 27 note 6} Loc. cit., p. 53, pl. x. Miss Smith very kindly gave me information about their frequency.

^{page 27 note 7} Loc. cit., pl. IX.

^{page 27 note 8} S.A.C. lxxi, 65, pl. VI, no. 10.

^{page 27 note 9} Briscoe, , Proc. Cambs. Arch. Soc. xlvii, 20, fig.Google Scholar

^{page 27 note 10} S.A.C. lxx, 50, pl. VIII, 2.

^{page 27 note 11} S.A.C. lxxii, 136, pl. XI, 14.

^{page 27 note 12} Ross-Williamson, , S.A.C. lxxi, 71, pl. X, no. 28.Google Scholar

^{page 27 note 13} Loc. cit., p. 70, pl. IX.

^{page 27 note 14} Antiq. Journ. xiv, 113, fig. I4 and 115, fig. 23.

^{page 27 note 15} S.A.C. lxxi, 72, pl. XI, no. 35, etc. They were not so common as at Abingdon. I am grateful to Miss Smith for giving me information as to their frequency. She tells me that some of the T-headed rims may be undecorated Ebbsfleet ware.

^{page 27 note 16} Antiq. Journ. xiv, 113, fig. 6; cf. ibid. viii, pl. LXXII, fig. 2o.

^{page 27 note 17} Ibid, xiv, 118, fig. 41.

^{page 27 note 18} S.A.C. lxxi, 66, pl. VII, 16.

^{page 27 note 19} Loc. cit., p. 71, pl. X, no. 28.

^{page 27 note 20} S.A.C. lxxvii, 76, fig. 2; Antiq. Journ. xiv, 113, fig. 8.

^{page 28 note 1} S.A.C. lxxi, 64, pl. IV, no. 2.

^{page 28 note 2} Loc. cit., p. 64, pl. IV, no. 1.

^{page 28 note 3} Mildenhall, , loc. cit., pp. 20–22, figs. 4 and 6.Google Scholar

^{page 28 note 4} Neo. Cult., pp. 73, fig. 11, no. 4.

^{page 28 note 5} Mildenhall, loc. cit., fig. 4 c.

^{page 28 note 6} I am grateful to Miss Smith for letting me examine a large unpublished collection from Milden hall in the Institute of Archaeology, and for giving me access to her notes on the frequency of types of rim.

^{page 28 note 7} Mildenhall, loc. cit., figs. 3, 4, 6.

^{page 28 note 8} Inst. of Arch. 10th Ann. Rep., p. 31, fig. 2, no. 1.

^{page 28 note 9} Information kindly given by Miss Smith.

^{page 28 note 10} Neo. Cult., pp. 67–70.

^{page 28 note 11} Prehistoric Communities (1949), p. 38.

^{page 28 note 12} Neo. Cult., pp. 97–99.

^{page 29 note 1} The assemblage from the Trundle stands apart, being apparently closer than the others to the primary Neolithic of Windmill Hill.

^{page 29 note 2} Typologically, the comparatively smaller assemblage from Clegyr Boia, Pembroke, should be earlier than Lyles Hill. Here are heavy rims (although no T-heads), and shouldered bowls including open ones. The rare decoration is con fined to transverse hatching on the rim by grooving or finger-nail marks. Corky as well as stone-gritted wares occurred; calcite was used, as at Lough Gur and Vila Nova de S. Pedro. Reddish as well as grey sherds occurred. Williams, Arch. Camb. cii, 34–42.

^{page 29 note 3} A shell-gritted sherd from the Nympsfield Long Barrow, Glos., reported as Neolithic B but of Abingdon ware, has concentric semicircles on the rim. Clifford, P.P.S. iv, 193, fig. 4, no. 20, 205, 211.

^{page 29 note 4} Neo. Cult., pp. 187–8. See also Piggott, , Anthroplogie, lviii, 8–14 for the French and Breton material.Google Scholar

^{page 29 note 5} Neo. Cult., pp. 168–70.

^{page 29 note 6} Atkinson favours such a derivation (Ant. xix, 176). Piggott favours the reverse, and discusses the arguments for and against in Neo. Cult., pp. 116–17.

^{page 29 note 7} Arch. J. lxxxviii, 43.

^{page 29 note 8} Jalhay, and do Paço, , Adas … Sociedad Española de … Prehistoria, XX, pl. XI, nos. 1–5.Google Scholar

^{page 29 note 9} Megalithgräber, 425–8, 432–4, pl. CLXIII. Good matches with British types seem rare in Portugal, but the Portuguese type with concavo-convex long edges and a slight process on one or both edges at the widest point in plan can be matched in northern Ireland; Cascais, pl. XV g; Megalithgräber, 426, fig. 14, no. 4.

^{page 30 note 1} Buick, J.A.A.I., 5th ser., v, 47. There are more than 55 Irish examples in the Ashmolean Museum, probably all from Co. Antrim.

^{page 30 note 2} Loc. cit., fig. 41, no. 12. But the same type was found with a bronze riveted dagger in a grave deposit at Aldbourne, Wilts. W.A.M. xxviii, 263.

^{page 30 note 3} Loc. cit., p. 51, and fig. 20, no. 21.

^{page 30 note 4} e.g. Adas …, pl. IX, no. 5. Cf. Cascais…, pl. XV m.

^{page 30 note 5} Loc. cit., no. 27.

^{page 30 note 6} Avebury Museum.

^{page 30 note 7} Information from Professor J. G. D. Clark, who also kindly examined the fragment from Abingdon, and gave his opinion that it might have been part of a javelin.

^{page 30 note 8} Antiq. Journ. vii, 447, fig. 5 a. Published as a dagger of Beaker Culture type.

^{page 30 note 9} P.D.A.E.S. i, 184, pl. XIX, nos. 754, 936.

^{page 30 note 10} Fragments of greenstone (cf. Great Langdale) were found at Mildenhall, but unfortunately on the surface. Loc. cit. xlvii, 20.