^{page 219 note 2} Grid ref. 302190.

^{page 219 note 3} Stenton, F. M., First Hundred Tears of English Feudalism (1932), pp. 194–5Google Scholar, cites it as an example of a twelfth-century castle known from documentary sources but of which no obvious traces remain on the ground at the present day. We first observed it from the train; it is noted in F.C.H. Oxon. ii (1907), 321–2. His other example at Oversley, Warws., has since been traced from the air (Trans. Birmingham Archaeol. Soc. lx (1938), 145, pl. XIII; ibid., lxvii (1951), 17–18). This Ascot Doilly castle should be added to R. A. Brown's ‘List of castles, 1154–1216’ in Eng. Hist. Rev. lxxiv (1959), 261.

^{page 220 note 1} In this work we have received much help. Our first debt is to Mr. Oliver Watney, the landowner, for allowing the excavation, and particularly to Mr. and Mrs. R. A. B. Shaw, the occupants of the manor, for their great kindness throughout this work. The excavation was carried out initially by us, with Mrs. Margaret Jope (who has reported on the fauna), Dr. Eric Gee, and Dr. Hans Hoch. Many residents of Ascot have helped us. In particular Mr. Reginald Edginton has maintained and stimulated interest for many years, and the knowledge of the village archaeology is largely due to his observations. We should also like to thank Mr. Desmond Pratley, Mr. and Mrs. John Sampson, Mr. and Mrs. Morris, and two successive vicars. We should also like to thank for their help and interest Dr. D. B. Harden, the late Mr. B. H. St. J. O'Neil, Mr. G. C. Dunning, Mr. P. S. Spokes, Mr. W. A. Pantin, Mr. Brian Hope-Taylor, Mr. H. M. Colvin, Mr. D. M. Waterman, and Mrs. M. E. Cox for her kindness in completing several pottery drawings. This Society provided a generous grant of £50 and much assistance has been received from the Ashmolean Museum.

^{page 220 note 2} Shipton-under-Wychwood is merely named, Siptone (= sheep-tun), in a charter of A.D. 777 (Birch, Cart. Sax., no. 222).

^{page 220 note 3} Eng. Place-Name Soc. xxv (1956), 109.

^{page 220 note 4} Oxoniensia, xvii–xviii (1953–4).

^{page 220 note 5} Mr. Reginald Edginton has recently observed and partly explored the site of a Roman house on the slope to the north of the River Evenlode, Nat. Grid ref. 296193. This may have been the source of the few fragments of Roman material found in the castle excavations.

^{page 220 note 6} DB, fols. 156b, 158b; V.C.H. Oxon. i, 407, 414.

^{page 220 note 7} This suffix seems to be derived from the le Despensers, earls of Winchester (Eng. Place-Name Soc. Oxfordshire (1954), p. 336).

^{page 220 note 8} A tradition preserved in a thirteenth-century part of the Oseney Cartulary (iv, 1): ‘fratres iurati & per fidem & sacramentum confederati …’. They were jointly associated in the founding of St. George's chapel within the castle of Oxford; Stenton, F. M., in V.C.H. Oxon. i, 383.Google Scholar

^{page 221 note 1} Ascot: Salter, H. E., Facsimiles of Oxford Charters (1929)Google Scholar, no. 58, etc., Kencot: Cart. Oseney, iv (Oxf. Hist. Soc. 1934), 502; Hundred Rolls (Rec. Comm. 1818), ii, 699; cf. also Bampton Aston, DB, fol. 158b, and Cart. Eynsham, i (Oxf. Hist. Soc. 1906–7), 73, n. 3.

^{page 221 note 2} Registrum Antiquissimum (Lincoln Rec. Soc. xxvii (1931)), pp.13–14, no. 6.

^{page 221 note 3} DB, fol. 168A Cart. Eynsham, i, no. 45, ‘… Newentone, quod est de feodo Rogeri de Oili…’ (1148–61); see also ibid., nos. 72, 73.

^{page 221 note 4} F.C.H. Oxon. i, 383, n. 1.

^{page 221 note 5} Nigel was still living in 1115 (Chron. Mon. Abingdon, ii (Rolls Ser. 1887), 63) and the wording of the Pipe Roll 31 Henry I (p. 139 (under the Honour of Wallingford)), ‘Et idem Brienti debet clxvj li. xiij s iiij d pro ministro parte terre Nigeli de Oilli‘, suggests he had not been dead all that long in 1130. For Brientius see also Chron. Mon. Abingdon, ii (Rolls Ser.), 109, 111.

^{page 221 note 6} Ascot d'Oilli is described frequently as ‘Estcote de feodo quod pertinet ad castellum Oxonie’ (Salter, , Facs. Oxf. Charters, no. 58Google Scholar; Cart. St. Frideswide's, iv, no. 1012 = Cart. Oseney, iv, 528; etc.).

^{page 221 note 7} Unless they are reflected in the 1 knight's fee held in Ascot of the bishop of Lincoln, in Inq. P.M. 12 Edward III, no. 154. Mr. H. M. Colvin has discussed the complexities which may lie concealed behind grants of this kind. His main example of Holme Lacy, Herefordshire, has some features in common with this at Ascot (in Essays presented to Rose Graham, ed. A. J. Taylor).

^{page 221 note 8} This is shown by Henry I's confirmation charter to Oseney (Salter, H. E., Facsimiles of Oxford Charters (1929)Google Scholar, no. 58; 1123–33, probably 1127).

^{page 221 note 9} Cart. St. Frideswide's, ii (Oxf. Hist. Soc. 1896), no. 951; see also i, no. 8.

^{page 221 note 10} Cart. St. Frideswide's, ii, no. 1012; Cart. Oseney, iv, 528.

^{page 221 note 11} See table 1 for descent of main line of the d'Oillis.

^{page 223 note 1} Cart. St. Frideszvide's, i, no. 8.

^{page 223 note 2} Feet of Fines Oxon. (Oxf. Rec. Soc. xii (1930)), p. 197; Hundred Rolls, ii, 730–1: ‘Heredes Rogeri de Oyly tenent manor de Estcote cum pert, de Hugone de Plessetis pro ij feodis militum et idem Hugo de domino Regis in capite quod quidem manor Beogo de Clar tenz, de heredibus predictis ad terminum vite reddendo uno cerotecar ad pascham et scutag' qu' cur' et tenz, in eodem manor ij hydas terre que valent per anno x libris.’

^{page 223 note 3} Died 1094, sine hered'; Cart. Oseney, iv, 1, 11. His wife Ealdyth (Alditha) was the daughter of a Saxon, Wigod.

^{page 223 note 4} Dead by 1093–1100.

^{page 223 note 5} Died after 1115, Cart. Oseney, iv, 416 n.; Chron. Abingdon, ii, 74.

^{page 223 note 6} Pipe Roll 31 Henry I; Cart. Oseney, v, 81.

^{page 223 note 7} Cart. Oseney, iv, 11. Edith Forne had been the mistress of king Henry I (see Round, J. H., Geoffrey de Mandeville (1892), p. 94Google Scholar, n. 4).

^{page 223 note 8} Thame Cartulary, i, 3; Cart. Oseney, i, 2; Cart. Eynsham, i, no. 65.

^{page 223 note 9} Thame Cart, i, 2–3; Sandford Cart, i, no. 62; Cart. Oseney, i. 1; iv, 27–28; Henry's reference to Edith as sororis mei in Cart. Eynsham, i, no. 69, is probably a mistake for matris mei.

^{page 223 note 10} Born 1161–2; d. 1232; Cart. Oseney, iv, no. 92, pp. 132–3, makes it clear that the father of Matilda, wife of Maurice de Gant, was still alive in 1219–23; Cart. Oseney, i, xxvi; Fine Roll, sub anno, p. 231. Constable, c. 1220: Eng. Reg. Godstow, ii, 570. See also, however, Round, J. H. in Genealogist, v (1888), 80–81.Google Scholar Henry d'Oilli II describes Guido de Oilli as cognato mei in 1184–1205, Cart. Oseney, v, 210, 237.

^{page 223 note 11} Facs. Oxf. Charters, no. 89; Cart. Oseney, iv, 55, alive 1220.

^{page 223 note 12} Wife of Henry d'Oilli II, 1228–9; Cart. Oseney, vi, 9; d. 1261.

^{page 223 note 13} Facs. Oxf. Charters, no. 87; Cart. Oseney, iv, 19, 55; dead in 1220.

^{page 223 note 14} Henry d'Oilli II never speaks of Margery in his charters, only of Matilda. Through Margery's marriage in 1206 the d'Oilli Honour passed to the earls of Warwick.

^{page 224 note 1} Cart. St. Frideswide's, ii, no. 851.

^{page 224 note 2} He held land of Osbern fitz-Richard in Naunton, Glos.; DB, fol. 168b.

^{page 224 note 3} Chron. Mon. Abingdon, ii (Rolls Ser. 1858), 127.

^{page 224 note 4} They witnessed royal charters together; ibid., pp. 73, 105, 114; Nigel was evidently often with the king in this district, ibid., passim.

^{page 224 note 5} In the Pipe Roll for 31 Henry I (p. 6) he is recorded as having paid 56s. 8d., though for a reason not specified.

^{page 224 note 6} Red Book of the Exchequer (Rolls Ser.), p. 812.

^{page 224 note 7} Round, J. H., Geoffrey de Mandeville (1892), p. 125 n.Google Scholar

^{page 224 note 8} On the position of men with the rank of baron, though not tenants-in-chief, as compared with the great barons, see V.C.H. Oxon. i, 385; Stenton, F. M., First Hundred Years of English Feudalism (1935), p. 90.Google Scholar

^{page 224 note 9} Cart. Eynsham, i, no. 66 (a grant by Robert d'Oilli, d. 1142); nos. 70, 74, 163. For Hawis, see ibid, i, 420.

^{page 224 note 10} Cart. Eynsham, i, nos. 74, 75.

^{page 224 note 11} Cart. Eynsham, i, nos. 72, 73; Hugo held lands in Naunton, Glos. A Hugo d'Oilli was fined 20s. under the Forest Laws in 1176 (Pipe Roll 22 Henry II, 33).

^{page 224 note 12} Cart. Eynsham, i, nos. 165, 166, pp. 421–2.

^{page 224 note 13} Pipe Roll 2 Ric. I, 14.

^{page 224 note 14} Cart. St. Frideswide's, ii, nos. 1017, 1019, 1020; in no. 1022 (c. 1229) the prior and convent of St. Frideswide's agree to find a chaplain to say mass in Roger d'Oilli's chapel at Ascot whenever Roger d'Oilli, his wife, or his family (successores suorum) are present, and this arrangement appears in the Hundred Rolls of 1279 (ii, 731).

^{page 224 note 15} Cart. Oseney, iv, no. 507; for John d'Oilli see also Feet of Fines Oxon., p. 11. It is possible that Roger d'Oilli might have gone abroad with the king.

^{page 224 note 16} Feet of Fines Oxon. (Oxf. Rec. Soc. xii, 1930), p. 197.

^{page 224 note 17} Hundred Rolls, ii, 730: Bogo de Clare died in 1294 (Cal. Pat. R., sub anno).

^{page 225 note 1} Held land in Naunton, Glos., of Richard fitz-Osbern, DB, fol. 168b; probably the same as the constabularius of 1105 (Chron. Mon. Abingdon, ii, 127) and who was evidently at times with Henry I.

^{page 225 note 2} He may have come into his lands in the 1120's (he was paying money for an unspecified purpose as entered in the Pipe Roll of 31 Henry I), and was in possession of Ascot c 1130. He was probably the same as the Roger d'Oilli, sheriff in 1135, who was with the empress at the siege of Winchester in 1141.

^{page 225 note 3} Perhaps the daughter of Roger de Chesney and Aliz de Langetot (Cart. Eynsham, i, nos. 124, 164, and p. 420).

^{page 225 note 4} Probably died a little before 1190, as a Roger d'Oilli paid 20 marcs in 1190 pro terra sua. It must have been Roger C who was the first husband of Aveline; she married secondly Richard Buckerel (Feet of Fines Oxon. 1222, p. 66). She had possessions in Ascot as her widow's dower.

^{page 225 note 5} A monk before c. 1142, see Cart. Eynsham, nos. 66, 70, 72, 74,163; no. 66 is a grant by Robert d'Oilli who died in 1142.

^{page 225 note 6} This is probably the Roger of the 1229 record, though it is not clear what place was occupied by John d'Oilli in the 1190's; perhaps Roger went overseas with the king.

^{page 225 note 7} Cart. Eynsham, i, no. 165.

^{page 225 note 8} Cart. Eynsham, i, nos. 72, 73; he held land in Naunton. A Hugo d'Oilli was fined under the Forest Laws in 1176 (Pipe Roll 22 Henry II, 33).

^{page 225 note 9} He came of age after Michaelmas, 1182 (Salter, H. E., Facs. Early Oxford Charters, no. 86).Google Scholar His name appears regularly in the Pipe Rolls from 28 Henry II onwards, and his confirmation charter of gifts in his fee is witnessed by Robert de Witefield, ‘tunc vice comes’, who was Sheriff in 1182–5.

^{page 225 note 10} Pipe Roll 13 Henry II, 105; ‘In terra Henrici de Oilli que est in dominio Regis, viijs. ivd.’

^{page 225 note 11} Pipe Roll 11 Henry II, 71. The confirmation by Hugo de Plogeneio of a grant by Roger d'Oilli c. 1178 may also reflect wardship during this minority (Cart. Eynsham, no. 165).

^{page 225 note 12} Pipe Rolls 9 Henry II, 50; 10 Henry II, 8.

^{page 226 note 1} See Round, J. H., Family Origins (1930), 217–236Google Scholar, for discussion of the relation between reliefs payable at $c, per knight's fee, and the fines paid on entering into possession of baronies.

^{page 226 note 2} Stenton, F. M., First Hundred Tears of English Feudalism (1932), 220Google Scholar; for discussion of wardship in connection with reliefs and fines, see Round, J. H., Family Origins (1930), esp. 230, n. 52.Google Scholar

^{page 226 note 3} Pipe Roll 13 Henry II, 15.

^{page 226 note 4} Printed in Red Book of the Exchequer (Rolls Ser.), 305.

^{page 226 note 5} Pipe Roll 13 Henry II, 15.

^{page 226 note 6} Pipe Roll 22 Henry II, 31.

^{page 226 note 7} J. H. Round, in Introduction to Pipe Roll 22 Henry II (Pipe Roll Soc, 18); Salzmann, L. F., Henry II (1917), p. 148.Google Scholar

^{page 226 note 8} Pipe Rolls 23 to 30 Henry II, passim, and especially Pipe Roll 31 Henry II, 106.

^{page 226 note 9} Pipe Roll 30 Henry II, 108.

^{page 226 note 10} Pipe Roll 2 Ric. I, 14.

^{page 226 note 11} Ibid.

^{page 226 note 12} Pipe Roll 3–4 Ric. I, 100.

^{page 226 note 13} Pipe Roll 5 Ric. I, 123.

^{page 227 note 1} Cart. St. Frideswide's, ii, no. 951; Cart. Eynsham, i, 73, n. 1.

^{page 227 note 2} Cart. St. Frideswide's, i, no. 8; compare the confirmation by Pope Innocent II, ibid., no. 15.

^{page 227 note 3} Ibid., nos. 1009, 1010.

^{page 227 note 4} Round, J. H., Geoffrey de Mandeville (1892), pp. 328–46.Google Scholar

^{page 227 note 5} e.g., as of Hastings on the Bayeux Tapestry; in the Aberconway charter of c. 1200, ‘usque ad monticulum que in similitudinem castelli apparet’ is a prominent motte-like hill near Llyn Dinas in Nant Gwynant to the south of Snowdon (Archaeol. Cambrensis, xciv (1939), 154). See also Mackenzie, W. M., The Medieval Castle in Scotland (1927), 3, 6.Google Scholar

^{page 227 note 6} Chron. Mon. Melsa, I (Rolls Series), xxiv, 107.

^{page 227 note 7} Cart. St. Frideszeide's, ii, no. 1019.

^{page 227 note 8} Ibid., no. 1022.

^{page 227 note 9} Hundred Rolls, ii, 731.

^{page 228 note 1} Stubbs, W., Select Charters (9th ed., Davis, H. W. C., 1913), p. 180Google Scholar, clause 8.

^{page 228 note 2} Pipe Roll 23 Henry II, 144; 12s. was spent on 100 picks for its demolition. This castle has never been properly described: it has the remains of a keep 41 ft. by 44 ft., of flint masonry with oolite dressings (Bath or Taynton Stone) on the corner clasping and mid-side pilaster buttresses. A large mass of the flint masonry lies where it fell, perhaps during the operations of 1176.

^{page 228 note 3} Salzmann, L. F., Henry II (1917), 168.Google Scholar

^{page 228 note 4} Cart. St. Frideswide's, ii, 251; c. 1229; St. Frideswide's agreed to provide a chaplain whenever Roger d'Oilli, his wife or members of his family should be there. Later in the century (1222) Matilda, widow of Roger d'Oilli, had the capital messuage at Ascot as part of her widow's dowry (Feet of Fines Oxon. (Oxf. Rec. Soc. xii, 1930), p. 66).

^{page 228 note 5} The frequent description from the 1120's onwards of the d'Oilli estate at Ascot as ‘de feodo que pertinet ad castellum Oxonie’ perhaps suggests that the lordship of this Ascot manor carried with it in some way the obligation of castle duty in Oxford Castle.

^{page 228 note 6} This type of service and the relation of a small private castle to its neighbourhood is illustrated in the mid-twelfth century at Weston Turville, Bucks. (Stenton, F. M., First Hundred Tears of English Feudalism (1935), pp. 206–7, 281–2).Google Scholar Here is a fair-sized motte which probably never carried a stone structure; its top seems to have been sliced and levelled in garden improvements. The castle is noted as ‘antequam prosterneretur’ in the Pipe Roll for 1174, p. 82.

^{page 229 note 1} Based on reports by Dr. W. J. Arkell, F.R.S., who visited the site several times during the excavations, and to whose interest we are very greatly indebted.

^{page 229 note 2} Proc. Geol. Assoc. lviii (1947), map opposite p. 108.

^{page 233 note 1} Dr. G. W. Dimbleby kindly reports that it is cut from radially split oak, part of a sizeable log.

^{page 233 note 2} Sections cut through the outer banks at various times have produced pottery closely comparable with that from the tower (e.g. Fig. 11, F i) but none later, which reinforces the view that these banked and ditched paddocks are contemporary with the castle.

^{page 234 note 1} Roy. Comm. Hist. Mons. Oxford City (1939), p. 158, pl. 211, which gives, however, no section, elevation, or plans above ground floor.

^{page 234 note 2} We are most grateful to Mr. J. G. Scott, lecturer in soil mechanics, Queen's University, Belfast, for his advice on the capabilities of this subsoil.

^{page 234 note 3} This paragraph is based on notes by Dr. W. J. Arkell, F.R.S.

^{page 235 note 1} Map in The Oxford Region (ed. A. F. Martin and R. W. Steel, 1954), p.114.

^{page 236 note 1} Oxoniensia, xi-xii (1946–47), 165–7.

^{page 236 note 2} The Times, 13 Nov. 1958. We are grateful to Mr. A. J. Taylor and Dr. M. W. Thompson for showing us the results of the Ministry of Works excavations at Farnham.

^{page 237 note 1} We are greatly indebted to Mr. A. D. Saunders for the information about his excavations conducted for the Ministry of Works at Lydford. Dating of the stone tower at Lydford is difficult; it must be fairly late (perhaps 1195), and the double-splayed windows need not really compel an early date, for similar ones occur in other towers, at at Bridgnorth (probably 1168–9), or Devizes, Sherborne, Kenilworth, and Portchester. In these thick walls they must have been built to give better lighting than single splays.

^{page 237 note 2} Archaeol. Cambrensis, lxxxii (1927), 161 ff.

^{page 237 note 3} Discussed by Mr.Hope-Taylor, Brian in Archaeol. J. cvii (1952), 42.Google Scholar

^{page 237 note 4} Ibid., 27 ff., pl. v.

^{page 237 note 5} Herrnbrodt, A., Der Husterknupp (Cologne, 1958), pp. 178 ff.Google Scholar

^{page 237 note 6} Trans. Devon Assoc. lxxxvi (1954), 236–7.

^{page 237 note 7} Ber. van de Rijksdienst voor het Oudheidkundig Bodemonderzoek in Nederland, iii (1952), 13–14.

^{page 237 note 8} Ulster J. Archaeol. xvii (1954), 103 ff., pls. x, xi.

^{page 237 note 9} Ibid., pp. 164 ff.

^{page 237 note 10} Trans. Dumfries and Galloway Archaeol. Soc. xxix (1952), 169–71.

^{page 237 note 11} Excavated by Mr. J. W. Hunt; unpublished; quoted by Riordain, S. P. Ó in Antiquities of the Irish Countryside (3rd ed., 1953), p. 20Google Scholar, where he points out that the motte-like aspect has been already achieved in pre-Norman times.

^{page 237 note 12} Medieval Archaeology, iii (1959).

^{page 237 note 13} Herrnbrodt, A., Der Husterknupp (1958).Google Scholar

^{page 237 note 14} Benson, G., Later Medieval York (1919), p. 24Google Scholar; Armitage, E. S., Early Norman Castles (1912), pp. 244–5.Google Scholar

^{page 238 note 1} Two or more earthworks of the mound and bailey class can be found scattered within the lands of one vill, as at Clifford, Aston and Ashton, Herefordshire (Roy. Comm. Hist. Mons. Hereford), or West Woodhay, Berks. (Trans. Newbury F.C. vi (1932), 115–26). It is not so easy to find examples of them both sited beside the main nucleus of population: Winkleigh, , Devon, (Trans. Devon Assoc. xxix (1897), 250Google Scholar, 262, 270), East Chelborough, Dorset, (Roy. Comm. Hist. Mons. W. Dorset (1952), p. 90)Google Scholar, and perhaps Knighton, Radnor (Armitage, E. S., Early Norman Castles (1912), p. 293)Google Scholar, may be quoted. The question of siege-castles must also be raised here; the usually quoted examples are Corfe, Dorset (E. S. Armitage, Early Norman Castles, fig. 13), and Pampudding Hill in Oldbury beside Bridgnorth, a most dubious example, for it is a regular motte (Trans. Shropshire Arch. Soc. lii (1948), fig. 6; Armitage, E. S., Early Norman Castles, 34).Google Scholar But an example needs to be excavated; some castles of the Empress Matilda's campaign were more slight affairs (Stenton, F. M., English Feudalism (1932), pp. 201–2).Google Scholar A village enclosure may be seen round two churches at Lee, , Roy. Comm. Hist. Mons. S. Bucks., p. 17.Google Scholar

^{page 238 note 2} See, e.g., Medieval Archaeology, ii (1958), 113, 124–5.

^{page 238 note 3} DB, fols. 156b, 158b.

^{page 238 note 4} Hundred Rolls, ii (Rec. Comm. 1818), 730–1.

^{page 238 note 5} Based largely on material collected by Mr. Reginald Edginton.

^{page 238 note 6} There is part of a jug with broadly frilled base, of hard grey ware with olive glaze, found by Mr. Desmond Pratley on the site 150 yds. east of the church.

^{page 239 note 1} e.g. Ruislip, , Roy. Comm. Hist. Mons. Middlesex, p. 107Google Scholar; Staughton, Great, Roy. Comm. Hist. Mons. Hunts., 252Google Scholar; Thorpe, Castle and Lavendon, , Roy. Comm. Hist. Mons. N. Bucks., pp. 81Google Scholar, 164 (the motte at the latter was destroyed in 1944, yielding much twelfth-century pottery). Expenditure on digging ditches such as those near the Ascot Doilly manor site are recorded in the Pipe Roll of the Bishopric of Winchester, 1207–8 (ed. H. Hall, 1903).

^{page 239 note 2} These were most generously taken for us by Prof. R. J. C. Atkinson in 1946. The rig-and-furrow patterns to the north of the river were first noted by Mr. B. H. St. J. O'Neil in the snow in the winter of 1946.

^{page 239 note 3} Armitage, E. S., Early Norman Castles (1912)Google Scholar, fig. 1; Roy. Comm. Hist. Mons. Middlesex, p. 95; Bull. Barnet Rec. Soc. no. 10, Nov. 1957.

^{page 239 note 4} Carried out by Mr. Reginald Edginton and Mr. and Mrs. John Sampson.

^{page 240 note 1} Cf. Trans. Bristol and Glos. Archaeol. Soc. lxxi (1952), 66.

^{page 240 note 2} Oxoniensia, xxiii (1958), 51, 55.

^{page 240 note 3} Ibid., xiii (1948), 71–72. Type (b ii) may be found outside this area, e.g. Trans. Leicester Archaeol. Soc. xxviii (1952), 34.

^{page 241 note 1} The map, fig. 10, revises that in Oxoniensia, xiii (1948), 71, through a more careful assessment on this basis. It seems to us possible, after a recent re-examination, that some of the vessels found in the excavation of the ‘citadel” in the Herefordshire Beacon should really be considered of the twelfth-century class (R.C.H.M. Herefordshire, iii (1934), p. xlviii, A. 1–3, A. 6).

^{page 241 note 2} A feature widely observed on English pottery of this period, e.g. Oxford (Oxoniensia, xxiii (1958), figs. 11, 12, 16, 17; xiii (1948), 68, no.4); Gloucestershire, (Trans. Bristol and Glos. Archaeol. Soc. lxxi (1952), 62).Google Scholar

^{page 241 note 3} We are exceedingly grateful to Mr. and Mrs. Terence Crowley for the benefit of their practical experience and thought devoted to this matter, which provide confirmation of these conclusions on the extent of wheel-throwing, and on convex base formation.

^{page 242 note 1} Cf. Beere, , Tawton, N., Devon, ; Med. Archaeol. ii (1958), 127–8.Google Scholar

^{page 242 note 2} Cf. Trans. Bristol and Glos. Archaeol. Soc. lxxi (1952), 91, fig. 1. 1.

^{page 242 note 3} e.g. Filkins, , Oxoniensia, xi–xii (1946–1947), 170Google Scholar; or Bourton on the Water, Trans. Bristol and Glos. Archaeol. Soc. lxxi (1952), 91Google Scholar, fig. 1.4.

^{page 242 note 4} Berks. Archaeol. J. 1 (1947), 55–57.

^{page 242 note 5} Oxoniensia, xv (1950), 62.

^{page 242 note 6} Fig. 10; revising Trans. Bristol and Glos. Archaeol. Soc. lxxi (1952), 65–68.

^{page 242 note 7} Trans. Worcs. Archaeol. Soc. xxxii (1955), opp. p. 14.

^{page 242 note 8} Grid ref. 046560, Ash. Mus. unpublished; also an inturned-rim bowl.

^{page 242 note 9} 21; Wilts. Archaeol. Mag. ccvi (1938), 46, fig. 4.17.

^{page 242 note 10} We are grateful to Mr. G. C. Dunning for lending drawings of these Dorset examples.

^{page 242 note 11} 13; Bull. Bd. Celt. Stud, xiv (1952), 14, fig. 6.7, pl. x.

^{page 242 note 12} Shepton Mallet Museum.

^{page 242 note 13} Inf. Group-Capt. G. M. Knocker.

^{page 242 note 14} Brewster, T. C. M., Two Medieval Habitation Sites in the Vale of Pickering (1952), pp. 37Google Scholar, 45; Mr. L. G. Hurst informs us that the type occurs at Wharram Percy.

^{page 242 note 15} P. A. Rahtz, Chew Valley Lake, forthcoming.

^{page 243 note 1} Cf. one from among the twelfth-century pottery pre-dating the Oxford City Wall in New College, Oxoniensia, xvi (1951), 38Google Scholar, fig. 15.16.

^{page 243 note 2} Oxoniensia, xv (1950), 44, 54, fig. 18.11.

^{page 243 note 3} See p. 259; Trans. Bristol and Glos. Archaeol. Soc. lxviii (1949), 30–44; lxxi (1952), 92–94.

^{page 243 note 4} Oxoniensia, viii–ix (1943–4), 104; xvii–xviii (1952–3), 220–1.

^{page 243 note 5} Antiq. J. xix (1939) 303–12.

^{page 243 note 6} Ibid, xxxi (1951), 48–49, no. 14; cf. Oxoniensia, xxiii (1958), 42, 69 (B. 7.2); 58, 71 (Z. 7).

^{page 243 note 7} e.g. eastern England, see Proc. Cambridge Antiq. Soc. I (1957), 53–60; south and west, Hordle, Hants, and Bath, Somerset, large vessels with applied strips, both unpublished.

^{page 244 note 1} Shepard, A. O., Ceramics for Archaeologists (1955), pp. 74–93Google Scholar; transitory subjection to higher temperatures (over 1,000°C.), fleetingly attainable in pit kilns, has possibly occurred with some pieces.

^{page 244 note 2} Cf. Arkell, W. J., Geology of Oxford (1947), pp. 224–5.Google Scholar

^{page 245 note 1} Recognition of this seems implicit in the German term Muschelgruskeramik.

^{page 245 note 2} Proc. Cambridge Antiq. Soc. xlix (1956), 43 ff.

^{page 246 note 1} Recovery of this pottery-making evidence is due to the observations of Mr. Reginald Edginton.

^{page 247 note 1} A few sherds from the tower filling were shown to fit others from the latrine sump.

^{page 257 note 1} Cf. fig. 17, Oxf. 2; Oriel Record, Jan. 1942.

^{page 258 note 1} Oxf. 1; Oxoniensia, xxiii (1958), figs, 19, 21; pl. II.

^{page 258 note 2} Oxoniensia, xvii–xviii (1952–3), 96; Dark Age Britain (ed. D. B. Harden, 1956), 254–6; Proc. Cambridge Antiq. Soc. li (1958), 38 ff.

^{page 258 note 3} Oxoniensia, xxiii (1958), 54–55.

^{page 258 note 4} Antiq. J. xv (1935). 189.

^{page 258 note 5} Oxoniensia, xv (1950), 44 ff.

^{page 258 note 6} Oxoniensia, iv (1939), 98, fig. 22 D, E; cf. also Oxoniensia, xxiii (1959, no. Z. 21).

^{page 258 note 7} Antiq. J. xv (1935), 323–34.

^{page 258 note 8} Cf. Shepard, A. O., Ceramics for Archaeologists (Washington, 1956), p. 189Google Scholar, fig. 18 e.

^{page 259 note 1} Oxoniensia, xv (1950), 50–52; cf. Proc. Cambridge Antiq. Soc. xlix (1956), 54–56.

^{page 259 note 2} Oxoniensia, iv (1939), pl. x. 1.

^{page 259 note 3} Trans. Bristol and Glos. Archaeol. Soc. lxxi (1952), 93; now a little of this type of glaze has come from the early waster dumps at the Brill kilns, Bucks, , Rec. Bucks, xvi (1933–1934), 39–42.Google Scholar

^{page 263 note 1} Oxoniensia, iv (1939), pl. x, nos. 1 and 2; ibid. xv (1950), 50–52.

^{page 265 note 1} For the Germanic and Saxon background of inlaying metals in iron, see Antiq. J. xxxv (1955), 20 ff., pls. v, vi.

^{page 266 note 1} Lond. Mus. Med. Cat. (1940), pp. 135, 142.

^{page 266 note 2} Lond. Mus. Med. Cat. (1940), p. 69, fig. 17; cf. Ulster J. Archaeol. xviii (1955), 99, fig. 11, no. 5.

^{page 266 note 3} Norfolk Archaeology, xxxi (1955), 98, fig. 24.2; cf. also Ulster J. Archaeol. xvii (1954), 141, fig. 13.2. That from Knaresborough is also made in one piece, Antiq. J. xxxiii (1953), 213, fig. 1.23.

^{page 266 note 4} e.g. Lond. Mus. Med. Cat. (1940), p. 277.

^{page 266 note 5} For other early examples of this plain type, see Lond. Mus. Med. Cat. (1940), p. 113, fig. 36.7; Ulster J. Archaeol. xvii (1954), 136, fig. 11.16; xviii (1955), 99, fig. 11.2.

^{page 266 note 6} Antiq. J. xi (1931), 254, pl. XXXVI.

^{page 267 note 1} e.g. Archaeologia, xlvii (1883), 450.

^{page 267 note 2} Ulster J. Archaeol. xvii (1954), 141, fig. 12. 18–21; xviii (1955), 94.

^{page 267 note 3} Excavations in 1947–9, unpublished.

^{page 267 note 4} We are indebted to Mr. H. Roberts for letting us examine this material.

^{page 267 note 5} Pipe Rolls, later twelfth century, passim.

^{page 267 note 6} We are grateful to Mr. G. C. Dunning for bringing to our attention some of these examples.

^{page 268 note 1} Analysis showed this glass to contain 12 per cent, potassium and 3 per cent, sodium (it had probably lost some alkali due to weathering). For technique of spectrographic analysis, see Ahrens, Quantitative Analysis of Silicates. We are most grateful to Dr. Taylor, Geology Laboratory, Oxford, for his generous help in this matter.

^{page 268 note 2} But note Old Sarum, one piece in pit fillings of c. 1100, which might, of course, have been Roman (Antiq. J. xv (1935), 179).

^{page 268 note 3} Cal. Liberate Rolls, passim; references in Turner, and Parker, , Domestic Architecture in the Middle Ages, i (1851), 182–243Google Scholar, and in M. E. Wood, Suppl. to Archaeol. J. cv (1950), 30–31.

^{page 268 note 4} Cal. Lib. R., 1243/4, 1245/6.

^{page 269 note 1} A window is mentioned in 1207–8 in the bishop of Winchester's house at Marwell near Twyford, Hants, but there is no suggestion that it was glazed, nor is there any mention of glass in the expenses connected with the upkeep of buildings throughout these extensive estates at this date. (Pipe Roll of the Bishopric of Winchester, 1207–8 (ed. H. Hall, 1903).)

^{page 269 note 2} See, for instance, Le Couteur, J. D., English Medieval Painted Glass (1926), pp. 58–67.Google Scholar Some twelfth-century painted glass may still be seen in Dorchester Abbey church, Oxfordshire.

^{page 269 note 3} Salzman, L. F., Building in England (1952), pp. 173–4.Google Scholar At Witney, for instance, in 1217, 9d. was spent on ‘linen cloth for the windows of the church’, and linen was similarly being used in St. George's Chapel, Windsor, in 1426–7.

^{page 269 note 4} Garrod, H. W., Ancient Painted Glass in Merton College Chapel (1931), pp. 39–43Google Scholar, to which may be added that Thomas le Verrer was renting a shop in the Drapery (no. 16 Cornmarket) in about 1220 (Sandford Cartulary, i (Oxf. Rec. Soc. xix, 1937), p. 95).

^{page 269 note 5} Cart. Hosp. St. John, i (Oxf. Hist. Soc. lxvi, 1914), pp. 420–1.

^{page 269 note 6} Cart. Hosp. St. John, i, 419; ii, 74; Eng. Reg. Godstow Nunnery (Early English Text Soc), p. 475.

^{page 269 note 7} Cart. Hosp. St. John, i, 408.

^{page 269 note 8} Cart. Oseney, iii (Oxf. Hist. Soc, 1931), p. 105 (1260).

^{page 269 note 9} Winbolt, S. E., Wealden Glass (1933), pp. 7 ff.Google ScholarParker, Eric (The Countryman, xxiv (1941), 134–5)Google Scholar reported the finding of scraps of fused glass with twelfth-century pottery at the Tolt, Hambledon, near Chiddingfold. We have seen this pottery, and while some is twelfth-century, much of it is later. Mr. Brian Hope-Taylor considers that the Tolt may have been a motte (Archaeol. J. cvii (1952), 17–18).

^{page 271 note 1} Compare the buttresses with weathered offset at Sutton-at-Hone, Kent, off. 1234; Archaeol. J., cvii, Suppl. (1950), 40; Archaeol. Cantiana, xxii (1897), 255–60; xlvii (1935), 205–10.

^{page 272 note 1} We are most grateful to Mr. Lawrence Stone for his opinion on this carving: it would have been c. 1200–20 at Wells.

^{page 272 note 2} M.E. Wood in Archaeol. J. cvii, Suppl.(1950), 55, 112 n. 21.

^{page 272 note 3} Ibid., p. 60, pl. iv c.

^{page 272 note 4} Ibid., p. 112 n. 19.

^{page 272 note 5} Ibid., pp. 112–13; Trans. Exeter Diocesan Archit. Soc. xv (1929), pl. xxvi.

^{page 272 note 6} Radford, C. A. R., Guide to Acton Burnell (M. of Wks. 1958).Google Scholar

^{page 272 note 7} Mrs. Kaines-Thomas (Dr. M. E. Wood) has suggested Tyting, near Guildford, as a parallel, and has kindly put her photographs at our disposal here; the building itself was somewhat wantonly destroyed with little protest in 1957, and Mr. John Harvey has generously obtained notes and drawings for us. But though the structural parallel for a window in the gable is valuable, there seems no certainty that this was a chapel either. The Ascot window could just have been a triple lancet, for the crucial evidence is destroyed by the inserted windows. Compare also Martock, fourteenth century (Archaeol. J. cvii, Suppl. (1950), 72–74), and Mr. W. H. Godfrey's arguments concerning Swan-borough, Sussex, (Sussex Arch. Coll. lxxvii (1936), 1–14)Google Scholar, where he considers the chapel probably lay to the east of the hall off. 1200.

^{page 272 note 8} We owe this suggestion to Professor Sir Maurice Powicke.

^{page 272 note 9} SirFox, Cyril and Raglan, Lord, Monmouthshire Houses, ii (1953), 105.Google Scholar

^{page 273 note 1} This type of subdivision is not seen in the Monmouthshire houses nor commonly in the south-west; it is a recognizable type in the Oxford region.

^{page 273 note 2} This term refers to an entrance opening directly on to the butt end of a chimney-stack; Berks. Archaeol. J. lvii (1959), in press. This layout is of wide occurrence in the sixteenth and seventeenth centuries, and can be found also in the New World (Morison, H., Early American Architecture (1952), 21Google Scholar, 54, 61–65, 172, 329).

^{page 273 note 3} Berks. Archaeol. J. lvii (1959).

^{page 273 note 4} SirFox, Cyril and Raglan, Lord, Monmouthshire Houses, iii (1954), 48.Google Scholar

^{page 273 note 5} Oxf. Rec. Soc. xxi (1940), 177.