Two Kentish Hospitals Re-examined: S. Mary, Ospringe, and SS. Stephen and Thomas, New Romney

TWO KENTISH HOSPITALS RE-EXAMINED: S. MARY, OSPRINGE, AND SS. STEPHEN AND THOMAS, NEW ROMNEY* By S. E. RIGOLD, F.S.A. THE hospitals here considered were both products of the great age of medieval hospital-founding, in the late twelfth and early thirteenth centuries. From that age fifteen out of the twenty-five known hospitals in Kent originate, five being earlier and five later. Perhaps more than any other of these (except the unfortunate house of Sweynester in Sittingbourne), these two were unable to adapt themselves to the changing conditions of the fourteenth century, yet they differed widely in purpose and status: Ospringe had royal patronage and performed several of the various functions later assigned to more specialized institutions that have inherited the name of hospital; Romney was of comparatively humble foundation and solely a refuge for lepers. Both houses have already been the subjects of detailed studies in Archceologia Cantiana, since when it has been part of the writer's official duty to investigate their physical remams and he has taken the opportunity to submit the limited, documentary evidence, as it were, to a second pressing. I t is this, rather than any attempt to test the relevance of two such divergent samples to the general problem of the social history of the medieval hospital, that is the reason for considering them together here. THE HOSPITAIL OE ST. MAKY OE OSPRINGE, COMMONLY GALLED MAISON DIET/ The late Charles H. Drake pubUshed a valuable paper on this house in Arch. Cant., xxx (1913), pp. 35-78, followed by a shorter supplementary paper in Arch. Cant., xxxviii (1926), pp. 113-21. He collected an impressive amount of documentary evidence, some of it difficult of access, and gave plans and other material descriptions of the buildings as they were in his day, and particularly at the time (1922) of the rescue and repair of the building now in Guardianship of the Ministry * The Ministry of Public Buildmg and Works contributed to the cost of printing this Paper. 31 TWO KENTISH HOSPITALS RE-EXAMINED of Works. The historical part was ably summarized in V.C.H. Kent, Vol. II, p. 222, by R. C. Fowler. The present writer was most indebted to his predecessor when compiling the official guidebook to this building, and here offers a second supplement, to be read in the light of Drake's papers, incorporating: (i) further documentation that he has collected and suggestions for the interpretation of the whole; (ii) further evidence about the fabric, which has been repaired since it passed into Guardianship in 1947; (iii) the discoveries occasioned by a drainage-trench dug in 1957 across the line of the main range, on the north side of Watling Street. I. HISTORIOA4L EVIDENCE Four questions seem worthy of re-interpretation or amplification: (A) The precise conditions of the foundation of the Hospital. (B) The prolonged and intimate connection between the Hospital of Ospringe and that of St. John without the East Gate of Oxford. (C) The constitution of the Hospital and the appointment of its staff and inmates. (D) The succession of the earlier Wardens. A. The Foundation The claim to a royal foundation need not be taken absolutely at its face value. Henry I I I was fond of taking over and improving existing rehgious foundations and arrogating to himself the Founder's privileges. Netley Abbey had been colonized by donation of Bishop Peter des Roches more than a decade before Henry adopted it,1 and, among hospitals, the very house of St. John at Oxford (see Section B), had had an existence of some forty years before Henry gradually refounded it, commencing with a grant of land in 1231.2 About the same time, beginning with a Charter of Liberties in 1229, the King assumed the patronage and 'foundership' of the Maison Dieu at Dover,3 which Hubert de Burgh had actuaUy founded some ten years earlier and gently reUnquished, in anticipation of his fall. Hubert had also been Lord of Ospringe, which he surrendered with his other honours in August, 1232, having presented the Dover Hospital with the living, which was later transferred to the Ospringe Hospital. He did not recover Ospringe; instead, in 1234, the King gave it in dower to his betrothed Queen 1 Founded July, 1239 (Ann. Waverley), adopted by Henry III, March, 1251. 2 V.CH. Oxford ii, p. 158-9; Close Rolls, 1227-31, p. 600 and 1231-34, pp. 35, 74, etc.; Cal. Pat Rolls, 1292-1301, pp. 101-2. 3 V.C.H. Kent ii, pp. 217-19; Cal. Chart. Rolls, 1226-57, pp. 91 and 141, 191 (re rectory of Ospringe). 32 TWO KENTISH HOSPITALS RE-EXAMINED during the term of his mother's life,4 after a short occupancy by one Joldewin de Doe (Douai?).5 Drake prints in full (Appendix V) a list of private benefactions to the Hospital, confirmed by royal charter in April, 1247,6 but in his commentary he elaborates on one donor, Adam de Tamie of Sheppey, giving an unwarrantedly precise date for the royal foundation, namely 1235, as though this were given in the preamble, or elsewhere in the charter. I have examined the Charter Roll and the date is not there; it is a mere gloss by Daly in his History of Sheppey. The royal donations are in fact numerous in the years 1235 to 1240, there being confirmed by charter, and all are noticed by Drake, as also is the earliest recorded gift,7 in 1234. But the full implication of this earliest benefaction has not been appreciated, since it grants all the surplus corn from the Manor of Ospringe ad emendationem hospitalis, i.e. for the repair of something already in existence. In any case, these months of crisis and rapid change of tenure would hardly seem propitious for a new foundation. The inference is that the Hospital had already stood in embryonic form for some years (Drake's 'very soon after 1230' may well be correct) and that the real founder may have been Hubert, possibly to compensate himself for the loss of the Hospital at Dover. Again, the royal adoption was gradual, commencing in 1234. The Charter of Liberties of 1246 (abbreviated by Drake, op. cit., p. 41), enlarged and confirmed in 1267, was evidently the consummation of the process: it, and no earlier charter, is confirmed by an inspeximus of 1338, issued at the same time as the confirmation of the charter granted to Dover in 1229.8 B. The Link with Oxford Drake records how the errant Master, Nicholas of Staple, was sent to the Oxford Hospital in 1314 and his place as a brother taken by an Oxford man, William of Dewesbury;9 likewise, in 1332, the offending brother Thomas Urre was sent to Oxford.10 But he does not notice the 4 Hasted, 1798, vi, p. 505; it was also granted to Queen Margaret of France and the capital tenement is still called Queen Court; Cal. Chart. Rolls, 1226-57, p. 218; Cal. Pat. Rolls, 1272-81, pp. 348-9 and 1292-1301, p. 453. 5 Not Dol, pace Drake. Joldewin, or Joldan, was a French knight who had somehow forfeited his lands and was granted Wrestlingworth (Beds.) and Piddington (Oxon.) for a term of three years from 1232, to defray his expenses on a crusade (Cal. Pat. Rolls, 1225-32, p. 158); he probably held Ospringe on the same terms and evidently died on active service, as his brother resigned any claim to Ospringe in 1234 (Close Rolls, 1231-34, p. 488 and 1234-37, p. 31). 6 Cal. Chart. Rolls, 1226-67, p. 315; Drake's appendix, no. 5; it is strange that the careful Drake should not have checked Daly's obscure and journalistic little book. 7 Close Rolls, 1231-34, pp. 488, 492. s Cal. Chart. Rolls, 1226-57, pp. 91, 294-5, 1257-1300, p. 70, 1327-41, p. 44. 9 Cal. 01. Rolls, 1313-18, p. 55. 10 Cot. CI. RoUs, 1330-33, p. 551. 3 33 TWO KENTISH HOSPITALS RE-EXAMINED numerous instances under Henry III, when the two hospitals are mentioned in the Rolls simultaneously,11 and under precisely the same conditions. The possible implications of this will be drawn below: meanwhile a brief table of references follows in chronological order: 1. 1234—.Geoffrey, the Royal Almoner, is receiving gifts for both, though named as custos of Ospringe only.12 2. 1237.—Injunction against over-taxation of both, eodem modo.13 3. 1238—Contribution to both, for infirmaries (£10 to Ospringe, £20 to Oxford).14 4. 1238—A chaplain at both, at £2 10s. Od, per annum, for the soul of Wilham de Valence.15 5. 1241—Protection to both, not quite simultaneous.16 6. 1242—25,000(1) poor to be fed at each, at Id. per head.17 7. 1244—A silver cup to each.18 8. 1244—-Fifteen cows to each, from the goods of the vacant See of Winchester.19 9. 1245—A chaplain at both, for William de Valence's widow, on the same terms as her husband's (No. 4).20 10. 1246—Charters of Liberties to each, enrolled in sequence.21 11. 1253—-Three milliaria allecis (brine or salt fish) to both.22 12. 1253—William of Kilkenny custos of both, doubtless temporarily, but concurrently.23 13. 1266—Six oaks to each.24 14. 1266—A robe for the Master of each.25 11 On occasion the house of conversi (converted Jews) in London makes a third. 12 Close Rolls, 1231-34, pp. 394, 488. 13 Close Rolls, 1234-37, p. 569. " Cal. Lib. Rolls, 1226-40, p. 347. 16 Cal. Lib. Rolls, 1226-40, p. 436. 10 Cal. Pat. Rolls, 1232-47, pp. 248, 249. 17 Cal. Lib. Rolls, 1240-45, p. 124. i8 Cal. Lib. Rolls, 1240-46, p. 268. 19 Close Rolls, 1242-47, p. 214. 20 Cal. Lib. Rolls, 1245-51, p. 10. 21 Cal. Chart. Rolls, 1226-57, pp. 294-5; the consolidation of private donations by charter follows closely—ibid, pp. 296-304 (Oxon.), pp. 315-18 (Ospringe). 22 Close Rolls, 1253-54, p. 33. 23 Cal. Pat. Rolls, 1247-58, p. 185. 24 Close RoUs, 1264-68, p. 271. 20 Close Rolls, 1264-68, p. 278. 34 TWO KENTISH HOSPITALS RE-EXAMINED Furthermore, Henry of Wingham, a royal clerk, became vicar of Headcorn, a benefice of the Ospringe brethren, in 1251, and master of the Oxford hospital in 1254.26 Oxford was the senior, both in original foundation and royal adoption, and always the larger house,27 in matters of discipline she behaved much as a mother-house to Ospringe, but 'elder sister' will better symbolize the relationship. The names of the early masters of Ospringe include too many Kentish ones to allow that all the first brethren under the royal dispensation migrated from Oxford, but one or two at least may have formed the cadre. Neither the Chaplain, Adam of Worcester, admitted in 1243,28 nor the unfortunate Henry of Buckingham were local men. Their origins suggest they may have possibly come via Oxford. C. The Constitution The source of the account of the establishment given by Drake (Arch. Cant., xxx, p. 36, note 4) is quoted more fully in V.C.H. It comes from the registers of .Archbishop Warham29 and contains the depositions of two who remembered the last time a proper convent had existed, under Master Robert Darell (1458-70). At that time there were the Master, three professed fellow priests, wearing the habit of the Holy Cross,30 and two secular chantry priests, presumably successors of those appointed, in the first instance, for the souls of the de Valences31 —no mention of the sisters, of whom we hear in the earlier records and who seem to have been active members, not mere alms women. Fortunately, though no early constitution survives for Ospringe, there is one for Oxford,32 which, in the light of the preceding section, may be relevant, allowing perhaps, in the case of Ospringe, for fewer lay members. Again we have a Master (or prior) and three professed brethren (or chaplains), besides six lay brethren and six lay sisters, to attend the poor and infirm. This supports the inference that the Ospringe sisters were attendants, or nurses of a primitive sort, no 20 Cal. Pat. Rolls, 1247-58, p. 121, and Macray, Notes from the Muniments of Magdalen Coll., Oxford, p. 2; he was either the Henry of Wingham who died as Bishop of London in 1262 or, more probably, his namesake who died as Archdeacon of Middlesex in 1269; both apparently began as royal clerks and the Bishop has an article in D.N.B. 27. Compare, for instance, item 3, above; Oxford even acquired a maternity ward in 1240 (Cal. Lib. Rolls, 1226-40, p. 455)! 28 Close Rolls, 1242-47, p. 44. 29 Reg. Abp. Warham, f. 40b (not printed). 30 Reg. Abp. Robert Winchelsey (ed. R. Graham, Cant, and York Soc), p. 79, contains an injunction that they should make their profession after the manner of the Templars and Hospitallers. 31 v.s. section B, items 3 and 9. 32 Cal. Pat. Rolls, 1232-47, p. 38 (Oct., 1234). 35 TWO KENTISH HOSPITALS RE-EXAMINED matter that one was bUnd.33 There were probably lay brethren at Ospringe as weU. By the fifteenth century the lay establishment had evidently lapsed: after the final retrenchment, under the last two masters, only the secular priests remained.34 Though there may not have been an absolute distinction between the working lay members and the 'enforced corrodians'—old royal servants sent to the hospital for maintenance during their retirement, no doubt in much better comfort than the local sick fokk—the latter did not earn their keep and soon became a particular burden on the hospital. Admittedly other houses had the same trouble without the excuse of royal foundation. The first 'enforced corrodian' we hear of, in 1258, is a nephew of a royal waiting-woman, unnamed.35 Possibly she is the same as Juliana, a former maid of Queen Eleanor of Provence, who was herseU already in residence in 1278, when she received logs for her own private fireside.36 This Juliana is quite probably identical with Juliana of Wye, who had recently died in 1307, when her pension was taken over by a man, Robert of Ridware (Staffs.),37 at the King's nomination. If so, she had lived here in comfort for nearly thirty years. I t is tempting to guess that she may even be the same as the Juliana, sister of the Hospital, who was rewarded for gifts of milk and butter (?for ointments) in 1241.38 Did the young sister pass into the Queen's service at the Manor (Queen Court) and then return to the Hospital as a privileged pensioner? Ralph the Beadle, presented in 1292,39 was another of the Queen Mother's men, probably from Queen Court. There were certainly two pensioners at this time, and the two appointed in 1314, one from Bedfordshire and the other, perhaps from Queen Court,40 may replace these, but the number had risen to three if John Toght, recently deceased in 1335,41 is correctly reported as having been presented under Edward I and is not the same as John de Tot, presented in 1314. In 1330 it was agreed not to fill one vacant place.42 33 Helen of Faversham (Close Rolls, 1234-37, p. 48); Oxford had a blind chaplain, William of Faringdon (Close Rolls, 1254-56, p. 44). 34 v.s. note 29; Drake quotes the relevant part of Abp. Warham's visitation, Arch. Cant., xxx, p. 57. 36 Close Rolls, 1256-59, p. 337. 36 Cal. CI. Rolls, 1272-79, p. 445. 37 Cal. Pat. Rolls, 1307-13, p. 9; if it is the same Juliana she was admitted after 1272, i.e. under Edward I. 38 Drake, ibid., p. 39, without exact reference; it does not appear to be enrolled. 39 Cal. CI. Rolls, 1288-96, p. 250. 40 John de Tot, yeoman to Margaret the Queen Mother, possibly a Frenchman —they first thought of retiring him to Evreux (Cal. CI. Rolls, 1313-18, pp. 83, 90) and Henry le Lounge of Flitwick (Beds.) (Cal.Cl. Rolls, 1313-18, p.. 192); another man was retired to Oxford. 41 Cal. CI. Rolls, 1333-37, p. 606; the calendar says Edward I. « Cal. Pat. Rolls, 1327-30, p. 494 (Robert the Messenger, of Newington). 36 TWO KENTISH HOSPITALS RE-EXAMINED D. The Early Wardens The titles of 'Warden' (custos) or 'Master' seem generally to be interchangeable, but the temporary custodes may not have been 'masters'. 'Prior' is not found (compare Oxford, above). The lists given by Drake and by Fowler, including his appendix,43 as far as the mid-fourteenth century (after which they agree completely and I, in turn, have nothing to add) can now be amended and enlarged. Previously published evidence is briefly summarized. 1. GEOFFREY of SUTTON, the King's Almoner, occ. 1234. Clearly a temporary appointment, to put the house in order.44 2. HERVEY of COBHAM, occ. 1235.45 3. WILLIAM GRACYEN, occ. 1237-47.46 4. WILLIAM of KILKENNY, appointed warden of Oxford and Ospringe in 1253. An outside appointment and, it is to be hoped, temporary. He was a royal clerk, archdeacon of Coventry and in December, 1254, elected bishop of Ely. If these two custodies were already being disposed of as life emoluments for a high civil servant, like sinecure prebends, it speaks ill for the King's solicitude for his Hospitals. William would have, in effect, to appoint a deputy. Anyway, he died in September, 1256.47 5. ROGER of LYNSTED, Chaplain and apparently acting master (called proctor) in 1253-55, would have succeeded to the title, at latest, in 1256; certainly resigned in 1263, as his successor was appointed in October of that year; still alive in 1268 when he received a tenement as a pension.48 6. ELLIS (ELIAS) son of HERVEY, appointed 1263, previously a chaplain, i.e. professed brother. Still in office late in 1267.49 43 V.C.H. Kent, ii, p. 242. 44 A Templar, Almoner from 1229, Keeper of the Wardrobe from 1236; an efficient but rapacious official, deposed early in 1240; died soon afterwards, certainly by 1244. See L. E. Tanner, 'Lord High Almoners' in Journal of the British Archaeol. Ass., 3rd ser. xx-xxi (1957-58), pp. 72ff., where his colleague John Lewknor is wrongly named as warden of Ospringe; also Tout, Chapters in the Administrative History of England, I, p. 34. For his relations with Ospringe, Close Rolls, 1231-34, pp. 488, 492. 46 Feet of Fines to 1272 (Kent Records Soc, 1956), p. 123. 48 Close Rolls, 1234-37, p. 493 (1237); Cal. Lib. Rolls, 1240-45, p. 96, F. Fines to 1272 (v.s.), p. 172 (1242); Cal. Pat. Rolls, 1232-47, p. 496 (1247). 47 Cal. Pat. Rolls, 1247-58, p. 185; he is well documented and something of him is entombed beneath a fine Purbeck effigy at Ely and an article in D.N.B. 48 Cal. Lib. Rolls, 1251-60, p. 118 (1253); F. Fines to 1272 (v.s.), p. 257, acting on behalf of Brethren of St. John (1254); Cal. Pat. Rolls, 1247-58, p. 395—-for 'Reynold' read 'Roger'—(1255); Cal. Chart. Rolls, 1257-1300, p. 9, giving lands to the Hospital (1258); Cal. Pat. Rolls, 1258-66, pp. 284, 304 (1263); Cal. Pat Rolls, 1266-72, p. 182 (1268). 49 Cal. Pat. Rolls, 1258-66, pp. 284, 304 (1263); Cal. Chart. Rolls, 1257-1300, p. 70 and Lewis, History of Faversham . . . , p. 81 (1267). 37 TWO KENTISH HOSPITALS RE-EXAMINED 7. JOHN of STAPLE, appointed 1268, previously gate-keeper of Hyde Abbey, but apparently a man of Kent and the first of his family to be associated with the Hospital.50 8. HENRY OF BUCKINGHAM, professed brother in 1262, master, at latest, by 1271 when he was receiving a pardon (repeated in 1290), for trespasses—dilapidations and alienations —committed in that office. If the grant to Roger of Lynsted was the beginning of the rot, he may have become master in 1268. Apparently still misbehaving in August, but deposed by September, 1272.51 9. WALTER of THANET (Taneth), appointed 1272. Occ. 1274- 81, in the latter year with Brother Roger (?R. of Lynsted still active).52 10. PETER, occ. 1287-94.63 11. ALEXANDER of STAPLE, appointed 1295 and only ordained acolyte that year(!). Occ. 1309(?).54 12. NICHOLAS of STAPLE, appointed 1310 (acolyte in 1296); deposed and sent to Oxford, 1314.55 13. HENRY of TEYNHAM, appointed 1314. Died, at latest, 1319.56 14. ADAM of ASH (Esshe), appointed 1319. Died, at latest, 1330.67 15. JOHN of LENHAM, appointed 1330. Died 1349.58 The impression is one of a close community of local men, with more than a hint of nepotism and conspiracy. The complaints about the masters never come from the brethren, and, except in the appointment of WiUiam of KiUtenny, the King shows a pathetic willingness to trust yet another of the already compromised little group. Herein, above all, 50 Cal. Pat. Rolls, 1266-72, pp. 177, 232, 265. 81 Close Rolls, 1261-64, p. 152 (1261) and 1268-72, p. 384 (1271); Cal. Pat. Rolls, 1266-72, pp. 683, 707 (1272); Cal. CI. Rolls, 1288-96, p. 83 (1290). 62 Cal. Pat. Rolls, 1266-72, p. 683 (1272); Feet of Fines, Kent, C.98, file 56, no. 20—not yet printed (1274); Cal. CI. RoUs, 1279-88, p. 119 (1281). e3 Cal. CI. Rolls, 1327-30, p. 480 (refers back to 15 Ed. I, 1286-87); Cal. Pat. Rolls, 1292-1301, p. 117 (1294); Cal. CI. Rolls, 1330-33, p. 496 (posthumous). 64 Reg. Abp. Robert Winchelsey, p. 906; Cal. Pat. Rolls, 1292-1301, p. 148; the reference of 1309, given in V.C.H. cannot be traced. 00 Reg. Abp. Robert Winchelsey, p. 910; Cal. Pat. RoUs, 1307-13, p. 285; Cal. CI. Rolls, 1313-18, p. 55. 08 Cal. CI. Rolls, 1313-18, 55; Cal. Pat. RoUs, 1313-17, p. 105; Cal. CI. Rolls, 1318-23, p. 12; Placit. Abbreviatio. T.R. Ric. I-Ed. II, p. 322 (1316). 07 Cal. CI. Rolls, 1318-23, p. 12 (1319); Cal. Pat. Rolls, 1327-30, pp. 58 (appointment confirmed for new reign, 1327) and 600 (1330). 08 Cal. Pat. Rolls, 1327-31, p. 600 and 1348-50, p. 260. 38 TWO KENTISH HOSPITALS RE-EXAMINED rather than in the economic difficulties of the age, lies the sad failure of the Hospital to maintain its originally ample endowments. The inspeximus of charters, made in 1338 in favour of John of Lenham and. his brethren, makes a point that non-user shall not have rendered any liberties invalid.59 There bad already been a commission to enquire into abuses in 1331,60 the forerunner of many others, including three between 1414 and 1422. II. THE PRESENT STATE OE THE FABRIC Everything above ground of the main complex on the north side of Watling Street has disappeared. The only upstanding relics of the Hospital are the stone walls of two undercrofts on the south side. The name 'leper house' has become attached to one of these subsidiary buildings, but the tradition is most questionable, if only because, in a well ordered semi-monastic plan, lepers should not receive the watercourse before the uninfected.6011 They were, in fact, domestic undercrofts, built either to carry first floor halls, or, as they were later used, to carry solars of ground floor halls. They may have formed part of the residences of the secular priests. Drake published plans and elevations of both undercrofts, and, subsequently, photographs and details of the sixteenth century additions to the western building, which was preserved from destruction in 1922 and converted to a museum, in the care of trustees. When this building (Fig. 1) came into Guardianship of the Ministry of Works (1947) it was temporarily safeguarded, and then thoroughly repaired between 1952 and 1955. The north-east corner of the stone wall, removed in 1894 to accommodate a shop, was restored, using a corner post from a demolished wing of Temple Manor, Strood, and re-setting the original stone door-case, which had fortunately been preserved. Several concealed windows on the upper floor were re-opened. A more detailed analysis of the buildings is now possible. A4. The Eastern Undercroft This remains in private hands. I t probably occupies a plot of land, the conveyance of which was confirmed to the Hospital in 1255.61 The well coursed, knapped flint rubble, and the accurate ashlar are consistent with this date. The door arch has many short voussoirs and a simple chamfer. The moulded corbels for the jetty seem to be original, but not the present upper storey. 69 Cat. Chart. Rolls, 1327-41, p. 444. 89 Cal. Pat. Rolls, 1330-34, p. 207. ooa Xhere is no evidence for the presence of lepers at Ospringe, the idea of a leper-house seems to originate with J. Lewis Hist, and Antiq. of Faversham, p. 81. 81 Cal. Pat. Rolls, 1247-58, p. 393. 39 FIRST FLOOR MAISON DIEU Osprinae G r e a t crotim-post~i Chamber (d) § T f «—*. * o I f e e t 4 metres GROUND FLOOR • Titahar Upright C (3oo 16 th Centuru *i ED c )7oo Restored or .window a t '. R e s e t touer level v Under Croft ;7vw :h$ 1* Fia. I. 40 TWO KENTISH HOSPITALS RE-EXAMINED B. The Western Undercroft This is a deep cellar, floored over at the rear at ground level but open to the first floor at the front. The intact walling is of rougher rubble than in A.—the flint often unbroken, and the ashlar of inferior ragstone. The depressed door arch has long voussoirs and a coarse ovolo within the chamfer. The narrow slit lights have less internal splay than those in the other undercroft, and are grilled and rebated for shutters within. A window, not shown on Drake's plan, opens below street level, between the other two surviving windows, and shows that the flood-line was formerly lower than at present. Everything points to a date not earlier than c. 1300, when the fortunes of the Hospital were already in decline. C. Post-Dissolution Work The rest of the building is entirely subsequent to the dissolution of the Hospital in 1516, but it is not much later. The inserted plaster ceiling indeed dates from the later sixteenth century, to which Drake ascribed the whole, and there are also modifications of c. 1700, but, substantially, the work represents part only of a large house of the early sixteenth century, incorporating the older undercroft. (Plate IB; plans on Fig. 1.) The method of framing may be compared with a doubly jettied block in Canterbury, Nos. 40-44 Burgate, at the corner with the Buttermarket, 62 or, in less sophisticated form, No. 39 Strand Street, Sandwich. It is still allied to the earlier 'open frame' type, having exposed braces and widely spaced studs, but the braces are set very low,63 foreshadowing the small quadrant braces of the late sixteenth century; taken in pairs they form four-centred arches. This fashion is both cheaper and more conservative than the more widely distributed close-studded form, with which it runs concurrently in east Kent, both forms having their derivatives late in the century. This example may be conveniently contrasted with the buildings, including Arden's House, built on the site of the approaches to Faversham Abbey, probably immediately following the dissolution of 1538; these are close-studded and generally more elaborately finished, having heavily moulded fascias to the jetties, but have small four-centred windows like those in the south extension at Ospringe. The moulding on the tie, braces and wall posts spanning the Great Chamber is of usual late Gothic form—a hollow-chamfered fillet flanked by cymas (a more elaborate form would have a roll on the fillet); the 82 The frame recently exposed; on the site of the 'Great Stone House' identified by Dr. Urry from early rent-rolls. 03 Low-pitched tension-braces are laiown in Kent on rather earlier buildings, e.g. no. 124 High Street, Tonbridge. 41 TWO KENTISH HOSPITALS RE-EXAMINED crown-post64 they support is again of a typical late form, with high octagonal bell-base and plinth, a long cavetto above the neck-roll (Fig. 2 and Plate IIA). It is typologically less advanced than the o-'-- ^*W_! •-.•'^ ^ t ' ^^ V ':'-.. 1 1 • • 1 1 ' 1 1 1 I i>' f ;',*/ j^Jii^^ t r r O i z Fecit % o K ', ' Detail of Groir^Ti Vo Regi s / rmoe A t 50 3° F«W to 19 u> JO JH*r«# •Kl ">• u *e °AP /////;//; /z oor „..-*£. o\ «• \ V : Y : > . n B B' '///fates/ AS-//'A .:swv\' "J>p trench ,v ^ - ..A Soil Mtirh Slate i"VY")Tvrt:rTrj'ry "Period II Floor Period I Floor 1 0 1 2 3 4 5 feef /////A Topsail-With .0 ^ sfemefy Pzhhle Base O Wfl lAac\e-xip [/; "Puggy soil Met 2 iii T)olnris Dark, Soil •with. Slate- •res "Litne floor or Base Gravel Floor Graves or other Disturbances Fia. 7. over a 'lens' of small beach pebbles set in the 'pug' and loam that composed the embankment, tapering into a thin layer of pebbles which was traced for at least 3 ft. within the building and 2 ft. externally, a feature which remained even where the solid footings had been robbed. Above the pebbles two clear horizons were detected, both within the wall and without—internafiy some 9 in. in the thickness of clean sandy loam covered the pebbles and were capped by the broken lime floor, about 2 in. thick, found by the previous excavators (M-Bii); externaUy there was up to a foot in thiclaiess of dirty, loamy soil, mixed with the debris of a slate roof, but absolutely free of tiles and covered by a layer containing tile of the harder, later fabric. 55 TWO KENTISH HOSPITALS RE-EXAMINED The loam clearly represented a raising of the floor-level at a time when a tile roof replaced the slate roof. Features clearly belonging to the second phase (e.g. the brick grave) are at the horizon of the upper, lime, floor. Historically, the re-roofing in the harder tile would fit the re-foundation of 1363, but the two pieces of glazed jugs sealed below the slates in the loamy soil would permit an earlier date for the work. The wall-footings, which were unbuttressed and the slate roof would appear to be contemporary. The slate roof was certainly the earliest roof detected on the site. On the analogy of similar roofs at the period,23 it is unfikely to have lasted for two centuries without extensive replacement. This confirms the hypothesis that both roof and waUing date from weU back into the thirteenth century, but hardly into the twelfth. Six feet south of the internal angle the pebble floor was interrupted by a line of mortar sandwiched by lines of dark soil and curving round to a slight nib projecting from the Ume footing of the east wall. This might indicate the beginning of an arcade, perhaps of timber, and/or a parclose screen, 8 ft. within the north wall. If it was aisled on this side only, the nave would then have had the reasonable space of 20 ft. The junction is visible on Plate IIIC. The east wall showed the same section wherever examined, except that the internal pebble spread was absent in the south-east corner. Only the mortar footing remained and that had been robbed in the angle, but it indicated a primary building precisely 32 ft. wide externally. The distribution of the upper (tile) debris already reported (M-Bi) confirmed the same general width for the secondary building. This relatively wide building, as the presence of graves (see below) within it showed, was clearly the chapel in the second period of the hospital. In order to test the possibility that it may only have been the hall, or ward, in the earlier phase, an 8-ft. trench, north-south (x), was opened 6 ft. east of the east wall. This showed no indication of any eastern extension like a structural chancel, as the more usual hospital plan would have had, at least in the twelfth century. This again points to a thirteenth century date for the lay-out. The mortar footing of the primary north wall was traced, with interruptions for over 44 ft., as already found at this lower horizon in 1935. A gap between 37 and 41 ft. from the east was apparently intentional, suggesting an entrance. The south wall was only a little less certain; the pebble spread was found on the right ahgnment up to 27 ft. from the east, and the primary 23 The writer's exoavation at Carisbrooke Castle revealed a succession of five slate roofs between the late twelfth and early fourteenth centuries. Eor the distribution of medieval slate roofs see E. M. Jope and G. C. Dunning, Antiq. Journ., xxxiv, p. 209; the most easterly examples, Stonar, Dover (Arch. Cant., lxix (1955), p. 152) are mid-late thirteenth century and as far as is known the trade declined thereafter. 56 TWO KENTISH HOSPITALS RE-EXAMINED mortar footing was encountered between 25 and 27 ft., the secondary tile debris approximately overlying it. Towards the west end the pebble lens increased in depth up to a foot; apparently the wall had been demolished before the reconstruction, as a lime footing occurred on the right ahgnment but separated from the pebbles by the loam stratum. The west wall was more complex. The pebble spread, presumably primary, was traced irregularly, between 40 and 50 ft. from the east with the latest debris directly over it. This is consistent with a square set west waU just short of 50 ft. from the east and maintained in the second phase. But west-end lay over what looked like earlier footings on a different line. A line of massive, water-worn boulders (M-Cii) set in the clay (Plate IIIA), at the north-west angle and running slightly east of south, was picked up again between 10 and 12 ft. from the angle, and again, with a projecting 'batter', between 16 and 18 ft. These hardly suggest a buttress, as the earlier excavators thought, but could be remains of an earlier structure, on a different alignment, but still on top of the embankment and retained as reinforcement against the slope of the land. This conflicts with the slight indications that the original buildings lay below the embankment, and it may be an uncompleted lay-out. No other sign of it was found. The general picture remains of a unitary but probably aisled chapel, without a structural chancel, about 49 ft. long and exactly 32 ft. wide covering the latter part of the primary phase and the whole of the secondary. It was impossible to strip the area required to discover traces of the suspected aisle-posts. Graves Burials were found immediately outside the north and south walls of the main building, both during the excavation and in subsequent drainage trenching. The only grave found within the church (Fig. 7, I; Plate IIIB) had a brick lining which was traced down for 2 ft. and it was sunk from the level of the later lime floor. On the assumption that the building had a north aisle only, it would have been practically in the centre of the assumed chancel, a position that would be appropriate for the burial of the re-founder (Fraunceys). This grave was not the brick grave found and photographed in 1935, which had a solid capping; in spite of its massiveness, this grave escaped both the trenching and a fairly extensive probing of the eastern half of the building and all subsequent operations; it is possible that it had been robbed for its materials during the war. The bricks, mainly broken, of irregular yellow fabric were probably imported; they are similar in texture to those used on the inserted vault at Homes Place Chapel, Appledore. A headstone cross, described in the next section, was found in the area north of the main building during the building works and has been 57 TWO KENTISH HOSPITALS RE-EXAMINED set up in a wall near its find-spot by Mr Balchin, Clerk to Komney Marsh R.D.C. FINDS (I) BUILDING MATEKLADS, ETC. (a) Slates. Of the 'Cornish' variety (in fact very likely from Devon), usual in such contexts. The standard size was apparently about 4 in. wide, over 7 in. (probably 9 in.) long and relatively thick, with a single tapped hole and some trace of lime torching.24 (b) Rooftiles?i& Three types of plain tile were found: (i) soft red fabric; (ii) similar with the exposed half thinly covered with orange or olive glaze (both (i) and (ii) were found in the lower part of the East Midden and were evidently in use at the same time as the slates); (iii) harder, pale pink-buff (confined to the late horizon but stiU associated with medieval pottery). No tile was complete enough to ensure full dimensions but types (i) and (ii) were of the normal breadth (6J-7 in.) and probably of the normal length (10-11 in.) for tile of similar date. They differ from those found at Joyden's Wood,25 Eynsford and other north Kent sites, in that the holes were closer together (If in. apart). Ridge tiles, as usual in Kent, were glazed but not crested.26 (c) Floor tiles. As previously reported, from the upper floor-level only; thick—up to 1£ in., with yellow or dark green glaze. The largest fragment suggests they were 11 in. square and divided into quarters by an incised cross. (d) Hearth Louver (Fig. 8). A fragment from the East Midden, therefore probably from the northern building. The following note is kindly supplied by Mr. G. C. Dunning. The sherd is of hard grey sandy ware, brownish-red on the inside, the outside glazed dark green. It shows the lower edge of an aperture, with part of a projection on the left side. Enough is present to give the slope and for the diameter just below the aperture to he estimated at about 33 cm. (13 in.). The fragment is part of a large conical ventilator provided with several openings in the side.27 These usually have flanges or baffleplates, projecting outwards, as restored in the drawing. The apertures were probably triangular, as on large pieces of a similar louver from Canterbury.28 It is not possible to say how many apertures there were at this level; there is ample space for four or even for six. 24 But they seem to have been narrow and single-holed, like that shown in Arch. Cant., lxix (1955), p. 153. 24a jror a possible source, the productive file-kilns at Wye, see L. F. Salzman English Industries p. 177. 28 Arch. Cant., lxxii (1958), p. 28, Pig. 3, item 3. 28 Ibid., item 1. The glaze varied from 27 Arch. Journ., cxvi (1959), p. 176, Figs. 16-17. 28 Unpublished; from Prof. S. Erere's excavations. 58 TWO KENTISH HOSPITALS RE-EXAMINED / / / / I 1 1 1 1 1 1 1 1 1 1 j 1 1 I 1 1 1 1 I i FIG. 8. The diameter of this fragment suggests that the louver belonged to Dunning type 1, a separate structure fitted over an opening left in the roof. Louvers of type 2 are made in one piece with the ridge-tile and. in consequence, are smaller in size ;29 this type is not yet represented in Kent. (e) Windowglass. Only three small pieces of painted glass from the final building debris to add to those reported in 1938,30 and of precisely similar character (M-E). No glass whatever from the lower horizons. (f) Headstone Cross (Fig. 9). Of calcareous sandstone, with fairly Fia. 9. 29 'The Pottery Louver from Goosegate, Nottingham,' Trans. Thoroton Soc. of Notts., lxvi (1962), p. 20. 30 Arch. Cant., xlvii (1935), pp. 201-3. 59 TWO KENTISH HOSPITAALS RE-EXAMINED precise horizontal toohng; circular head, 10 in., diameter 4£ in. thick, incised on both sides, with plain open cross extending to edges of circle, l£ in. wide; shaft 5 in. wide with pointed foot; about 2 ft. 4 in. overall length. The design is elementary but would hardly be appropriate after the early thirteenth century: its appearance on this site is valuable as indicating that the type is not in fact earlier than c. 1200. There are similar crosses in the churches of Chislet, Barham (re-used as a bracket), Lyminge (several sizes and varieties), Lympne (three, very hke the present example). (g) Whetstone. A fragment of the usual type of schist whetstone came from the lower deposits of the East Midden.30a (h) Iron Nails. Not well preserved, apparently lath nails with small squarish heads. I I . POTTERY The house was a going concern from the 1190's until at least the 1320's, and again, from the 1360's until some time in the middle of the fifteenth century. The later limits of either phase are uncertain. I t is not certain that the Hospital was completely extinct in the interval: no obvious break is detectable in the pottery sequence or the accumulation of middens but certain types that on other evidence would be assigned to the mid-fourteenth century are rare or absent. For the latter end we know of the appointment of a master in 1458 and the grant of an indulgence in 1451, but most of the documentation concerns the first two generations after 1363. From this and the state of ruination in 1481 we may posit an effective terminus ad quern for everything on the site in the second-third of the fifteenth century. This would be most useful if the pottery were more abundant, since all the other late medieval sites known to the writer in Kent carry their deposits into the sixteenth century. At Romney the fully developed late medieval, ringing, hard wares are absent, though the texture of some of the latest jugs approaches them. Very little of the pottery was firmly stratified. However, beside the smaU quantity (A) sealed beneath the debris of earlier, slate roof of the main buUding, and (B) that just above the floor that overlay the debris of the earlier, slate roof of the northern building, the East Midden deposits can be divided into (C) lower samples, associated with fragments of slate and soft tile roofing and (D) upper samples associated with harder tiles, Uke those from the second roof of the main building. The north-east midden material (E) was unstratified, but fairly uniform in content and similar to (C). The West Midden was not SOa Arch. Cant., lxix (1955), p. 155 for Kentish examples of this commonest type of medieval hone 60 TWO KENTISH HOSPITAALS RE-EXAMINED properly sampled. In view of this uncertainty of association, the principal wares will be treated in one series, but varieties (f) to (i) are practically confined to (D). (a) Pink-buff sandy coarse-wares, often with a greyish core, with some shell-filling and usually a little flint grit; some are exactly paralleled from Potter's Corner near Ashford31 (brighter pastes) or Pivington32 (duller pastes), in late thirteenth century or c. 1300 contexts, and are therefore almost certainly east Wealden wares;33 a third variety, in colour much like the Pivington ware (?made at Egerton), contains pounded chaUc rather than shell and may represent another Wealden pottery nearer the Downs, possibly in the Wye area. Taken together, these wares constitute about 60 per cent, of the material from (E) and about 40 per cent, of (C), and comprise: (i) Cooking pots of various diameters, having flat rims with little or no upper bevel (Fig. 10, nos. 1, 2), as from Pivington34 and Ashford,35 but the inward inclination of the flat rim on no. 1 seems more usual in east Kent.36 A very neat textured pot (Fig. 10, no. 12) has a rim-form suggestive of examples from Eynsford of the mid thirteenth century. (ii) One or two pots with more archaic clubbed outlines and more noticeable chahk filhng (Fig. 10, nos. 3, 4). Rather eroded, these may simply be the predecessors of the flat-rims, but another source, possibly in Sussex37 should be considered. (iii) Bowls: one (Fig. 10, no. 5), in the brighter paste, has a flat rim with pronounced moulding underneath. A later example from (D), in grittier ware, has a slightly concave rim-flange. Both are straight-sided. (iv) Skillet (another was found in 1935), also from (D); the paste contains a good deal of grit but little shell. Tapering, pricked handle, turned down at the tip (not the commoner form with hole for insertion of wooden handle), and concave rim (Fig. 10, no. 6). (b) Grey sandy wares, resembhng those from west Kent and east Surrey, and of finer and darker texture than is usual in the grey wares 31 Arch. Cant., Ixv (1952), pp. 184, 187. 32 Arch. Cant., lxxvii (1962), pp. 38, 40. 33 Such ware is already noted from Brookland: Arch. Cant., Ixv (1952), p. 191. 34 Esp. nos. ii and iii. 30 Esp. Nos. 8, 9, 10, 13. 30 e.g. Tyler Hill, Arch. Cant., Iv (1942), p. 57, no. 10; Dover, Arch. Cant., lxix (1956), p. 157, nos. 5, 6. 37 No. 4, with its triangular rim-section, resembles a form common, in a very different paste, at Pevensey Castle in layers associated with the reconstruction of c. 1240. 61 "-V l\- u V-i K TWO KENTISH HOSPITALS RE-EXAMINED from e.g. Tyler HiU37a or Reculver.37b But since the Romney sherds are predominantly from cooking pots and the, generally more transportable, jugs so common, e.g. at Eynsford, are missing, the source would seem to be nearer than the Kent-Surrey border. They constitute about 15 per cent, of (E) and 30 per cent, of (C). The cooking-pot forms (Fig. 10, nos. 8, 9), flat-rimmed, usually with concentric wheel-marks on the flange, and without upper bevel, generally agree with those found c. 1300 at Joyden's Wood38 and Eynsford. Compare also Group (h). (c) Fine pale buff sandy wares, moderately hard, nearly all with an olive or yellow-green glaze but one or two with a plain white slip. A small jug (Fig. 11, no. 6) is glazed on both sides. All or most are certainly Rye wares;39 one of the embossed rosette patterns (Fig. 11, no. 7) found at the Rye kilns40 occurs, but other varieties found there particularly the elaborate incised designs, are absent. One jug, with wavy comb-marking and a red inner slip (Fig. 11, no. 3) probably belongs to this group. Numerous fragments—this is the commonest jug-ware from the site, constituting about 20 per cent, of (E) and rather more of (C), but hardly anything can be reconstructed. (d) Other fine jug-wares from the earlier deposits: various unplaced sherds, including a rather thick sandy ware with a buff lining and a bronze green glaze, striped with yellow bands over a white slip. The following can be described or identified: (i) A taU non-bulbous41 pitcher from the London area: wall-sherds only, of grey ware with cream shp, speckled green glaze and purpUsh stripes. (ii) Jugs with simple rims in soft, sandy buff paste containing a little chahk, with grey core and pale green glaze without slip. All are associated with the debris of the slate roofs, i.e. context (A) or similar. Fig. 11, no. 2, from the foot of the walls of the main building, has continuous vertical furrows on the handles, a treatment not found on anything published from Kent but known in Wessex and the south Midlands. Fig. 11, nos. 4 and 5 show robust squared rims, pricked handles and bold scoring on neck or rim, a treatment found at Rye and also at Tyler Hill, S7a Arch. Cant. Iv (1942) p. 57; a minority of wares from this site is grey rather than pink-buff. 37b Unpublished, from Mr. B. J. Philp's exoavations. 38 Arch. Cant., lxxiv (1958), p. 18—most examples in Fig. 5, p. 32; the Eynsford material is essentially similar. 30 'Medieval pottery and kilns found at Rye', Sussex Arch. Coll., lxxiv (1933), p. 45. 40 Ibid., PI. iii, p. 48. .An original in Lewes Museum is rather different in colour. 41 Either a baluster, as Antiq. Jour., xii (1961), p. 2, Fig. 1 (from Lesnes Abbey), or the more conical form, as B. Rackham, English Medieval Pottery, PI. 25. 63 1/ h r r ^ f (• ~~»- M^-SM FIG. 11. 64 f . 13 TWO KENTISH HOSPITALS RE-EXAMINED but not on wares of this colour. Seems to be allied to the finer and moderately chalky members of Group (a). (iii) Rim of a large (at least 14 cm. at neck) jug, in pink ware with cream body and thick orange glaze (Fig. 11, no. 1). (iv) One thin sherd of fine cream paste with overall green glaze, from (C), has been identified by Mr. G. C. Dunning as west French, c. 1300. (All the foregoing wares can be assigned to c. 1300, or in some cases to an earlier, but not very early, thirteenth century date, i.e. certainly to the first or leper' phase of the hospital.) (e) Pink-buff sandy wares, probably from the east Weald, as (a) but with shell-filling eliminated: only fifteen sherds including one small, flat-rimmed bowl (Fig. 10. no. 7). A comparable ware was much commoner at Pivington, where it was assigned to the first threequarters and particularly the middle of the fourteenth century, the very period when the hospital was depopulated. This at least reinforces the dating at Pivington. (f) Extremely fine, red, sandy jug-wares; from a probably late fourteenth century context but in the best thirteenth century tradition; origin unknown but probably the same in each case: (i) From above the lime floor in the north building; a tall (baluster?) pitcher; greyish core with a little grit and mica, coppery red slip, green and vermilion glaze; pricked handle, slight inner bead (Fig. 11, no. 9). (ii) Parts of a broad jug in identical ware with overall green glaze and horizontal striations (Fig. 11, no. 8). (iii) From (D), associated with tile, not slate; a broad jug with a carinate rim, similar ware and slip but even finer; orange speckled glaze (Fig. 11, no. 12). (iv) .Another squat jug from same context as (iii); similar but red core; rounded rim; orange glaze on strip of red slip down pricked handle (Fig. 11, no. 10). The thumb-pressed base (Fig. 11, no. 11) would fit, but the core is greyer, though the slip is identical. (g) Pale cream-buff sandy wares, progressively harder. A small (17 cm.) cooking-pot (Fig. 10, no. 14) has a delicate, flat, turned down rim with an inner bead. Many wall-sherds show rilfings or striations. Two bowls, one very large (56 cm.) (Fig. 10, nos. 10, 12), have a pale red slip, traces of external orange glaze and stabbings on the side and on the flat, slightly upturned flange. Quite numerous in (D). The paler may be later Rye wares, but the majority, often with a little flint grit, 5 65 TWO KENTISH HOSPITALS RE-EXAMINED seem to derive from the Wealden style, though not paralleled at Pivington. (h) Later grey sandy wares, in the tradition of (b); a cooking pot fromtheupper level of the East Midden (Fig. 10,no. 11) has a simple rim not unlike the more conservative late medieval pots from Pivington, and schematic ornament; a jug (Fig. 11, no. 13), broad, squat and equally simple in profile, has a deep internal thumb-press at the junction of the handle, a type that has occurred at Whitefriars, Sandwich. (i) Hard, dense jug-wares, approaching the late medieval style; vermilion body (in two cases dark red with a grey lining), or grey body with vermilion lining; red external slip, usually with stripes (as Fig. 12, no 3) or trellis-pattern in white slip; rather metallic glaze, sometimes internal, which appears red or olive-green according to the slip. No profile is reconstructable. Fig. 12, no. 1 is a large jug with normal ver- FT^ r 2 i / i i 4 / / V \ \ Li \ \ \" • ^ \ FIG. 12. milion lining, red core, softer than most, overall deep olive-green glaze and vertical strips, repousse" rather than applied. Fig. 12, no. 5, is a sagging-based vessel, possibly a cistern, glazed internaUy; Fig. 12, no. 4, shows a spread foot, of which no. 2 may be the relatively narrow neck. Most of the sherds seem to have come from large pots. None of the sherds reported in 1935 calls for special re-appraisal. The pottery is consistent with a division into two general categories: 66 TWO KENTISH HOSPITALS RE-EXAMINED (a)-(d) within 40 years on either side of 1300, and (e)-(i) within 40 years on either side of 1400.42 I I I . JETTON OR COUNTER FOUND IN 1935 RECONSIDERED This (Fig. 13) is a relatively early Tournai jetton, diameter 27 mm., pellet stops; the type (obv. sacred monogram ihs, in gothic letter: rev. cross paty, fleurs-de-lys in angles; as always, with crown as initial mark) though not exactly paralleled in Barnard, The Casting Counter and the Counting Board,iS has been noted from Rievaulx44 (two, one on 9 * W FIG. 13. Sc. 3/2. the footing of fifteenth century fireplace) and King John's Palace, Writtle, Essex45 (two from an early fifteenth century level, certainly antedating a late fifteenth century rebuilding), etc. The legends vary: e.g. MISIT.DNS.MANVM.SWM/XPE FILI DEI VIVI MISER [ere nobis] (here and Rievaulx); [IHC] AVTEM.TRANSIENS.P/ MEDIVM.EORVM.IBAT (Writtle). The early or mid fifteenth century dating is confirmed. CONCLUSIONS FROM EXCAVATION (a) The two buildings described in 1935 were rediscovered, lying on the northern part of the southern and higher end of the embanked platform. No other buildings were found. (b) Occupation of the platform and buildings can be traced to the second half of the thirteenth century, i.e. during the functioning of the 42 No exact published parallels for the later material, but see Sussex Arch. Coll., Ixxvi (1935), p. 222 (Bodiam, all after 1386), Fig. 5 (parallel lines of stabbing) and Fig. 4, 29 for a simple-rimmed cooking pot in grey ware (cf. Fig. 10, no. 11). 43 There is a full discussion of this class of jetton ('Le nom de Jesus') in Revue Beige de Numismatique, 1897, p. 185, and PI. IX ff. 44 Found in H.M.o.W. excavations in the 1920's. 45 Unpublished excavations by Mr. P. Rahtz. 67 TWO KENTISH HOSPITAALS RE-EXAMINED leper hospital as such, but the hypothesis stands that the primary occupation may antedate the embanking of the area. (c) The thirteenth century roofs of both buildings were of slate; one at least of the two successive tile roofs (but no significant modification of plan) can be attributed to the fourteenth-century refoundation. (d) A small bank to the west was perhaps the base of the precinct wall. ANALYSIS OF PLAN (a) The main, east-west, building survived through both phases as a single- ceUed structure, probably aisled on the north only, but was re-roofed and re-floored at the foundation. All evidence shows that it was a chapel and never anything but a chapel. (b) The northern, north-south orientated, building likewise had the same position, if not the same plan in both phases. Evidence of occupation suggests that it was a haU for the master, and probably for the other clerks. I t approached, but did not touch, the chapel, on the north side. The north aisle of the chapel was perhaps for the clerks' private devotion; the main body of it for the inmates. (c) These buildings bisected the embanked area; the level space to south of them, in particular, would have been suitable for other structures, though none were detected. In the usual basic plan of hospitals (e.g. St. Mary's, Chichester;46 St. John's, Winchester47), and of monastic infirmaries, there is a chapel and a common hall in series. At Romney there was, in effect, only the eastern half of this arrangement. But in most hospitals there would also have been a hall for the master and staff, and, in certain cases (e.g. Eastbridge, Canterbury48) also for more privileged guests or pilgrims. This was commonly at right angles to the chapel, and in cases (e.g. Kersey, Suffohk), where a hospital with staff following the Augustinian rule was enlarged into a proper priory, it would remain as the west range when a conventual frater and east range were added to complete the claustral plan. A hall of this type is represented by the north building at Romney. In this case a common hall would have been inappropriate both for the later phase, without inmates, and the leprosery. The one known overaU plan of an EngUsh leper-house, St. Mary Magdalene's Winchester, 49 where buildings from the late twelfth century survived until 1788, 40 R. M. Clay, The Medieval Hospitals of England, p. 113 and Fig. 19; W. H. Godfrey, The EngUsh Almshouse, p. 35, Fig. 3 and PI. 2. 47 Godfrey, op. cit., Fig. 17. 48 Ibid., p. 43, Fig. 26. 40 Clay, op. oit., pp. 118-19, Fig. 22 and PI. XXI; and Godfrey, op. cit., Fig. 1, with .some conjectural restoration. 68 TWO KENTISH HOSPITALS RE-EXAMINED shows, within a precinct-wall, an arrangement precisely as at Romney, an aisled, single-cell chapel (originally with a projecting chancel) and a master's hall at right angles to it; there was a storehouse beyond the hall and a row of cells for the inmates lay (the precise position is uncertain) roughly parallel with the precinct wall. Already Lanfrane had stipulated individual timber dwellings for the lepers at St. Nicholas of Harbledown.50 The practice then was to accommodate lepers in cells and at Romney the obvious position for these was around the effective courtyard south of the chapel. They were probably of timber and demohshed without trace at the refoundation. The area north of the Master's hall would doubtless have held a storehouse. 80 Eadmer, Hist. (Rolls Ser.), p. 16. 69