^{page 180 note 1} The Sassetti Chapel (1482—6) by Domenico Ghirlandaio in the church of Ognissanti, Florence, offers a further example. This chapel has no east window at all and is lit entirely from the transept of the church.

^{page 180 note 2} WC, p. 366. It should be noted that ‘choir’ has frequently been used to indicate what Etonians will simply know as ‘chapel’—that is, the entire eastward extension from the antechapel. It is sometimes used in appropriate contexts to mean the emplacement of the stall enclosure.

^{page 181 note 1} James, M. R., The Walpole Society, xvii (1928–9)Google Scholar (afterwards referred to as James), 10, argued that seat canopies were never built against blank walls in the middle ages and that the paintings would have formed a natural background to the stalls. These arguments ignore such a major example as the canopies in the Ulm Minster; and assume that the stalls were always in their present position. James's arguments were to a certain extent coloured by the circumstances in which he wrote.

^{page 181 note 2} The width of the individual scenes varies slightly between 71 and 75 inches (measuring from the centre of one ‘niche’ to the centre of the next). Each would contain either two seats of c. 37 inches each; or three seats of 24 inches. A normal stall would seem to occupy c. 26–c. 30 inches. Those at Eton (nineteenth century) are c. 29 inches in width; those at Norwich (medieval) c. 28 inches. For random foreign medieval comparisons, those at La Chaise Dieu are c. 27 inches, those at Saulieu c. 26–28 inches.

^{page 181 note 3} Compare for instance the fourteenth-century painted decoration of the stalls in Cologne cathedral.

^{page 181 note 4} WC, p. 354.

^{page 181 note 5} Smith, J. T., The Antiquities of Westminster (London, 1807), p. 200Google Scholar. In October 1352 six or seven men were paid for three and a half days' work ‘sawing wood for the finials of the stalls’.

^{page 181 note 6} The text of this contract is printed in WC, p. 597. The reading of the word ‘cowtre’ seems correct although it could be ‘contre’. In neither case is the precise meaning clear. It is possible that the scribe intended ‘centre’, in which case his pen slipped over the crucial vowel (pl. xcvi c). The original document is now catalogued in the Eton Muniments as ECR 38/309.

^{page 181 note 7} WC, p. 447.

^{page 181 note 8} There is no reason to suppose that the geography of the chapel had been rearranged between the Reformation and 1699. See below, p. 186.

^{page 182 note 1} BM. Add. MS. 4843, ff. 144 et seq. See below, p. 183.

^{page 182 note 2} Nevertheless, it is, of course, true, and slightly confusing that the existence of these large flat wall surfaces resulted from Henry's original plan to build stalls with canopies at this point.

^{page 182 note 3} James (p. 14) agreed that a screen stood at this point, but proposed an east choir screen similar to that which survives in St. David's cathedral.

^{page 183 note 1} It is hoped that this manuscript will re-emerge, It seems likely that it is somewhere in the Library at Eton, but in spite of a great deal of patient searching by the Librarian, Mr. Strong, it has not so far been discovered.

^{page 186 note 1} Since so many have been destroyed, it has only been possible to guess at the sizes of the stones. A few survive in the ante-chapel as follows. (The numbering refers to the table, fig. 1)

There are many other stones with indents for brasses, now lost. Richard Allestree's mural monument survives in the vestibule of the north choir entrance. No stones or monuments were taken to the new (nineteenth-century) parish church.

^{page 186 note 2} For the following see WC, pp. 380 ff.; and Brown, R. Allen and others The History of King's Works (London, 1963), i, 279 ffGoogle Scholar.

^{page 186 note 3} For the windows, see WC, p. 405. The large flat areas on which the paintings are executed must date from this early period; for they must almost certainly belong to a time when it was still intended to place the stalls at this point. See also p. 182, n. 2, above.

^{page 186 note 4} There have been three major changes of the floor level in the chapel since the Reformation. The first occurred in 1699–1703 when almost the entire floor was levelled off at a height corresponding roughly to the present sanctuary. This height was reached by five steps west of the eighteenth-century organ screen. The second change came c. 1847 when the chapel was restored, as it was thought, to its original appearance. The result may be seen in the illustration of WC (fig. 36 opp. page 450) where the high altar at its present level was reached by a steep flight of six steps adjacent to Lupton's chapel. More changes occurred in 1863—of what nature is not clear. Finally in 1905 as part of a South African War Memorial, a new restoration was undertaken—again, as was thought, reconstructing the original floor levels.

^{page 187 note 1} The equivalent mouldings in the south side of the third and fourth bays (see pi. xcv a) measure 8½ inches in height. Here in the first bay the moulding measures 5½ inches.

^{page 187 note 2} The north side is occupied by Lupton's chapel, the south by the nineteenth-century monument to E. C. Hawtrey († 1862).

^{page 188 note 1} Now catalogued as ECR 39/81.

^{page 188 note 2} WC, p. 366.

^{page 188 note 3} The steps are now separated. One lies to the east of the stalls, the other lies at the choir entrance. There seems no reason to suppose that the medieval level of the ante-chapel has ever been seriously altered.

^{page 189 note 1} WC, p. 354.

^{page 189 note 2} This figure is derived from the subsidy assessments of 1523 which are fortunately printed for Buckinghamshire (see Bucks Record Society, vol. 8,1944, The Subsidy Roll for the County ofBuckingham Anno 1524, edited by A. C. Chibnall and A. Vere Woodman). At that date there were sixty-six taxable inhabitants in the ‘Tonne of Eton’ (p. 18). It is unlikely that every taxpayer represented a household; against this, however, there were an incalshire culable number of very poor necessarily omitted. An allowance of four persons for every name on the list produces a population of 264.

^{page 190 note 1} Woods's order is of interest here since it was clearly the result of an orderly tour of the chapel. He dealt first with the north-eastern and then with the south-eastern tombs. This is followed by monuments in the second bay; and these are followed by Westbury. He then went back to a group under the third window on the south (perhaps they had been covered by benches which had to be moved). Finally he moved down to Henry Bost at the entrance to the choir and so out into the ‘nave’.

^{page 190 note 2} It is of interest that although the south wall of the third bay is liberally sprinkled with carved names (seventeenth century), the fourth bay is entirely bare. This is consistent with the idea that medieval stalls were built against the wall at this point.

^{page 190 note 3} ES 10. The parish clerk appears in the audit for 1484–5 (EA 20).

^{page 190 note 4} ES 31.

^{page 190 note 5} It is not intended to discuss the layout and dedications of the Eton altars at this point. Some information will be found in H. C. Maxwell Lyte, History of Eton College, ch. 5. Maxwell Lyte thought that these two nave altars were dedicated to Sts. Peter and Nicholas. This was based partly on a process of elimination and partly on a will of 1517 (EReg i (1457–1536), p. 136). This conclusion may not hold good for the 1480s; the documents are inconclusive but could indicate that these two altars were dedicated to the Virgin (north) and St. Nicholas (south).

^{page 190 note 6} ES 37.

^{page 191 note 1} WC, p. 354.

^{page 191 note 2} Here the main problem was the impossibilty of forecasting the future numbers of paying commensals.

^{page 191 note 3} Bond, Francis, Screens and Galleries (London, 1908), pp. 40–2Google Scholar, prints two contracts from Cambridge (Old St. Mary's 1518–23) and Stratton, Cornwall (1531), in which this happens.

^{page 191 note 4} See, for instance, Llanegryn, Merioneth (fifteenth century) illustrated in A. Vallance, Roodscreens and Roodlofts, pls. 221 and 224.

^{page 191 note 5} See EAB i (1506–29), 1506–7 under Ecclesia, p. 40: ‘pro factura altaris in pulpito et unius sedis ibidem cum ambone’ carpenter paid for 38 days' work. Also 1510–11 under Ecclesia, p. 107: ‘pro apparatu eneo et ferreo bibliae et librorum a magistro dale in pulpitum collatorum et firmandorum iiijd’.

^{page 191 note 6} The problem of the organ's medieval position is complicated by the fact that in 1699 it was nowhere near the site of the original screen but against the south wall of the third bay from the east (James, p. 5, thought its position was against the north wall further west). The building of that particular organ is amply documented in the Eton Audit for 1613–14 (EAB viii). Reference to this year, under Templum, shows that at the same time a loft with stairs was made for it (‘Item to Goodman dale for iii men iii daies and ii men ii daies in making the staires and loft for the organ xiiis’). The site of this loft was given by Martin in his account of the tomb of Hopton. It will be observed that at this point there is a scattering of names carved high up on the wall in a position consistent with an overhead loft or gallery. The names, where clearly identifiable, belong to the seventeenth century. It might therefore be supposed that the organ was always in this position; and this hypothesis is to some extent supported by an ambiguous reference in the accounts which at least suggests that the medieval organ and pulpitum had different staircases (EA 29, under ecclesia ‘pro factura parvi hostii in superiori parte graduum pulpiti vid et pro alio hostio ad inferiorem partem graduum assendencium ad magna organa viiid’). These two staircases might well have been at opposite ends of the same screen, however. And there are two particular circumstances which make it virtually certain that the medieval organ was actually placed on Nicholl's screen (probably at one end with its own staircase). First, the ‘Avyse’ provides for a screen with an organ on it (WC, p. 366). Second, the records for organ repairs run fairly continuously from 1482–3 (EA 19 under ecclesia) up to 1568–9 (EAB iv). In the 1560s in particular, annual payments were made to a Mr. Howe from London, to keep the organ in a proper state. The last payment was made in 1569–70—at what date is unfortunately not known; but this coincides with the demolition of the screen. Nothing further is then heard of any organ until the new one was built in 1613–14 (EAB viii). It is difficult to avoid the conclusion that the disappearance of organ and screen are connected. The circumstances in which a well-maintained chapel organ vanishes need some imagination to be explained. The college did not sell it since no such entry occurs in the relevant ‘Fortuita Proventa’ of the Audit. Five men worked a total of twenty-one days demolishing the rood-loft and it is possible that in the course of this action the organ fell to the floor with a shattering crash from which it never recovered. It is also possible that it was placed in store with the intention of a subsequent reerection which never happened.

^{page 192 note 1} See A. Vallance, Roodscreens and Roodlofts, pl. 97: Burghill, Hereford (fifteenth century).

^{page 192 note 2} Ut sup., pl. 64: Guilden Morden, Cambs. (fifteenth century).

^{page 192 note 3} The overhang marked on the south side by the curving incised line projects for about 22 inches.

^{page 192 note 4} For the south side, this result is represented by the sum(or 15): for the north side, by the sum(or 10). The space of one seat on each side must be discounted in order to accommodate the return stalls. There are obvious problems in combining return stalls with a transparent screen, but it was sometimes done at parish-church level; see, e.g. Ranworth (Norfolk) and Southwold (Suffolk).

^{page 193 note 1} There are five rows of seats in front of the present stalls.

^{page 193 note 2} The number of churches where a third form was necessary seems to have been severely restricted. The Sarum Customs assume two formae only (see Frere, W. H., The Life of Sarum (C.U.P. 1898), p. 13)Google Scholar; and in the examples of secular stalls surviving the middle ages, only two rows of ‘formae’ are to be found. Of these, the most complete is probably at Lincoln; here pre-restoration engravings show the medieval arrangement still intact, including the ‘first form’ of benches for the choirboys in the front. The only reason for increasing the number of formae would have been numbers of people to be seated and one can only assume that, in most cases, such an increase was unnecessary. However, there are two instances in which a third form is referred to. The first occurs in the customs of St. Paul's Cathedral, London, in the recension drawn up by the Deans Ralph de Baldock and Thomas Lisieux (see Simpson, W. Sparrow, Registrum Statutorum et Consuetudinutn Ecclesiae Cathedralis Sancti Pauli Londinensis (London, 1873), pp. 1 ff.)Google Scholar; pt. vi, ch. 9 (ibid., p. 91) is headed De officio Puerorum infesto Sanctorum Innocencium and dated 1263; its general sense runs to the effect that an essentially laudable custom has been getting out of hand. One would like to know what had happened in 1262 because the recommendations all protect the canons. The ancient customs are to be observed and the choir boys ‘tabulam suam faciant et legant in Capitulo. Hoc tamen adhibito moderamine, ut nullum decetero de Canonicis Majoribus vel Minoribus ad candelabra vel turribulum vel aliqua obsequia ejusdem Ecclesie, vel ipsius Pontificis deputent in futurum, set suos eligant ministeriales de illis qui sunt in secunda forma vel in tercia.’ This particular provision is clearly putting a stop to the joke of the dean swinging a censer or carrying a candlestick (contrast here Salisbury; see Wordsworth, C., Salisbury Ceremonies and Processions Cambridge, 1901), pp. 53 fGoogle Scholar.) and limiting the choirboys’ choice to the cantarists, vicars in minor orders, and poor clerks. But it also seems to make it clear that in the thirteenth century there were three rows of formae in St. Paul's.

The same was true of Chichester (see Walcott, M. E. C., The Early Statutes of the Cathedral Church of the Holy Trinity, Chichester, London, 1877)Google Scholar, according to the Statutes promulgated in 1232. Here the chapters De Offensa Divinis Obsequius (ibid., p. 38) and De Pueris in Tercia Forma (ibid., p. 41) make it appear that there were three rows of formae, but numbered in reverse so that the choirboys sat on number three rather than number one.

It is less clear why the increase was necessary but St. Paul's at least had an unusually large number of cantarists accommodate (see Edwards, K., The English Secular Cathedrals in the Middle Ages (2nd ed., London, 1967)Google Scholar, ch. iv passim). The main point is that to have three rows formae was not unknown in the Middle Ages.

^{page 193 note 3} Generalizations are difficult because foundations varied on the precedence accorded to minor canons and priest-vicars.

^{page 194 note 1} See Winchester College statutes, printed in T. F., Kirby, Annals of Winchester College (1892)Google Scholar, ch.2:‘De modo dicendi missas matutinas etc…. Ulterius statuentes quod eciam singulis diebus dominicis et aliis diebus solempnibus et festivis per annum contingentibus in primis et secundis vesperis, matutinis, missis, processionibus, et aliis horis canonicis de die, custos ipsius colegii…ac vicecustos presbiteri et scolares omnes et singuli necnon magister informator scolarium et hostiarius as tres clerici de capella…in eadem cape la personaliter intersint. Dictique custos, vicecustos, presbyteri, magister informator scolarium et hostiarius, nostrique consanguinei quintum decimum eta-tis sue annum attingentes ac alii ipsius colegii scolares provectiores stallos in choro capelle predicte occupent.’ Note that the wording of the equivalent Eton Statute makes it clear that ‘provectiores’ is here to be taken with ‘scolares’ and not ‘stallos’.

^{page 194 note 2} See New College statutes printed by Bond (editor), The Statutes of the Colleges of Oxford (1853), ch. 42: ‘De modo dicendi missas matutinas etc…. Ulterius statuentes quod singulis diebus Dominicis … et aliis horis de die, custos ipsius collegii … ac Vicecustos, necnon ejusdem collegii socii et scholares omnes et singuli … intersint proces personaliter …, stallosque in choro occupent capellae praedictae. Scholares vero qui infra annos probationis fuerint, similiter intersint in divinis in capella praedicta, ut praedictum est; qui non in stallis sed ante stallos in choro stare debebunt.’ The New College statutes appear to have been the earliest (1400) in which this type of definition was found necessary. Similar regulations are not for instance to be found in the earlier statutes of Merton, Queen's, or Oriel colleges.

^{page 194 note 3} See All Souls' statutes, printed by Bond (editor), op. cit., ch. 22: ‘De modo et temporibus devote dicendi horas canonicos … et standi in choro ejusdem [capellae] … Praeterea, statuimus, ordinamus et volumus, quod dicti collegii custos in dextera parte chori dictae capellae teneat locum primum; in sinistra vero parte primum locum occutis pet vicecustos; inde autem ex utraque parte ipsius chori, theologiae Magistri, deinde decretorum Doctores et exinde legum Doctores; et extunc Doctores Medecinae, si qui tales fuerint in collegio memorato. Subsequenter vero in theologia baccalaurei, ac succesive Magistri artium, postea juris canonici, et tune juris civilis ac posteris medicinarum et artium baccalaurei; singuli videlicet ipsorum omnium secundum quod in Collegio seniores existant. Et postremo caeteri socii et scholares ejusdem Collegii non graduati, juxta tempus quod habeant in Collegio, loca teneant, cujuscollegii cunque fuerint facultatis; quern quidem ordinem in procespersonaliter sionbus etiam volumus observari. Capellani vero ejusdem Collegii stent inter socios praedictos prout custodi, vel in ejus absentia Vicecustodi, videbitur melius expedire ….’

^{page 195 note 1} See Magdalen College statutes, printed by Bond (editor), op. cit., chapter (no numbers): ‘De modo dicendi missas matutinas … et ordine standi in choro.”

^{page 195 note 2} Victoria History of the Counties of England: Oxford-shire, iii, 158.

^{page 195 note 3} When the Eton statutes were delivered (1443) the college was actually using the temporary accommodation of the old parish church.

^{page 196 note 1} Exactly what happened to the screen in 1569–70 is not thereclear. The rood-loft was, of course, removed. (The entry concerning its removal is printed in full in WC, p. 433.) This demolition involved repairs at floor level (ibid., ‘Item to Glover and his Laborer for ij days repairing and wasshinge ye walles where ye rood loft stoode and pavinge ye same place with gret stone and brick iiis iiijd’) but this does not necessarily mean that the entire screen was removed. If the screen had in some sense a double thickness (as has been suggested) the western layer could have been removed, leaving that to the east intact, with its attached return stalls, In favour of this suggestion is Woodward's account of Bost's tomb-slab (see Cole's manuscript) as ‘at the entrance into the Choir’, which would make good sense if a transverse screen still existed in the 1690s with an archway in perthe middle (and rather less sense, if it did not). It would also make better sense of the 1699 appeal with its wish furthat ‘all the Children of the Schole may appear under one view’ (WC, p. 447). Nevertheless, the medieval layout had undergone a few disturbances. First, a pulpit had been erected by 1578–9 (when a ‘hed over the pulpet’ was made, WC, p. 443). It was also recorded in 1625 (WC, p. 445) that Thomas Weever, a fellow, built ‘ye great pew under ye Pulpitt for the use of ye Fellowes, Scholmr and their Families’. The position of this pulpit is never revealed, Chapalthough it received various repairs and refurbishings in the seventeenth century. The presence of the organ and the carved names in the third bay suggest strongly that services still gravitated towards the east end and it is therefore probable that the pulpit was also built east of the original screen. In this case, Weever's ‘great pew’ would almost certainly have involved the removal of some of the medieval woodwork. More surprising is an entry in the audit for 1615–16 (under Templum, EAB viii): ‘Item paide by the handes of Henry Sayres for mending of Mr. Provost seate on the north side of the church and for matte canvas leather floxe and nailes for the same ut per billam xvi^s xi^d.’ The traditional place for presidents' seats was always on the right or south side of a chapel (see indeed the Eton statutes) and the mention here of a seat on the north is at first sight very puzzling. Its express mention, however, may mean that at this date the Provost for some reason had two seats. Perhaps one was temporary; or perthe haps he had a special seat for sermons opposite, or adjacent to, the pulpit. However, it seems unlikely that much furthat ther information will emerge to clarify all this. The fate of these fittings in 1699 is, for the most part, unknown, Maxwell Lyte records that the altar-rails went to Burnham church (3rd ed., 1899, p. 288) where they apparently still remain. For the rest, the following entry (part of the accounts for the refitting of the chapel under Reparationes, 1699–1700, EAB xiv) must serve as an epitaph ‘payd for carrying of 35 loads of Timber and old stuff from ye Chapaell and Timberyard to a barne at ye Sun in Eton out of ye workmens way £1. 13s. 6d.’l.

^{page 197 note 1} WC, p. 596.

^{page 197 note 2} WC, p. 410.

^{page 197 note 3} EReg i, p. 114. This was the year in which roofing material was bought for the ante-chapel (WC, p. 411). The wording of Street's will is puzzling, and its interpretation depends largely on who the ‘antiquos socios’ were and where their bodies were in 1482. It is possible that these had been exhumed and re-buried in the new church.

^{page 197 note 4} Oxford Bod. Lib. Browne Willis MSS. vol. xxiii, f. 26: ‘on the south side of the chapel is a large churchyard where the people of the parish bury who frequent this chapel. The Antient Parochial Church having long since been suffer'd to fall to ruine, this place is a peculiar.’ ‘Eaton is a peculiar; here was antiently a parish church, which belonging to the Colledge, they suffer'd it to fall to ruine and now the Parish all frequent the College Chapel and bury here etc.’ It has not been traced whom Browne Willis was quoting.

^{page 197 note 5} EReg i, p. 112.

^{page 197 note 6} EA 18.

^{page 197 note 7} Since wills form an important part of the evidence for the occupation and furnishing of the church, something must be said aboutthem. In 1443 as part of the endowment of the college, the provost of Eton was given archidiaconal powers within the parish in return for an annual payment to the archdeacon. These powers of course included the right to prove wills; as a result, a large number of wills are duly entered and proved in the college registers. It must be said that the provost's monoply of Eton wills was singularly effective. For this period the registers of the bishops of Lincoln produce only one will (Thomas Spenser—1453. Register of Bishop Chedworth f. 21). The records of the archdeaconry of Buckingham produce none at all (but they only begin in 1483). Virtually all the Eton wills therefore appear to be in the college registers.

^{page 197 note 8} John Wolston 1464 (EReg i, p. 105) left 8d. ‘ad lumen beatae mariae’. William Blakden 1470 (EReg i, p. 109) left 3d. ‘summo altari ecclesiae’ but these are exceptions.

^{page 198 note 1} EA 18.

^{page 198 note 2} EA 17 is 1475–6; EA 19 is 1482–3.