The Jesse Window in New College Chapel, Oxford
By David J Critchley
New College was founded by William of Wykeham, Bishop of Winchester, in 1379. The foundation stone was laid on 5th March 1380, and the College moved into the buildings on 14th April 1386. Little is known about the chronology of the building of the Chapel and the completion of its stained glass. The Chapel’s marble floor was not laid until 1411–12, and work on the pinnacles in 1412–13 may represent the final completion of the building. The glazing may not have been completed until 1404, and we know that Thomas Glazier, who was responsible for the glass, dined in the College four, five, or six times a year every year from 1387 to 1398.1
The Chapel’s largest window is the great west window, containing two tiers of seven principal lights each. The earliest reference to its subject matter seems to be a payment in 1456 pro reparacione Fenestrae de le Jesse (for repairs to the window of the Jesse).2 The original glass was replaced in 1765 by a design by William Peckitt of York (1731–95), and Peckitt was given the discarded glass in part payment. Peckitt’s glass was removed in its turn c.1780 in favour of a design by Sir Joshua Reynolds (1723–92) executed by Thomas Jervais (d.1799) and in 1788–9 was re-erected in windows nII and nIII. Later, perhaps not until the 1790s, Peckitt re-used some of this glass to glaze a window in York Minster (sVIII), and to glaze some smaller windows (nII and nIII) at St James’, High Melton, for John Fountayne, Dean of York 1747–1802.3 It is clear from the portions re-used in these windows that the window did indeed contain a Jesse, and that its tracery contained a Resurrection of the Dead.
This re-used glass is therefore our primary evidence for the original window. The three main lights of the York window contain fifteen panels depicting kings or prophets (Fig. 1), while the tracery contains the remains of a further three figure panels (B1, B2, D1), and of six panels from the Resurrection of the Dead (A1–6), of which one contains a devil, and the others contain souls risen from the grave. The re-used glass at High Melton appears to comprise a Christ in Judgement (nII 1–2b) and a Risen Christ (nII 1–2a), both largely made up of fragments of different figures; the head of an Apostle (nIII 2a); and a group of three naked souls (nIII 2c).
A further source may shed light on the window. New College’s sister establishment, Winchester College, was designed on similar lines. Winchester College Chapel was given a large Jesse window, in the east wall rather than the west wall, which appears to have been installed in 1393 (Figs. 2–3). It would be surprising if there were not much commonality between the two windows. The Winchester window was replaced with a copy in 1822, although some portions of the original have since been recovered and are incorporated into a window in a side chapel and in Fromond’s Chantry nearby. The copy, in the opinion of observers such as Charles Winston and J. D. Le Couteur, and when compared with the surviving portions of the original, gives a good account of the original design.4
The Resurrection of the Dead
Knowledge of the New College window’s original stone tracery is dependent on Loggan’s print of circa 1670. Some of the stone tracery was removed in 1779,5 but was restored in 1848 on the basis of Loggan’s print. This tracery contains twenty-two half-lights, of varying depths, and numerous smaller openings. Glass survives from ten lights at most, and in no case does a complete light survive. It may therefore be possible to draw general conclusions about the Resurrection of the Dead, but there is no immediate prospect of identifying the exact original location of any of the surviving pieces.
The numerous half-lights necessarily affected the designer’s approach. Vertical half-lights are well-suited to the portrayal of figures standing or kneeling in an architectural framework, and this is what we find in the surviving tracery lights (Fig. 4). The architectural frameworks, moreover, are identical to those in the tracery lights in the east window of the Chapel’s north aisle (nVI). In any Resurrection of the Dead, a Risen Christ, or a Christ in Judgement, would be expected at the centre, so it is not clear how the existence of two Christs at High Melton is to be accounted for. At the apex of the tracery in the New College window are two half-lights, with a small quadrilobe above them. Presumably Christ occupied one of these two half-lights, with perhaps the Virgin Mary in the other. The suggestion that there were two Christs, one in each half-light, seems improbable.6 The head of an Apostle surviving at High Melton, with its unkempt hair, is most likely to have belonged to John the Baptist.
Souls looking up and to one side are likely to be looking towards Christ, so we may assume that the souls in York Minster tracery light A3, who look up and to the right, were on the left hand side of the window, and those in A2, A4–5, and in High Melton nIII 2c, who look up and to the left, were on the right hand side. The souls in A6, who have already been judged, are being dragged to Hell with a rope around them and must have been on Christ’s left, on the right-hand side of the window, as must the devil in A1, who has his hands on the same rope.
Some assistance can be provided by the Winchester window, although the two windows differ substantially in the design of their tracery. John the Baptist appears at Winchester alongside the Virgin Mary, immediately below Christ and interceding with him. If the New College head at Long Melton is that of John the Baptist, this was perhaps his role. At Winchester there are numerous souls, many of whom look towards Christ, although some are horror-struck by the imminent terrors of hell. The depiction of Hell at Winchester appears to be more restrained: there is no roped group of dammed souls. There are no architectural backgrounds at Winchester, except for two lights in which heaven is depicted as a walled city. Two Winchester lights portray pairs of souls, but there is nothing corresponding to the groups of three souls encountered in the New College glass. However, our knowledge of how Christ was depicted at Winchester is limited, since the glass in the relevant light was replaced before 1802 with a Risen Christ, standing and displaying his wounds. Apparently, the original glass showed ‘the Resurrection,’ so it is possible that it too depicted a standing Christ.7
Reconstructing the Tree of Jesse
Woodforde, in the first full description of the York window, recognised the importance of the comparative material at Winchester, and started his analysis with a detailed account of the Winchester window. When he came to describe each of the surviving panels, he carefully noted the position of the corresponding panel in the Winchester window.8 Unfortunately, he had no key to enable him to tell whether the New College panel occupied the same position in the window as its Winchester counterpart. However, the close correspondence between the tabernacles formed from the branches of the vine tree in the York window and in the Winchester window (Fig. 5) suggests that the pattern of the vine that grows out of Jesse in the Winchester window (Fig. 6) may provide the necessary key, enabling the format of the Oxford window to be reconstructed.
First, it may be noted that nothing now in the York window corresponds to the central pair of lights in the Winchester window, in which the vine divides into two stems, proceeding upwards through David to the Virgin and Child and the Crucified Christ. The surviving York panels must therefore come from the lateral lights. In the Winchester vine, all the tabernacles on the lateral branches are formed by shoots which start by growing away from the centre lights and their depiction of the Virgin and Child and then return towards them, ending in two tendrils, which twist around the parent branch to complete the tabernacle. In the left-hand half of the window, therefore, these tendrils are on the right-hand side of each tabernacle, and in the right-hand half of the window they are on the left hand side. The shoots in the York panels terminate similarly and allow panels to be allocated to the left- or right-hand half of the Oxford window. On this basis, it can be argued that panels 1b, 2a, 2c, 3c, 4b–c, and 5a come from the left-hand half of the window, and panels 1a, 1c, 2b, 3a–b, 4a, and 5b–c from the right-hand half.
The lateral branches of the Winchester vine do not grow straight and level; when they branch off to left and right from the two central stems, they rise to form the first tabernacle, then descend to form the second tabernacle, and finally rise to form the third and last tabernacle, giving rise to the prominent lattice effect visible in the window. This is clearly replicated in the York panels and enables each panel to be allocate to a given vertical pair of lights. Thus, 2c belongs to the first pair of lights; 2a, 3c, 4b, and 5a to the second; 1b and 4c to the third; 2b, 5b, and 5c to the fifth; 3a, 3b, and 4a to the sixth; and 1a and 1c to the seventh.
In the Winchester window, the rows of kings ascend in chronological order, with each row comprising kings later than the kings in the row below. The only king breaking this pattern is Zerubabel, who is inserted out of sequence, but he does not survive among the York panels, so perhaps can be left aside at this stage. The arrangement of the kings at Winchester was additionally constrained by the placing of the prophets, the Crucifixion, and other features, and if the allocation of the York kings is refined on this basis, it is likely that the original Oxford placing is revealed. This is achieved if 5c (Abia) is placed in Row 3; 4b (Ioas), 4c (Ioram), 5b (Ochosias), and 3a (Ioathan) in Row 5; 2a (Manasses), 1b (Achaz), 2b (Ezechias), and 3b (Iosias) in Row 6; and 3c (Ieconias) and 4a (Sedechias) in Row 7. Rows 4 and 8 are the cusped window heads. The Winchester prophets were not arranged on a similar chronological basis, but the surviving Oxford prophets at York share their tabernacle designs with their Winchester equivalents, and it may be best to assume that all the prophets occupied the same positions at Oxford as at Winchester. This would put 2c (Helias) and 5a (Samuel) in Row 2; 1c (Amos) in Row 5; and 1a (Daniel) in Row 6. Some confirmation that we are on the right track is provided by the fact that each light now has a consistent background colour, alternating with that of its neighbours. On this basis, all fifteen Oxford panels surviving at York can be tentatively allocated to their original positions in the New College Jesse window (Fig. 7).
The York window contains three other figures, cut down to fit the tracery lights (Fig. 8). Panel B1 contains the upper half of St John the Baptist, surrounded by a seemingly random jumble of vine pieces. At Winchester he appears twice, once in the tracery interceding for souls, and once with King Richard II at the foot of the window. The York Minster Baptist is not interceding, but is pointing to the Lamb of God, and so he may also have come from a panel depicting Richard II and the Baptist, probably at the foot of the window. Panel B2 contains a figure labelled Amon Rex, against a red background. If this is Amon, son of Manasseh and father of Josiah, there is no suitable space available. However, he may be Amnon Rex, the son of David, who at Winchester is immediately to David’s left. Carlo Vercellone, in his work on variant readings in the Vulgate, noted apropos 2 Samuel 3:2, that some codices and most of the older editions read Amon for Amnon, so the use of this form of the name would not be surprising.9 Panel D1 contains a king against a blue background, but his label is a jumble of pieces. The most likely identification is perhaps Iosaphat, who at Winchester is immediately to the right of Abia. On this basis, Amnon and Iosaphat can perhaps be added to the fifteen panels. That still leaves numerous panels unfilled, from which no glass survives; but the surviving panels and the need to allow for essential features such as the Crucifixion, reduce the number of possibilities, and if these empty panels are assigned to the names of the kings and prophets occupying them at Winchester, we shall perhaps not be far from the truth (Fig. 9).
Two differences from the Winchester window emerge in this exercise. The first is that Ezechias (York panel 2b) occupies the position (panel 6e) which Zorobabel occupies at Winchester. Winchester omits Ezechias altogether, which is unusual given his historical prominence, and there is no evidence for Zorobabel in the Oxford window. It may be that Zorobabel replaced Ezechias at Winchester, for iconographical reasons that are yet to be determined. The second difference is that the Winchester window seems to have switched round the figures that at Oxford occupied the second and sixth lights in the upper tier. Thus Ioas, Manasses, and Ieconias occupied the second light in this tier at Oxford (i.e., using from now on the CVMA codes for the Oxford window, 5b–7b) and Ioathan, Iosias, and Sedechias the sixth light (5f–7f). At Winchester the position is reversed. The reason for this switch is obscure: it must have occurred at the planning stage since the Winchester panels are correctly drawn for their positions.
The Selection and Ordering of the Kings
The reconstruction sketched above is, of course, hypothetical, but it prompts two observations. The first is that the list of names has not been taken from the list of the ancestors of Christ in Matthew 1:6–11, which omits Amnon, Ioram, Ochosias (Ahaziah) and Sedechias (Zedekiah), not to mention Absalom, but from the list of the House of David in 1 Chronicles 3:1–16, which incudes David, Amnon, Absalom, Salomon, Roboam, Abia, Asa, Iosaphat, Ioram, Ohozias, Ioatham, Achaz, Ezechias, Manasses, Iosias, Iechonias and Sedecias.
The second is that it is possible to follow part of the planning process which preceded the construction of the Oxford window by comparing the sequence of the names in I Chronicles 3 with their positions in the vine. At some point in that process the designer must have transferred the names to some sort of diagram giving the proposed layout of the window. Perhaps he sketched his plan out on a board or on parchment; it may even be that he discussed the matter with Wykeham. Whatever form this process took, it involved – in the case of the kings – splitting up the list of names extracted from 1 Chronicles 3 into six rows, each comprising between one and five names (Fig. 10).
Thus, after starting with Jesse (1c–e), the designer placed the next king David (2d) on the row above. He then put Amnon on David’s left (2c) and Absalom on David’s right (2e). Solomon came next, in the centre and one row higher (3d). Again, the designer put the next king, Rehoboam, immediately on Solomon’s left (3c) and Abijah immediately on Solomon’s right (3e). For Asa, the designer went back to the left and put him on Rehoboam’s left (3b), and then back to the right for Jehoshaphat, putting him on Abijah’s right (3f). The designer followed the same principle for the next two levels, except that the central spaces (5d–6d) were unavailable, being already assigned to the Virgin Mary. So first he filled the space immediately to the Virgin Mary’s left (5c), then the space immediately to her right (5e), then back to the left (5b), and finally back to the right (5f). He repeated the procedure in the next row (6c, 6e, 6b, 6f). When he came to the topmost level, if it is assumed that the three spaces in the centre had already been assigned to the Crucifixion, he first put Jeconiah on the extreme left (7b), and finally Zedekiah on the extreme right (7f). At every level, therefore, he started at the centre and worked outwards, beginning with the left, and then crossing back to the right. The result is that, apart from David and Amnon, and Solomon and Rehoboam, no king sits alongside his predecessor or successor. The designer has systematically broken up any trace of a single continuous sequence.
The changes to the order introduced at Winchester upset the clarity of this procedure, which tends to confirm the long-held supposition that the Winchester window is the later of the two.
All this suggests that the designer did not see his kings as a single genealogical sequence, and did his best to prevent the viewer from reading the figures in that way. The viewer’s first impression will therefore have been of the members of the House of David, arranged only so that the later kings are higher in the window than the earlier ones.
Gestures and Attitudes
Le Couteur noted that mediaeval glass painters generally used action rather than facial expression to indicate emotion.10 For this reason the attitudes and gestures of the figures deserve our attention. All three surviving prophets in the outermost pairs of lights (1a–8a and 1g–8g) look inwards, either towards the figure next to them or up towards Christ. Elijah has the index finger of his right hand extended as if he were talking, as more obviously has Daniel. Of the seven surviving figures in the second and sixth pairs of lights (1b–8b and 1f–8f), nearly all face the prophet in the outermost light. Samuel looks outward towards Elijah, Manasseh faces halfway outwards towards Ezekiel, Jeconiah faces outwards towards Zechariah, Joathan faces outwards towards Amos, Josiah faces the observer, and Zedekiah faces outwards towards Malachi. Joash is stroking his beard as if in deep thought. Manasses, Joathan, and Josiah have the index fingers of their right hands extended. Of the five surviving figures in the third and fifth pairs of lights (1c–8c and 1e–8e), Joram looks up towards where Christ should be and has his right hand extended in a gesture of receptivity; Ahaz looks towards Christ and has his hand flat against his breast in a gesture of interiority; Abijah looks inwards towards Solomon; Hezekiah looks towards Christ and holds his right hand up in a gesture of receptivity; and Ahaziah faces the observer and may have his hand flat against his breast in a gesture of interiority.11 In short, the figures in the first two pairs of lights (1a–8a, 1b–8b) and in the last two (1f–8f, 1g–8g) are for the most part turned to face each other, and indicate by their gestures that they are talking. The figures in the inner pair of lights on each side (1c–8c, 1e–8e) are turned towards the infant Christ, or if lower down, towards their neighbour, and for the most part show that they are deeply affected by the experience.
The author is currently working on a study of the iconography of the Winchester window, for which there is much more evidence and which is likely to reflect the iconography of the Oxford window. To avoid needless repetition, it is proposed to restrict comments here to the iconographical issues raised by the features discussed above.
The Oxford depiction of the Resurrection of the Dead has substantial differences in detail from the Winchester window and appears to have been conceived on a grander scale. It is possible that at Winchester Wykeham was conscious of the young age of some of his scholars and toned down the depiction of Hell. In that case, the Oxford window may reflect a conviction that the college’s fellows and scholars were sufficiently adult to be exposed to the full implications of the life choices before them.
The designer’s recourse to the book of Chronicles for his list of kings makes it quite clear that the window does not depict the ancestry of Joseph. It is likely therefore that the emphasis was on the ancestry of Mary, who is described in an office lection for the Feast of the Nativity of the Blessed Virgin as being descended both from King David and from a priestly line.12 This sheds a new light on the presence of the kings and prophets in the vine. They are not there as ancestors of Christ. The reason for their inclusion is revealed rather by the fact that the kings are for the most part depicted either in discussion with the prophets or in contemplation of the arrival of the Christ child, the flowering of the stirps Iesse (the trunk of Jesse). They are, in other words, overcome by an extraordinary event, the sudden and unexpected bursting into life of the stirps Iesse and the emergence of the virga (the rod), which is the Virgin Mary, and of the flos (the flower), which is Christ himself. The kings and prophets are asking what the event means, explaining it to each other or even to us, or just gazing on it in rapt admiration. It is within the House of David that the miracle has occurred, and it is the House of David whose reactions we see depicted.
The chapel’s west window was second in size only to the chapel’s elaborate reredos and must have played a central role in conveying the chapel’s iconography. Given the prominence of the Virgin Mary in the iconographical programmes of Wykeham’s two colleges,13 it should not be surprising if we find that the west window’s principal theme is the lineage of the Virgin Mary.
The author is grateful to Dr Michael Stansfield, New College Archivist, for his help, to the editor of Vidimus and the anonymous reviewer for prompting clarification of the argumentation, and to Painton Cowan, the late Gordon Plumb, and York Minster for permission to reproduce photographs.
David J Critchley has a degree in Classics and a diploma in Theology. Recent publications include an account of the painted prophets on the North Crawley rood screen and a translation of La messe: une forêt de symboles: Commentaire allégorique ou mystique de la messe romaine traditionelle, by Claude Barthe.
- Gervase Jackson-Stops, ‘The Building of the Medieval College’, in John Buxton and Penry Williams (eds), New College Oxford 1379–1979, Oxford, 1979, pp. 147–92.
- Christopher Woodforde, The Stained Glass of New College, Oxford, Oxford, 1951, p. 7.
- Sarah Brown, ‘Our Magnificent Fabrick’: York Minster, An Architectural History c.1220–1500, Swindon, 2003, p. 283; Trevor Brighton and Brian Sprakes, ‘Medieval and Georgian Stained Glass in Oxford and Yorkshire. The Work of Thomas of Oxford (1385–1427) and William Peckitt of York (1731–95) in New College Chapel, York Minster, and St James, High Melton,’ in The Antiquaries Journal, 70/2, September 1990, pp. 380–415; and Brian Sprakes, The Medieval Stained Glass of South Yorkshire, Corpus Vitrearum Medii Aevi (Great Britain), Summary Catalogue 7, Oxford, 2003, pp. 45–50.
- Charles Winston, ‘A Short Notice of the Painted Glass in Winchester and its Neighbourhood,’ in Memoirs illustrative of the Art of Glass-Painting, London, 1865, pp. 66–7; J. D. Le Couteur, Ancient Glass in Winchester, Winchester, 1920, p. 71; ‘Notes on the East Window of College Chapel,’ in Supplement to the Wykehamist, no. 626, 21 November 1922, p. 273. See also Sarah Brown, ‘Medieval Stained Glass and the Victorian Restorer’, in Gareth Atkins, Jasmine Allen and Kat Nichols (eds.), Reframing Stained Glass in the Nineteenth Century British World: Culture, Aesthetics, Context, 19: Interdisciplinary Studies in the Long Nineteenth Century, 30, 2020, https://19.bbk.ac.uk/article/id/2901/ [accessed 24 March 2022]. For the recovered portions of the original, see John H. Harvey and Dennis G. King, ‘Winchester College Stained Glass,’ Archaeologia, 103, 1971, pp. 149–77; and Sarah Griffin and Eleanor Townsend, Medieval Glass at Winchester College, Winchester, 2021, pp. 7–10.
- Woodforde, op. cit., pp. 44–5.
- Brighton and Sprakes, op. cit., p. 403.
- John Milner, The history, civil and ecclesiastical, and survey of the antiquities of Winchester, Winchester, 2, 1809, p. 137.
- Woodforde, op. cit., pp. 102–5.
- Carlo Vercellone, Variae Lectiones Vulgatae Latinae Bibliorum Editionis, Rome, II, 1864, p. 327.
- J. D. Le Couteur, English Mediaeval Painted Glass, London, 1926, p. 42.
- For the gestures, see François Garnier, Le langage de l’image au moyen âge: Grammaire des Gestes, Paris, 2 vols, 1982–2003.
- Francis Procter and Christopher Wordsworth (eds.), Breviarium ad usum insignis ecclesiae Sarum, 3 vols, Cambridge, 1879–86, 3, col. 775.
- See Anna Eavis, ‘The Commemorative Foundations of William of Wykeham,’ in The Medieval Chantry in England: Journal of the British Archaeological Association, 164, 2011, pp. 169–195, and Veronika Decker, William of Wykeham als Collegegründer und Bauherr: Architektur und Glasmalerei zur Zeit Richards II., Kiel, 2017, pp. 246–74.