The Date of Beverley Minster and its Role in the Development of Northern Gothic in the Late 12th and Early 13th Centuries

Abstract

Drawing on close archaeological examination, a study of masons’ marks and especially new dendrochronological data, this article defends and develops Christopher Norton’s argument, published in 2009, that Gothic construction at Beverley Minster commenced soon after a well-documented fire in 1188. This dating was widely accepted until the 1860s, but since then there has been widespread consensus amongst scholars that Gothic construction at Beverley began no earlier than 1220. This radical re-dating disassociates Beverley with Lincoln Cathedral, and instead locates its design in relation to early Gothic construction at the abbeys of Fountains, Byland and Jervaulx, together with a wider group of ‘northern’ Gothic churches. New documentary and dendrochronological evidence also provide a more secure dating for Beverley’s 14th-century nave.

Keywords:

• Beverley
• dendrochronology
• Fountains
• Lincoln
• masons’ marks
• Purbeck
• syncopated arcading
• York

In 2009 Christopher Norton gave the annual lecture to the Friends of Beverley Minster and then published a short article suggesting that Beverley was started shortly after a documented fire largely destroyed the church in 1188.¹ This was a seismic shift from the accepted art-historical dating of the building, which envisaged a start in 1220 at the earliest.

Arguments for an early dating for Beverley depend heavily on the similarity between its Minster and Fountains Abbey, first pointed out in 1898 by John Bilson.² It was also mentioned by Larry Hoey in 1984, and explored in more detail by Christopher Wilson in 1991.³ These articles collectively broke new ground in our understanding of the development of northern Gothic architecture.

Stuart Harrison worked extensively on the loose stone collection at Fountains and Jervaulx abbeys in the 1980s and created new reconstruction drawings of the elevations at both sites.⁴ In 2010 Harrison expanded Norton’s arguments further to mention other sites which had similar architectural features and, in the case of Jervaulx Abbey, similar dating.⁵ In 2016 John Phillips published his book about Beverley Minster in which he adopted the new dating suggested by Norton.⁶ This dating has proved contentious, though no rebuttal has been published and work over the last few years on the dendrochronological dating of the Beverley roof timbers has now clarified the situation and confirmed the revised dating. It is therefore the purpose of this paper to review the development of Beverley Minster, introduce the evidence of the dendrochronology, and then to consider its architecture with reference to other buildings in the northern region.

history of beverley minster

Beverley Minster was one of three mother churches in the archdiocese of York, together with Southwell and Ripon. All three had collegiate status and remarkably all have survived relatively intact. Beverley is by far the largest and most splendid of the three. It has Anglo-Saxon origins dating from the time of St John of Beverley, who as Bishop John of York retired to the monastery in Beverley in 714 and died there in 721. King Athelstan visited the shrine of St John on his way to fight the Scots in 937 and granted the minster his tax, known as thraves, from the plough lands of the East Riding. This made Beverley a rich institution.⁷

The present church is almost a black hole in terms of knowledge of the structures which preceded it. The earliest description of any buildings at Beverley comes from a 16th-century manuscript chronology, which may be a copy of a much earlier source, now lost, and mostly ignored by previous commentators.⁸ This gives dates for the site from 687 to 1064:

AD 692 in the ancient place called Deirwuda, in Latin silva Deiorum [wood of the Deirans] now Beverley, the Blessed John was received into the parish church of this place, long since built in honour of St John the Evangelist, how the parochial and conventual church began to expand in the same year; founding the same Abbey he constituted monks about to serve God, in which he placed in command as father and first Abbot, the Blessed Berthinus. To whom he also associated seven secular priests’ (priests) or elders, with seven clerks (‘Levites’ or priests.) And in the chapel of St Martin on the south side of the church he set up nuns, with Blessed Polfrida in charge.

Bede refers briefly to the monastery in his accounts of John’s Miracles.⁹ Kynsige, archbishop of York (1051–60) later added a tower with two bells and Archbishop Ealdred also improved the church. This and the likely arrangement of the Anglo-Saxon church have been discussed by Richard Morris and Eric Cambridge.¹⁰ No further references to buildings occur until much later.

Recently, small-scale but hugely important excavations against the aisle walls and buttresses of the nave showed them to be Romanesque in their lower parts, implying that the notable misalignment between the nave arcade piers and buttresses north to south must have originated in the Romanesque aisled nave, which was the same width as the present building.¹¹ The location of these excavated buttresses is shown in Figure 1.

Fig. 1. Plan of Beverley Minster, with that of Romanesque Southwell Minster superimposed S. Harrison

Burials datable to the late 10th century, found at the same time, were aligned on a different axis to the present building, which may indicate a change from the pre-Conquest building. A building break by the Highgate porch in the NW section of the nave suggests the length of the Romanesque nave. Superimposing the plan of Romanesque Southwell over the present building shows a remarkable fit, aligning with the nave building break towards the west, the nave aisle buttresses, the crossing piers and location of the high altar at the east end (Fig. 1).¹² This implies that the proportions of the ground plan of the current building partially follow those of the Romanesque building.

The building contains a considerable amount of reused Romanesque ashlar, identifiable by its tooling marks.¹³ New foundations were constructed and no doubt used quantities of ashlar and rubble retrieved from the demolition of the Romanesque choir, crossing, and transepts. Evidence for this reuse can be seen most easily in the west wall of the main north transept, where Romanesque ashlar is visible three to four courses above the benches. It is assumed that this continues on the east side, although the walls are not visible because of shop fittings, and then into the north aisle wall of the choir and the chapter-house entrance. Some reused Romanesque ashlar blocks can also be seen in the south choir aisle, used as infill for the Gothic blind arcading. There is also a substantial block of reused ashlar in the south aisle of the nave.

The plan has an aisleless east end or Lady Chapel of two bays and an eastern transept with two bays in each arm that flank the second bay of the Lady Chapel (Fig. 2). The eastern crossing is rectangular in plan with the long axis north to south. West of the eastern transept is the choir or presbytery of four bays to the main crossing and western transept. This western transept has an exceptional length of four bays in each arm and also features eastern and western aisles. While it seems highly likely that the crossing piers occupy the same position as the Romanesque ones and that the main body of the transepts were as wide, it is unlikely that they were as long or had a western aisle. The width of the main transept bays is also remarkable as they are equal to the presbytery bays, when usually transept bays are significantly narrower.¹⁴ It can also clearly be seen that the first bay of the nave was part of the same build as the western transept and crossing to support the crossing and tower. It must have abutted the Romanesque nave, which presumably either survived the 1188 fire intact or was reroofed immediately after it.¹⁵

Fig. 2. Plan of Beverley Minster by J. Bilson in Architectural Review, 3 (1894–98), adapted with revised dating by the authors

The overall design is remarkably homogenous, with relatively narrow bays of 16’1” (5.016 m), most likely based on the medieval perch, and also perhaps the bay spacing of the Romanesque church.¹⁶ The aisle walls have wall arcades with moulded trefoiled arches supported on foliate capitals with moulded bases and Purbeck shafts. The windows are single lancets flanked by blind pointed arches in each bay. The main arcade is tall in relation to the bay spacing and occupies around half the overall height (Fig. 3). The triforium has a double arcade of syncopated form in which a lower arcade of pointed arches with quatrefoils in the arch spandrels is set behind an upper arcade of trefoiled arches. The single detached Purbeck shafts of the lower arcade are very short, with moulded capitals. The upper arcade capitals are set level with the apex of the lower arches and have triple shafts. The capitals all have marble abaci. The arches are blind and the mouldings of the trefoiled arcade are decorated with dogtooth. The height of the triforium is relatively low and this allows greater height to the clerestory arcade. Structurally, the triforium arcades front a rear wall that is quite thin and hides large triforium arcade arches behind it in each bay.¹⁷

Fig. 3. Beverley Minster: main bay elevation in the presbytery, showing the tall proportions in relation to the bay width

S. Harrison

The bays are divided by Purbeck ‘marble’ detached shafts, which start as single shafts in the spandrels of the main arcade supported on small corbels (Fig. 3). A foliate capital set at the top of the main arcade arches allows the shafts to be increased to a triple group that rises to the base of the clerestory, where there is another moulded capital. A second triple group rises to the level of the clerestory capitals where they support the high vault. Notably, the vault springers are built with a tas-de-charge of five courses of level bedded sections before the angled bed-joints for the voussoirs begin. Externally, the clerestories are supported by a fully developed system of flying buttresses, which may indicate that the master mason was familiar with French buildings and their adaptions in early Gothic England.

The clerestory has a wall passage fronted by an arcade of five arches that are graduated in height. They consist of two small narrow arches at each side flanking a wider central lancet window arch opening. The arch mouldings are decorated with bands of dogtooth, are supported on single Purbeck shafts and octagonal moulded capitals with Purbeck abaci. In the presbytery the tallest shafts flanking the central lancet are octagonal in form, with seven of the facets hollowed and a flat one facing the wall passage.¹⁸

The rear wall has a similar blind arcade of simply chamfered arches supported on polygonal moulded corbels. The central lancet window is chamfered with a roll-moulding set in hollows worked on the internal screen angle. The front screen arcade is strengthened in places by strong iron bars spanning from the outer wall and sometimes from capital to capital.¹⁹ This iron reinforcement allows for structural daring, as seen in the relative tallness and lightness of construction of the front arcade of the clerestory, holding it firmly in place. The space between the inner and outer arcades is spanned by a single narrow transverse arch or barrel vault.

This design is used throughout all the main arcades and into the first bay of the nave. It was clearly a very successful design in which the relatively narrow bays give a far greater impression of height than is in fact the case. So successful, in fact, that when the new nave came to be built in the 14th century it was carried on in the same style, largely unaltered, albeit with larger tracery windows. This consistency led John Leland to comment that the minster was ‘of a fair uniforme making’.²⁰ Such historicism is relatively rare in England, except at Westminster Abbey, Worcester Cathedral and St Albans, and may be the work of Oliver de Stainfield, who was master mason in 1308 when work on the new nave is thought to have started.²¹

Historical evidence

Two major events are significant in the development of the present building. A great fire was recorded in 1188, in which the minster was burned together with most of the town. This was documented at the time by Roger of Howden and also by an inscription on a lead plaque, found when the tomb of St John was opened in 1664 and reported by William Dugdale. The plaque also recounted a search made in 1197 for the tomb of St John:²²

Anno ab incarnationem domini M^oC^oLXXX^oVIII combusta fuit haec ecclesia in mense Septembri in sequente nocte post festivum sancti Mathaei apostolic: et in anno M^oC^oXCVII VI idus martii facta fuit inquisition reliquarum Beati Johannis in hoc locus et inventia sunt haec ossa in orientali parte sepulchri et his recondite et pulvis cement mixto ibidem inventus est et reconditus.

In the year 1188 from the Incarnation of Our Lord this church was burned in the month of September on the night following the feast of St Mathew the Apostle [that is 26 September], and in the year 1197 on 6 Ides March [9 March] an enquiry was made into the relics of the blessed John in this place, and his bones came to light in the east part of the grave, and these concealed, mixed with mortar and dust, were discovered and concealed in the same place.

One of the miracle stories attributed to St John also mentions a tower collapse and its consequences. This event is undated but is one of a group of miracles that are said to have occurred within five years. Susan Wilson discussed this at length, suggesting that the events took place between 1211 and 1219 and that those described in the second miracle story must have taken place after the lifting of the interdict on 2 July 1214, because it describes the singing of hymns and ringing of bells, activities banned during the interdict. She concluded that the miracle stories were arranged chronologically, and it was likely that they all took place after the interdict was lifted.²³ That might suggest the collapse occurred sometime between 1215 and 1220. This is an important conclusion with regard to the sequences and dates of the high roofs delineated by the dendrochronology, as we will show.

The full Latin text was given by James Raine:²⁴

De ruina cujusdam turris ecclesiæ Beverlacensis

Erat eodem tempore, in medio crucis ipsius basilicæ, præcelsa quædam turris exstructa, admirabilis pulchritudinis et immensæ, adeo ut in eo se jactitaret virtus et subtilitas artis cæmentariæ. Processum erat eatenus in ipsius turris fabrica, ut completum esset opus lapideum; hoc solum totius superesset consummationi, ut superponeretur tectum lapidei operis, proportionatæ celsitudinis. Artifices qui præerant operi, non tantum quantum oporteret circumspecti; non tam prudentes, quam in arte sua subtiles; magis invigilabant decori, quam fortitudini; magis delectationi, quam commodo stabilitatis. Qui cum columnas quatuor erigerent cardinales, velut totius supercollocandæ molis fulcimina; eas subtiliter, quamvis non firmiter, inserebant operi antiquo, eorum more qui pannum novum assuunt inveterato. Unde factum est, ut nec bases, nec stylos columnarum illius efficerent firmitudinis, ut sistinere sufficerent molem immensam tam admirabilis et tantæ arduitatis; quarum debilitas, quamvis in processu operis satis posset adverti per hiatus et rimas partium, per columnarum quarundam marmorearum fissuram in longam a base usque ad epistylium; ab operis tamen incepti nequaquam arbitrabantur continuatione desistendum, cum tamen certum sit, paratum esse ad ruinam illud ædificium, quod super debile collocatur fundamentum. Quanto igitur lapidum ampliorem superponebant cumulum, tanto magis accelerabant turris præcipitium: eo immensiores factæ sunt stylorum et basium rimæ, quo plus eos præsumpserunt onerare.

Tandem factum est, ut, metu imminentis ruinæ, tam cleri quam plebis desisteret pars magna ab ingressu ecclesiæ. Sacersdotes nihilominus et Levitæ, et alii qui ex injuncto sibi officio ad chorum frequentandum erant astrici, horis statutis et debitis eo convenerunt, ut Divinis vacarent obsequiis. Confisi sunt quod non permitteret eos obrui, vel morte subitanea aliquatenus præoccupari, cui totis viribus, pura mente, interna devotione satagebant famulari. Non diffidebant de Confessoris sancti subsidio, cujus relliquiæ in eo, qui præsens erat, continebantur inclusæ locello. Patefecit autem, qui sequebatur eventus, quod Ejus Confessoris est miserator Dominus. Nam cum circa mensis Octobris initium media nocte surgendum esset ad confitendum Domino, ac nocturnum officium de more celebrandum in choro; quidam sacerdotum, qui, Deo, ut reor, disponente jacens in lecto, partem noctis transactam præter morem solitum duxerat insomnem; diuturnitate jacendi, et dormiendi etiam impotentia fastiditus, surgens, etiam adiit matricularios; quos reperit dormientes; a somno suscitavit, eosque ut pulsarent induxit, cum superesset adhuc quasi unius horæ spatium, ut eis signum surgendi et pulsandi præberet horoscopium. Cum igitur, congregato ad ecclesiam clero, nocturno decantaretur synaxis, corruit non longe a decantibus lapidum pars magna de turri; quorum audito fragore omnes nimirum magno sunt timore perculsi. Summa igitur festinantia se transferebant a stallis suis, et inceptum continuabant officium, stantes ad alterutrum latus altaris. Non multo post auditus est alius fragor priore major, lapidum plurimorum iterum de turri corruentium; ac si præmissa levi quadam et simplici admonitione ut recederent, sequeretur edictum peremptorium, cui contumaciter supersedentibus minitaretur supplicium.

Relicto igitur choro, locum tutiorem, magis a turri distantem, arbitrabantur adeundum, et transeuntes sub pendente ruina descenderunt in ecclesiæ naviculam; ubi consistentes a latere fontium, ad finem usque cœptum perduxerunt officium. Vix completo officio ad domos suas, licet satis vicinas ecclesiæ, pervenerant; et ecce! tota turris fundotenus collapsa, partes adjacentes secum trahens ad casum, vehementem dedit fragorem, in auribus humanis horrendum. Advertere licet in hoc eventu quam admiranda fuerit gratia Salavatoris, quantæ fuerit efficaciæ virtis sancti Confessoris. Disponente namque Domine, partes suas interponente Sancto, turri ruituræ provisum est illud tempus ruinæ, quo turbæ laicali læsio nulla posset inferri. Cum ministris ecclesiæ facta est admiranda dispensatio, cum et præter morem consuetam horam surgendi prævenirent, et bipartitus lapidum casus, præconatus vicem gerens, eis persuasisset ut alio se transferrent. Indultum erat etiam ipsis, qui intra septa ecclesiæ recubuerunt, matriculariis, dum, ut amoverent thoralia sua, transferrent stratoria; eis nulla ingereretur læsio corporalis. Deventum erat igitur adhuc, ut orientalem partem naviculæ ipsius ecclesiæ in chorum oporteret accommodari, altare vero supra tumbam erigi, feretrumque ultra medium tumbæ directe collocari. Quod cum factum esset, satis efficaci conjectura deprehensum est, quod lumen ad tumbam dicendum est aperuisse, et eum portendisse qui futuris erat eventum; quod scilicet corpus sancti Confessoris merita, decurso præsentis vitae stadio, nobis annuatur bravium permansurum, ab Eo Qui vivit et regnat Deus per infinita, sæcula sæcularum. Amen.

Wilson translated this as:

At that time there was a very high tower which had been built over the crossing in the middle of this church, which was of such remarkable beauty and size that it boasted of the skill and competence of the stonemason’s craft. The building of this tower had advanced to such an extent that the stonework had been completed; all that remained to be done was that a stone roof of proportionate height should be put on. The masons who presided over the work were not as careful as they should have been, much less prudent than they were subtle in their art; they paid more attention to ornamentation than to strength; more to aesthetics than to proper stability. For when they were erecting the four main pillars as supports for the whole massive superstructure, they inserted them elegantly, rather than firmly, into the old building works in the manner of those who sew a patch onto old clothing. Therefore it happened that they made neither the bases nor the uprights of the pillars of a strength sufficient to sustain a huge structure of such a remarkable lofty height, in spite of the fact that their weakness could be seen well enough in the course of construction by the crevices and cracks, and by the splitting open of some marble columns along the length from the base all the way to the architrave. Yet they did not at all consider stopping from continuing the work they had started, although it was certain that the building, which was built on a weak foundation, was ready to collapse. Thus the more they laid on the mass of masonry above, the more they hastened the fall of the tower: and the more they presumed to overload the uprights and bases, the bigger the cracks became.

At length it happened that a large number of the clerks as well as other people stopped going into the church because the fear of imminent collapse. Nevertheless the priests and Levites and others who were obliged to go to the choir regularly as a consequence of their bounded duty assembled at the appointed hours so that they might perform divine office. They trusted that he, whom they served industriously with all their strength, with pure mind and inward devotion, would not allow them to be crushed, or to be somehow seized by sudden death. They did not despair of the help of the holy confessor, whose relics were preserved enclosed in a chest in that very place. However, the event that followed revealed the Lord is one who pities his confessor. For, about the middle of October, when it was necessary to get up to praise God in the middle of the night and to celebrate the customary office in the choir, one of the priests who (by God’s will as I believe) as he had been lying in bed and had spent part of the night awake contrary to his usual practice, disdained having to lie in bed for a long time and unable to sleep, got up and went to the watchmen whom he found asleep. He aroused (them) from sleep and led them in to ring the bells, although nearly an hour early, to give them the signal to get up and strike the hour. Then, the clergy having assembled in the church, while the nocturnal office was sung, a large part of masonry fell down from the tower not far from those who were singing. When they heard the noise of this they were all totally seized by great fear. So with the greater haste they moved from their stalls and continued the office they had begun, standing on the other side of the altar. Not long after another noise was heard, greater than the first, again of a great many stones falling down from the tower, just as if, some mild and simple warning having been sent in advance so thy might draw back, a peremptory command would follow which would threaten punishment to those who stubbornly ignored it.

Therefore, abandoning the choir, they decided to move to a more secure location farther from the tower, and passing under the overhanging ruin, they descended into the nave of the church, where, taking up a position at the side of the font, they conducted all the way to the end of the office they had begun. When the office was complete they had scarcely reached their homes, although they were fairly near to the church, when lo, the whole tower which collapsed into ruin down to the ground, pulling to destruction with it the adjoining parts, made a violent noise, terrible to human ears … Therefore it came about that the eastern part of the nave of this church had to be turned into the choir, the altar had to be erected over the tomb, and the shrine had to be set up directly behind the middle of the tomb.²⁵

In summary, the masons were said to be heightening a crossing tower that was completed apart from a stone spire.²⁶ Cracks had appeared as the tower was erected and were ignored, causing the marble shafts to split along their length. The presence of marble shafts under load suggests that there were internal galleries, like the clerestory arcades, within the tower.²⁷ The new masonry of the piers was said to have been poorly worked into the older fabric. The dramatic fall of the tower allegedly started as a night service was taking place and the total collapse occurred after it had been completed. Following the collapse, the canons set up the high altar next to St John’s tomb in the nave and returned his shrine to that location. That would hardly be possible if the central tower had fallen, given the proximity of the nave tomb to the central crossing. Moreover, as we shall argue below, the dendrochronology shows that the roofs of the transept were erected in this period, which would be impossible if it was the central tower that had collapsed. There is also no evidence in the central crossing to indicate a collapse and rebuilding, but there is in the eastern crossing. Until a late edition of Thomas Rickman’s book, it seems that earlier commentators had accepted that the present church was begun after the 1188 fire. Yet this account led numerous authors since to suggest it was the central tower that fell and necessitated a total rebuilding, starting in the 1220s.²⁸

the eastern crossing

Beverley has two crossings because there is an eastern as well as a western transept. Evidence of possible collapse in the eastern crossing has until recently been ignored in connection with the miracle story. It first came to attention in a footnote by Robert Willis in J. L. Petit’s article about the church.²⁹ The western arch of the eastern crossing above the vaults shows spandrels decorated with quatrefoils and blind arches supported on Purbeck shafts (Fig. 4). Purbeck marble was also used for the string-course above the arches, and mortar residue on top shows that masonry has been removed from above the string-course. These features were clearly meant to be seen from the crossing below but are now closed off by a high vault. The spandrels are riven by large vertical cracks and also have open putlog holes for scaffolding. The corners show eastward returns with responds for similar arches. Willis suggested they were a remnant of an earlier east front, replaced when the east end was enlarged.³⁰ Bilson, closer to the mark, explained away their presence by suggesting they were part of an intended lantern tower abandoned in construction; an analysis apparently accepted by a number of more recent commentators.³¹ Lantern towers seem to have been what late-12th- and 13th-century patrons wanted in their churches, not least because the bells were usually housed in a belfry above and rung from ground level. Ripon, Selby, Southwell, Salisbury and Lincoln had open lanterns, and we suggest that so did both the crossings at Beverley before they were later closed with vaults.

Fig. 4. Beverley Minster: the north spandrel of the western arch of the eastern crossing, showing the quatrefoil decoration and blind arch with Purbeck marble shafts. The string-course in the corner above is also of Purbeck. Note the open putlog hole and wide fissure in the masonry

J. Phillips

As Hoey first pointed out, at each corner of the eastern crossing are stair turrets rising from clerestory level, their loop windows visible externally.³² Clearly, a lantern tower was intended, with access to an upper stage at each corner. One can therefore assume a gallery or triforium stage with an arcaded lantern with a wall passage above it, and most likely a belfry above the lantern fitted with bells rung from ground level—virtually a standard arrangement at the time.³³

In reviewing the past literature on this subject it seems that there was a kind of groupthink regarding the evidence in the eastern crossing. Commentators understandably associated the description of the tower that collapsed ‘in medio crucis ipsius basilicæ’ with the western crossing, yet the eastern crossing is equally in the middle of the church; in this case the canon’s church east of the main transept. Revising this meant overturning the long-accepted dating of the construction of the building to after 1220; Bilson, for example, discussed the dating and on his plan bracketed the main building between 1225 and 1245.³⁴ Commentators also fixed on the indulgence issued by Archbishop Walter de Grey in 1232, which speaks of the ‘Ecclesia Beverl miserabili ruina et enormiter deforma’ (the church of Beverley, miserably ruined and enormously deformed).³⁵ This indulgence has surely been blown out of proportion as it was for only twenty days and referred to damage in the past, the fall of the eastern crossing tower. Henry III twice granted protection to collectors for the fabric fund in 1235, and in 1252 he granted forty oaks from Sherwood Forest.³⁶ On 18 May 1261 Archbishop Geoffrey de Ludham consecrated the high altar.³⁷

It was Norton who in 2009 revived the idea that building actually started immediately after the fire of 1188, and that it was an eastern crossing tower that had fallen. He went further and suggested, like Willis, that the eastern transept and the eastern crossing were an addition to an already completed four-bay presbytery.³⁸

Interpretation of the eastern crossing piers has proved contentious, with some commentators suggesting they were radically altered following the abandonment of the lantern tower above them (Fig. 5). Christopher Wilson discussed these piers at some length and agreed with Peter Brieger that they had been altered.³⁹ They are unusual in their form, with conventional responds to the arcades and transept walls that abut them, although the faces towards the main vessels do not have the usual piers with clusters of shafts but instead nearly flat pilasters decorated with tiers of blind arches and marble shafts that are supported on moulded corbels, with a mixture of foliate and moulded capitals, and round and polygonal abaci, that are progressively corbelled outwards as they ascend. The short uppermost part has a conventional arrangement of shafts supporting moulded capitals to accommodate the crossing arches.

Fig. 5. Beverley Minster: the north-west crossing pier and eastern crossing seen from the south-east transept

S. Harrison

The corbelling out of the crossing piers increased the space at ground level. According to Norton’s theory these piers form part of a second phase in the development of the east end of the building, but we shall show this is unlikely to be the case.⁴⁰ Hoey also discussed and described variations in detail such as foliate and moulded capitals in the design of these crossing piers at some length, but, like Bilson, he accepted them as primary features. He also commented on the regularity of the masonry coursing. ⁴¹ Despite what has been said before it is notable that the piers course regularly with the masonry of the adjoining arcades and walls, with very few minor exceptions. Moreover, as we shall see, the evidence of the masons’ marks suggests that these are the original piers. The unusual design may have been intended to highlight the area surrounding the high altar and the shrine of St John. Yet today it is difficult to imagine the original vision behind this crossing as the tall lantern tower, which would have flooded the space with light, has been lost.

The evidence of masons’ marks

Beverley Minster is fortunate in having a complete set of masons’ marks from end to end and floor to vault.⁴² More than 250 banker masons were employed at one time or another throughout the east end.⁴³ While many of them may have only worked for a season or two, a substantial core of individuals can be tracked around the building, providing a good indication of how construction progressed. It is clear that some masons were employed to make mouldings for piers and voussoirs while other groups simply cut ashlar blocks. This will be clarified in the following discussion, where it must be remembered that the numbers indicated reflect the core group, which in all cases is augmented by further, less frequently occurring marks.

The first group identified is that of fifteen men who worked continuously on mouldings for the arcade piers (Fig. 6a). Their marks are present on virtually all the piers on both north and south sides of the eastern part of the building, including, significantly, the four piers of the eastern crossing. A second group of four masons is responsible for the two sixteen-shaft piers situated in the eastern transept arcades (Fig. 6b). Hoey tried hard to explain the changes in some of the arcade mouldings in the eastern transept and the presence of these exceptional piers.⁴⁴ They can now be seen as part of a systematic programme carried out by specific teams of masons. These masons then join the first fifteen men working on the arcades. Notably, masons’ marks and masonry coursing can be followed around the eastern crossing piers, particularly in the case of the south-west pier, and form a homogenous group. Had the piers been altered one would expect detectable variations intruding into the record, but there are none. Marks on voussoirs and mouldings at higher levels are much harder to see, but enough have been found to show that the same men largely worked throughout the whole elevations of the east end, main transept and into the nave.

Fig. 6. a–d showing various masons’ marks

J. Phillips

A third set of marks of interest are all on ashlar. This represents a group of five men, distinguished by their very large banker marks, which largely cover the surface of the blocks they cut (Fig. 6c). These marks are found in a clearly defined area on the north side of the building. The men responsible for them cut ashlar for the chapter-house entrance, which may indicate that they also cut ashlar for the chapter-house itself.⁴⁵ The marks then continue upwards from the spandrels of the choir arcade into the clerestory. They are also present in the east side of the main north transept at higher levels, including the stair turret in the north-east corner of the transept. It is perhaps relevant to note that the north aisle of the presbytery and the east side of the main north transept has a base plinth that is different from the rest of the building. This appears to be another example of a team of masons working a specific part of the building and to some extent ‘doing their own thing’. No doubt the overall work initially started as a partial envelope around the burnt-out Romanesque structure, which would have been demolished piecemeal to provide recycled stone for the new building.

Another notable point is that the style of the capitals at the top of the buttresses in the same walls is distinctly different from those around the rest of the building.⁴⁶ One of the third set of marks is also present on two stones of the ashlar filling of one of the blind arches of the surviving crossing arch of the collapsed eastern lantern, and also in the stair turret above the north-west pier of the east crossing, indicating that this was always part of the planned original construction, and not a later addition as Norton suggested (Fig. 4).⁴⁷ This is confirmed by the fact that the first group of fifteen masons were also cutting all the mouldings for the four piers of the eastern lantern.

Another group of particular interest worked in a clearly defined area of the upper levels of the north and south sides of the eastern transepts and the Lady Chapel. This group of five marks (Fig. 6d) may give a clear indication of the extensive area of damage from the tower collapse because of the way the marks are distributed, and the masons themselves may have been brought in at the repair stage during the 1220s (though they also appear in the upper part of the south transept).⁴⁸ It is perhaps also notable that the two high vault bosses of the north-east transept are the only ones of figurative rather than foliate design. That might indicate that the vaults were rebuilt after the tower collapse.

This group of five masons worked, with others, on the north-east transept from the spandrels of the arcades up into the clerestory in the east wall and the bays that join the west wall of the eastern transept to the north-west crossing pier. One of them also appears in the stair turret above the north-west pier of the east crossing, indicating possible repair in that area. The transept north wall appears to have remained intact.

On the south side their area of work is more extensive, which may indicate the main area of damage. They are found on the lower levels of the two south bays of the Lady Chapel and also at clerestory level. In the two east arcade bays of the south-east transept they are present from the arch spandrels upwards into the clerestory and roof space. They are also found in the stair turret in the south-west corner of the transept that leads from this transept up into the roof space at levels above the arcade. Strangely, they also appear in the upper level of the stair turret in the south-east corner of the main south transept and also in the adjacent transept clerestory, which might indicate that they had first worked on the completion of the south transept before the tower collapse.

The tentative conclusions we can draw from this evidence is that three core teams of masons worked on the building over time. One team was present throughout the main build period, which saw the whole structure completed into the first bay of the nave. The second worked at the same time on the north side of the presbytery and east side of the north transept, and may also have built the chapter-house. The third and smaller team with other seasonal masons seems to have worked on repairs to the fabric after the fall of the eastern crossing tower, but may have already appeared during the final stages to finish the south transept. It is also clear that the eastern transept, Lady Chapel and crossing were not additions to an existing presbytery but part of the same phase of building.

the development of the nave

In 1292 the canons issued a contract to Roger of Faringdon, a London goldsmith, to make a new splendid and elaborate gold and silver shrine for St John.⁴⁹ The work was not completed until 1308, when the balance of the money in the shrine fund was transferred to the fabric fund for the building of a new nave, and it is assumed that work began soon afterwards.⁵⁰ Archbishops of York issued indulgences for the fabric in 1302 and again in 1308, and the chapter appointed Elias de Lumby as collector for the fabric in the dioceses of York, Durham, Carlisle and Lincoln.⁵¹

In 1319 William Canon of Corfe supplied four Purbeck marble shafts to Exeter Cathedral for their new pulpitum, the details recorded in the Exeter fabric rolls. An identical pair of Purbeck marble shafts and bases was used for the new reredos behind the high altar at Beverley, most likely at a similar time.⁵² Masons’ marks on the reredos are identical to masons’ marks on the ogee arches carrying the sculptures of the musicians in the north aisle of the nave, which suggests they may be contemporary. The flowing tracery, mouldings and streamlined shafts of the reredos are consistent with Decorated architecture in Yorkshire c. 1320, and also match the ogee arches and window tracery of the north nave aisle. The reredos must have been finished before work started to assemble the Percy canopy which abuts it (note the fillets of the reredos spiral staircase, which had to be cut back when the Percy tomb was inserted). It was built to commemorate Lady Eleanor Percy, wife of the 1st Baron Percy, who had predeceased her in 1315.⁵³ She died in July 1328, but the canopy contains the arms of England quartering France, adopted by Edward III when he claimed the throne of France in 1337, giving a terminus post quem for the canopy.⁵⁴

The parish of St Martin had an altar in the nave of the Romanesque church, moved in March 1324 to a new building to the south-west of the old nave and built on the first floor above a charnel house.⁵⁵ Work was probably interrupted by the Great Famine of 1315–17, which may have stopped building completely, and probably did not resume for some time. Following their victory at Bannockburn in 1314, the Scots terrorized the north of England for years, and in October 1322 inflicted another defeat on Edward II at Sutton Bank near Byland Abbey. Fearful for their safety, the canons immediately petitioned Robert the Bruce, asking him to take them and all their possessions under his protection ‘that they may receive no damage from your men’.⁵⁶ The Scots were noted for destroying crops, and to build their church the canons depended to a great extent on their tax of sheaves of corn from every ploughland in the East Riding.

If work on the nave resumed after 1322, then progress was halted by the Black Death, which reached Beverley before June 1349.⁵⁷ How far up the building work had gone before the plague struck is difficult to ascertain. There is a series of beam slots on the south clerestory that may indicate that a temporary roof had to be put in place to cover the building site until work could be resumed at some time in the future. During building works parts of the nave had to be partitioned off.⁵⁸

There are three distinct building periods in the nave. As mentioned above, the first bay and a half is part of the first building of the east end, providing abutment to the crossing and central tower.⁵⁹ The second phase is the replacement of the Romanesque nave as far as the bay which at present contains the porches, and the final phase continues from this point to include the west towers and complete the building. There are also three distinct phases for the construction of the nave roof, indicated by the distribution of carpenters’ assembly marks on the roof trusses. The first two sections of the roof extend from the central tower to the bay containing the porches (Fig. 7). These have been discussed at length by Alison Armstrong and David Cant.⁶⁰ Each phase is delineated by carpenters’ marks, which are placed differently in each phase. Trusses T1–19 have them on the soffit of the timbers; in the second phase, trusses T20–42 have them on the west side of the timber and in the third phase trusses T43–61 have them on the east side of the timbers. Reused timbers in trusses T1–42 have been dated to 1177–1202 and will be discussed in detail below.⁶¹ The tie-beams that support them have been dated 1328–64 (truss T9) and 1322–58 (truss T20). ⁶² The final trusses from the porches westwards have been dated 1367–92, with a tie-beam after 1365 (truss T59).⁶³ This fits in broadly with other dating indicators discussed below. Some of the tie-beams in the nave were replaced in the 18th century and date to the winter of 1736–37.

Fig. 7. Beverley Minster: plan showing the three phases of building and dendrochronological dating in the nave roof

S. Harrison

Allowing for a period of recovery after the Black Death, work on the final phase of the building must have started c. 1360–75 or perhaps as early as the 1350s. The first priority seems to have been to complete the addition of the clerestories, to a design which is still essentially Decorated with its complex tracery, but which anticipates the change to Perpendicular. It was done in two sections, the first of which includes the bays from the central tower to the east side of the bay containing the porches. There is a clear building break in the stonework on the exterior at this point, and a marked change in the size of the ashlar used between the east and west spandrels of the main arcade in the bay containing the porches and also in the nave walls above the vault.⁶⁴

We have a number of different ways to date this final phase of building from the porches, up to and including the west towers. The first is the ‘tricking’ of the heraldry of the original west window glass by Sir William Dugdale, Garter King of Arms, in 1641. His drawings are preserved in the British Library and show the arms of various people connected to the minster, including the then archbishop of York, Thomas Arundel, and Richard II, king of England. This dates the design of the glass in the window from 1388, when Arundel became archbishop, to 1399 when Richard was deposed by Henry Bolingbroke.⁶⁵

Archbishop Richard Neville, who had preceded Arundel, was accused of treason four times between November 1387 and 3 February 1388, when he was finally convicted and fled to France. All his assets, both personal and those of the archbishopric, were frozen.⁶⁶ Very soon afterwards, on 10 June, the canons petitioned the Crown for the release of oaks from South Burton Wood, given to them by the archbishop and ‘cut down before the judgement of forfeiture rendered against him’.⁶⁷ Dendrochronology of the western section of the roof dates the timbers to no later than 1392, which probably indicates that these were the timbers from South Burton Wood. The roof would have needed to be complete and watertight before the builders contemplated inserting the vaulting.⁶⁸

In 1394, the governors of the town granted a licence to the canons for the movement of stone from the beck to the minster by sledge ‘for the repair of the vault beyond St John’s tomb’, which clearly indicates the nave vaulting.⁶⁹ Masons’ marks on the vault ribs indicate that they are contemporary with the structure of the western part of the building, including stairs and windows, and together with the brickwork that comprises the filling of the webs between the ribs, the marks put the date of this part of the building towards the end of the 14th century.

The webs of the nave vault are built with ‘walltiles’. These ‘walltiles’—or ‘tegulas’ as they are referred to, as the word brick was not commonly used until the early 1400s—can also be fairly accurately placed in the timeline.⁷⁰ The tiles used are a unique size, 10.5 × 5.25 × 2 inches, which was peculiar to the Hull area during the 14th century and the early part of the 15th.⁷¹ Very similar sized tiles were used for the walls of Hull earlier in the century, made in a tileyard owned by the De la Pole family that ceased production as soon as the walls were finished. Tiles of a similar size were also used to build parts of Holy Trinity church in Hull around 1340, and also the gatehouse at Thornton abbey in north Lincolnshire after 1382, where the tiles are recorded as coming from Hull. In 1391, citizens of Beverley were granted a license to produce 3,000 ‘walltyles’ per year at Grovehill, which may have been the tiles for the minster vault.⁷² Tiles of the same size were used at the friary, possibly at Watton Priory, and at the North Bar of Beverley, which dates to 1409.⁷³ After that they seem to go out of use in favour of more standard sized bricks.

The evidence of dendrochronology

The minster roofs have been the subject of extensive dendrochronological sampling in recent years, firstly by Ian Tyers and considerably more by Alison Arnold and Robert Howard of the Nottingham Tree Ring Dating Laboratory (NTRDL). This has tended to confirm the 18th-century dates for the reconstruction of the high roofs, and gives felling dates for many of the reused medieval pieces of timber and the medieval nave roof trusses.

The high roofs of the minster were the subject of extensive reconstruction in the 18th century so that today only the nave roof retains its medieval form. In reconstructing the roofs, new principal trusses were introduced at each bay division and were linked by purlins supporting common rafters. These were largely recycled rafters and collars from the original roofs, often sawn down the length to make two common rafters. They retain evidence of their former use in the form of numerous notched lap and mortice joints that often allow reconstructions of the trusses to be made.

The north transept roof must have been the first to have been taken down because there was a huge engineering operation in 1718–21 to restore the dangerously leaning north transept façade to the vertical.⁷⁴ That involved partially taking down the adjoining clerestories and triforiums, and the removal of the aisle and high vaults together with the flying buttresses. Plaster vaults with wooden ribs replaced the stone originals. The medieval roof must have been largely dismantled and replaced with what we see today, though there is documentary evidence for some later-18th-century restoration works to the high roof.⁷⁵ Contemporary accounts show that timber was supplied in 1721 from the Sandbeck Estate at Maltby, near Roche Abbey, for the roof of the north transept.⁷⁶ The roofs over the presbytery, eastern transept and south transept must also have been reconstructed in the 1720s and 1730s, and the dendrochronology has confirmed felling dates for the new timber used in the rebuilt roofs (see Table 1 below).

Table 1. Dendrochronological felling dates in chronological order

Reused timbers in the medieval roof frames of the nave		NTRDL	1177–1202
		Tyers	1175–1212
Western bays of the presbytery		NTRDL	1181–1206
South transept		NTRDL	1193–1218
North transept		NTRDL	1214–39
Eastern bays of presbytery and south-east transept		NTRDL	1234–59
North-eastern transept		NTRDL	1233–58
Nave east bays	tie-beam T9	Tyers	1328–64
Nave mid bays	tie-beam T20	Tyers	1322–58
Nave west bays	tie-beam T59	Tyers	After 1365
Nave west bays		NTRDL	1377–92
The 18th-century roofs
Treadwheel crane in the crossing tower		NTDRL	1724
Nave mid bays	tie-beam T31	Tyers	After 1727
Nave mid bays	tie-beam T37	Tyers	Winter 1736
Choir	tie-beam T4	Tyers	1734–67
Choir	tie-beam T8	Tyers	1733–68

While this extensive work leaves reconstructed roofs that, apart from the nave roof, incorporate much earlier timber and no original trusses, their form has been largely recovered by measurement of the recycled material to reconstruct the trusses on paper. Stefan King made detailed measurements and drew reconstructions of the medieval roof trusses in the 1990s. This study remained unpublished until it was edited with added information and additional reconstruction drawings by Harrison and published in Phillips’ book.⁷⁷ The reconstructions show typical frames of the late 12th and early 13th centuries, and compare well with the contemporary high roofs at Lincoln Cathedral (Fig. 8). Moreover, an important lost aspect of the minster was identified during the study of the roof timbers. It was clear that the present eaves parapet walkways in the eastern arm and eastern and western transepts represent a remodelling of the original arrangements. As built, the frames had rafters that most likely overhung the eaves without the benefit of walkways and parapets, and the trusses were reduced in length when the parapets were introduced.⁷⁸ The original absence of walkways is reflected in the internal arrangements of the roofs, where there were no access points for parapets and instead a crawl space with entrances at the top of the staircases to the roofs allowed access along the internal wall top. Even today, access to the external parapets is through hatches and doorways set into the leads of the roofs.⁷⁹

Fig. 8. Beverley Minster: reconstruction drawing showing the type of roof trusses from the choir, east transept and west transept

S. Harrison

Here we must also consider the most likely process by which the medieval roofs were replaced. It seems very unlikely that whole roofs were taken down in a single phase, because providing temporary watertight roofing over such large areas would have been virtually impossible. It is far more likely that they were taken down a bay at a time, the new trusses were erected and the sarking boards and leads reinstated. Each bay had been boarded separately in the 18th century; a fact recently confirmed when the south-east transept and the nave roofs were stripped to their boards for releading. That would also mean that the recycled timbers most likely came from somewhere close to their new location. This seems to be borne out by the discrete dating phases relating to the various parts of the church, outlined below. In certain areas the original roofs also appear to have been in a very poor state, with either rotten or worm-eaten rafters, which explains why a substantial number of reused common rafters are scarfed to make up the required length for the new common rafters, often with three pieces of timber and in extreme cases using up to five pieces of old timber.⁸⁰

Successful dating depends on having the heartwood to sapwood boundary within the sample. In the north-east transept, for instance, the average heartwood/sapwood boundary ring date is 1219, allowing an estimated felling date to be calculated for the four timbers sampled to within the range 1234–59. This felling date range is calculated using the estimate that 95 per cent of mature oak trees in this region have 15–40 sapwood rings. Exact felling dates can only be calculated when the whole of the sapwood is present to the bark in the sample.⁸¹

There is therefore a margin within the date brackets where the actual felling date is likely to exist. In some instances these bracketed dates can perhaps be refined by the use of documentary evidence. Grants of timber from Sherwood Forest in 1244 and 1252 must relate to the eastern bays of the presbytery and the eastern transepts. Similarly, the timber used in the western nave roof was already felled in June 1388, when, as mentioned above, the chapter applied to the king for its release.

These dates show that there is considerable leeway within the suggested date ranges; care must therefore be exercised when interpreting the results (summarized in Fig. 9). The chart above indicates that the western choir roof had most likely been built by 1206 at the latest, and possibly a few years before that. It confirms that the building must have been started relatively soon after the devastating fire of 1188, as Norton suggested.⁸² It was followed in sequence by the roofing of the south transept, complete by 1218 and probably some years before that. The north transept was next and complete by 1239, but likely earlier. It is at this point we encounter a return to work on the eastern bays of the presbytery and the east transept. This was roofed sometime between 1233 and 1259, most likely in 1244 and 1252, given the grant of oaks from Sherwood Forest in those years.

Fig. 9. J. Bilson’s plan of Beverley Minster, with the felling dates of the timbers in various parts of the roofs superimposed

S. Harrison

At first sight this return to work on the east end of the church might seem an aberration in what otherwise appears to have been a regular and conventional east to west building sequence. Something clearly happened and the collapse of the eastern crossing tower immediately comes to mind. It must have occurred somewhere in the sequence and, given the apparently regular flow of the building work from one part of the building to the next, without any obvious interruption, particularly in the main transept, it seems likely to have fallen somewhere between the completion of the south and north transepts. Given the felling dates for the north transept roof timbers of 1214–39 it would seem likely that the work there continued uninterrupted despite the fall of the tower. Norton’s suggestion that the Lady Chapel and eastern transept were added later would fit this sequence neatly, but the evidence of the fabric and masons’ marks shows this interpretation to be unlikely, as we have demonstrated.

We must also consider the evidence of reused timbers in the later nave roof, from which roof truss designs can also be reconstructed. These reused timbers show a variety of types with common rafters, inner rafters, soulaces and collars. King first reconstructed a truss design from some of these reused timbers. It was very similar to the roof built later over the new 14th-century nave. That roof design, it should be noted, was different from the truss design used in the presbytery and transepts, also reconstructed by King, which had common rafters with ashlar pieces rising from a sole plate and a large collar supported by soulaces below and struts above. There was also an upper collar and struts (Fig. 10). Further research has expanded on King’s work and added more information and truss reconstructions. Three in situ tie-beams in the north-east transept show that frames with tie-beams also existed on the bay divisions (dated 1233–58). The later-14th-century nave roof is similar, with much larger tie-beams on the bay divisions, but the regular trusses had an additional feature of an inner rafter rising from a sole plate to cross the soulaces and joint to the lower collar. Both framing techniques were current designs in the late 12th century, as the roofs over the chapter-house vestibule and St Hugh’s choir at Lincoln clearly show, and by the 14th century were possibly rather old-fashioned in terms of truss design.⁸³

Fig. 10. Reconstructed trusses from reused timbers in the nave roof, with group felling dates 1177–1202 and 1175–1212

S. Harrison

Explaining variations in truss design is difficult. The timbers that have been dated from the western choir roof are collars that are seemingly long enough to show that they did not have inner rafters. It might be the case that trusses with inner rafters were only employed in conjunction with the tie-beams on the bay centres. One common rafter on the south choir roof has a reused soulace or inner rafter with a scissor lap joint, which may indicate that trusses with inner rafters did exist within the western choir roof.

Several common rafters in the west side of the north transept roof have similar reused timbers with scissor lap-joints, suggesting that trusses with inner rafters were also employed in that roof. Notably, a recycled collar supports the axle of the treadwheel crane in the central tower and it shows joints for upper struts, a pair of soulaces and also a pair of inner rafters. That is likely to have been put in place in 1724, which is the exact felling date of the crane axletree—known because it has the full heartwood to sapwood to bark surviving. At that date the only roof which had been rebuilt was that of the north transept, which is therefore the most likely original source of the reused collar.

King did not recognize that the reused timbers in the nave roof probably came from two sources, because many of the recycled timbers used as inner rafters on the 14th-century nave roof have angled mortice joints that indicate they came from a more steeply pitched roof. In fact, their pitch can only be matched by the roofs over the eastern transepts (Fig. 10). They were originally common rafters and have a mortice joint for the collar and a notched lap joint for the soulace. Moreover, there are also reused soulaces or inner rafters with scissor lap-joints whose angle of intersection match those of the east transept roof frames. It would seem likely therefore that these roof trusses had inner rafters as well.

One is then drawn to the conclusion that the reused timbers in the nave roof originated from two sources. Firstly, it seems clear that the reused timbers from the steeper pitched roofs must have been from those of the eastern transepts. No other roofs have such a steep pitch. Secondly, the timbers from a roof design similar to that of the later nave, with inner rafters crossed by soulaces, could have come from the Lady Chapel roofs and therefore also originated in the original eastern roofs of the church.

It is likely that the original late-12th- or early-13th-century Lady Chapel and eastern transept roofs were damaged by the fall of the eastern crossing tower and were dismantled on grounds of safety while repairs took place. No doubt in the immediate aftermath of the collapse considerable quantities of timber were needed to prop up various parts of the structure, for which the roof trusses would have proved very useful. We can be certain that these reused timbers were either stored under cover until they were reused in the 14th-century nave roof or that they were redeployed to another roof shortly after their dismantling. Had they been left exposed to the weather and sunlight for any length of time they would have changed colour. It may be the case that the Romanesque nave roof was rebuilt using some of this timber and it was then reused a second time in the later nave roof. Whatever the case, the timbers retain only one set of joints from their original framings, with no trace of any secondary workings until they were reused in the present nave roof. That might mean that some complete roof trusses were salvaged and reused over the old nave, but they would have been limited in number to those that covered the Lady Chapel roof, minus any damaged by the tower collapse. Further dendrochronological sampling of the nave roof trusses might give a better indication of how much of their timber is actually reused and how much is 14th-century.

the development of the northern style of gothic

This reassessment and new evidence for the development of Beverley Minster changes the accepted dating of the building, moving it over thirty years. It now remains to set it firmly in place within the development of Northern Gothic.

Beverley Minster is a precious survivor as the only intact and largely unaltered building in a particular strand of the northern style of Gothic that developed in the late 12th century. The main arcade design was carried largely unaltered throughout the presbytery, eastern transept, choir, main transept and into the first bay of the nave. The piers are mainly of eight shafts and of two designs that alternate in the arcades, one with four large round shafts at the cardinal points and smaller keeled shafts in the angles, the other with filleted round shafts at the cardinal points and pointed nibbed round shafts of similar size in the angles (Fig. 11). They stand on octagonal pier bases that have a mid-height moulding around the base and water-holding moulded pier shaft bases. The capitals are plain and have fillets on a moulded rim. They support arches with complex mouldings that feature filleted rolls and hollows (Fig. 12). The triforium stage and clerestory have already been described in detail and the latter is a key structural element of the other churches in the style. Purbeck marble is used prolifically for abaci and shafts, with single shafts on corbels in the main arcade spandrels and triple shafts in the triforium and clerestory rising to support the high vault.

Fig. 11. 3D models of the arcade piers (from left to right) at Byland, Jervaulx, Beverley, Fountains and Meaux

S. Harrison

Fig. 12. Profiles of the main arcade mouldings at Jervaulx, Beverley and Fountains shown as J, B and F

S. Harrison

The aisle bays have trefoiled wall arcades with largely stiff-leaf capitals supported by Purbeck shafts.⁸⁴ There is a single lancet window in each bay flanked by pointed blind arches with moulded capitals with round abaci. The aisle windows of the choir bays have been enlarged and tracery inserted but enough remains to show they were like those in the transepts, which retain their original form. One of the most distinctive elements of the design is found in the north aisle of the choir, where the bays incorporate a double staircase that gave access to the now lost chapter-house. Its polygonal plan and dimensions were discovered and reported by Bilson.⁸⁵

Byland Abbey

The fountainhead of the northern style of early Gothic architecture was the great Cistercian church at Byland Abbey, where building work commenced sometime in the early 1170s and finished around 1190 or shortly afterwards.⁸⁶ With double aisles to the main transept, as at Beverley, Byland is also a remarkably ambitious building. The earliest parts of the building at the east end and transepts show use of a variety of waterleaf, palmette and crocket capitals. Yet this style evolved during construction and chalice capitals gradually ousted the leaf forms and in turn began to evolve in design. We can see this in the western parts of the nave, with distinct differences between capitals in the north and south aisle and in the western arcade responds. None of the in situ capitals show the full evolution, which is now represented by one complete capital and also some loose fragments broken from main arcade pier capitals. They show a moulded form with a distinct bell rim under a squared impost (Fig. 13). The aisle capitals show imposts and abaci of rounded moulded form (Figs 14 and 15), but if combined individually with the bell rim they would closely resemble the type of capitals employed at Beverley, Fountains, Jervaulx, Meaux and Revesby.

Fig. 13. Byland Abbey: half-pier capital, showing individual bell-rim segments under square imposts

S. Harrison

Fig. 14. Byland Abbey: north nave aisle respond capital showing single circular rim over three shafts

S. Harrison

Fig. 15. Byland Abbey: south nave aisle respond capital showing circular imposts and abaci

S. Harrison

Byland was also one of the first buildings in the north to employ lancet windows, with three large lancet windows set in a lancet arch wall arcade on the west front, and further lancets in the clerestory in the south transept. This suggests the clerestories had lancets throughout. The west front also has filleted roll and scroll mouldings and the hood-moulds of the doorways, rose window jambs and internal rere-arch have prominent dogtooth decoration. Nail-head decoration and rolls with fillets were also employed on the tracery of the great rose window.

Jervaulx Abbey

It seems no coincidence that all these details also appeared at Byland’s daughter house of Jervaulx, which had copied earlier architectural details from Byland such as its cloister arcading and the arrangement and detail of the chapter-house entrance.⁸⁷ With double aisles to the main transept, as at Beverley, Byland is also a remarkably ambitious building. The plan of Jervaulx was also an adaption of that at Byland, which replaced the presbytery ambulatory arcade with a full height east wall. This plan in a Cistercian church is the earliest example known in Britain of what became a common Cistercian design, with an aisled presbytery of four bays. Yet its architect went further and also adopted the eight-shaft pier design and an octagonal pier base design from the mother house, as well many other design details, as we have seen. This also included new, more varied designs for some of the pier shafts (Fig. 11).

The piers at Beverley, Fountains, Jervaulx, Meaux and Revesby show distinct shaft designs with combinations of plain round shafts and round shafts with fillets, some with pointed nibs and plain keels standing on octagonal bases. This seems to be a reference to the piers in the east arcade of the south transept at Byland, and in particular two bases employed there. It was a form that the Byland master only briefly used before reverting to circular bases, but which became a notable feature of a later series of churches in the north (Fig. 11).

Dogtooth and rolls decorated with fillets were features that appeared at Beverley and Jervaulx at virtually the same time. Beverley also employs shafts with a sort of moulded fillet on the cardinal shafts of the main piers in the eastern transept and notably the presence of the fillets and pointed nibs on pier shafts is articulated on the base mouldings and also in the capitals. These particular Beverley piers are notable in having sixteen shafts and at Jervaulx a pier in the south transept has twenty-four shafts, with a similar fillet on the keeled cardinal shafts, albeit bifurcated as a double fillet (Fig. 11). A similar double fillet appears on the central soffit moulding of the main arcades at Beverley (Fig. 12).

The main piers at Beverley have circular moulded capitals with a double-filleted rim, similar to some fragmentary and one complete half-capital at Jervaulx, as well as less complete examples (Fig. 16). The fragmentary sections include parts of the capital from the twenty-four-shaft pier in the south transept or some other identical pier. It seems to have had a single circular rim rather than individual circular rims to each capital shaft (Fig. 23, right).⁸⁸At Jervaulx we are unfortunately hampered by the loss of many of the piers and their bases, but enough survive to show a variety of types similar to those at Beverley (Fig. 11).

Fig. 16. Jervaulx Abbey: large moulded capital from the main pier

S. Harrison

Some details at Jervaulx, such as the mouldings of the main arcades, were first published by Edmund Sharpe in 1848.⁸⁹ He assembled loose mouldings lying around the site to recreate the arcade profile. The voussoirs employ fat rolls decorated by fillets, similar to those seen in the main arcades at Beverley (Fig. 12). More loose stone fragments at Jervaulx can be assembled on paper to show that the aisle windows were virtually identical to those at Byland, with external nook-shafts standing on moulded bases and internal keeled rolls to the jambs. They differ from the Byland model in being of lancet form rather than round-headed design.

Yet amongst the numerous architectural fragments excavated in 1805 at Jervaulx are pieces which relate to the same clerestory design seen at Beverley and Fountains Abbey, described briefly above. Like them, the clerestory at Jervaulx had a wall passage with a rear external wall that was articulated with chamfered blind arches supported by moulded polygonal corbels (Fig. 23). The front of the wall passage was screened by a wall arcade supported on single shafts with moulded bases and capitals.⁹⁰ The roof of the wall passage was a single broad transverse pointed arch spanning the bay. This design was a distinct departure and architectural innovation from early Gothic clerestory arcades at Byland Abbey and Ripon Cathedral, which employed cross-passage lintels to support individual arches from the front arcade to the rear wall.

Elements of this design exist as loose stones scattered throughout the stacked-up architectural fragments at Jervaulx. There are examples of the polygonal corbels that supported the arcading on the rear wall (Figs 17 and 18), fragments of the blind arcading, elements of the clerestory windows and many fragments of the front wall arcades with their supporting polygonal capitals and bases (Fig. 19). Springers and ribs from the aisle vaults survive and show variation in the rib profiles (Fig. 20), indicating changes in design and possible phasing of construction. They also show that the vaults were built with tas-de-charge level bedding.

Fig. 17. Jervaulx Abbey: a moulded polygonal corbel from the rear wall of the clerestory

S. Harrison

Fig. 18. Jervaulx Abbey: a double polygonal corbel from the cross-passage lintel set on the clerestory bay division

S. Harrison

Fig. 19. Jervaulx Abbey: moulded polygonal capital from the front screen wall arcade of the clerestory

S. Harrison

Fig. 20. Vault rib and other profiles from Jervaulx (J), Beverley (B), Fountains (F), Meaux (M), Bridlington (BN) and Guisborough (G). J1–2, diagonal ribs; J3–5, transverse ribs; F6, diagonal and transverse aisle vault ribs; F9, high vault transverse and diagonal ribs; F10, single transverse high vault rib; B7, aisle diagonal and transverse rib; M8, aisle vault rib; J11, B12, F13, BN14, G15, clerestory front arcade mouldings

S. Harrison

Evidence for a high vault is strikingly lacking. A main clerestory capital set between two bays lacks any provision for a bay dividing shaft and suggests that there were no high vaults or any shafts demarcating the bays (Fig. 21). Sufficient elements of the clerestory screen arcades survive to show that it is likely that both three-arch arcades, like the clerestory at Fountains, and five-arch arcades, like the clerestory at Beverley, were present in the church.⁹¹ Evidence for iron bars like those at Beverley and Fountains is prolific, with sockets in the tops of capitals that in some instances show evidence for two bars dogged and leaded into sockets and channels set at right angles to each other (Fig. 22). This suggests that bars spanned the clerestory passage and also the front screen arches. Dogtooth decoration appears on the south nave doorway and also in larger format on loose window jambs from the façades of the church, which appear to have featured wall passages. All the evidence can be brought together to show the full elevations (Fig. 23).

Fig. 21. Jervaulx Abbey: main clerestory double capital set on the bay division. This shows that no bay dividing shaft was provided. Note the capital is upside down

S. Harrison

Fig. 22. Jervaulx Abbey: top of a single polygonal clerestory capital, showing the sockets and channels for iron reinforcing bars set at right angles to each other

S. Harrison

Fig. 23. 3D reconstruction of Jervaulx Abbey, showing the three possible clerestory designs

S. Harrison

Fountains Abbey

At Fountains Abbey, the 12th-century aisleless square-ended presbytery was demolished and replaced with a five-bay aisled presbytery, which extends two more bays into an eastern transept that had nine altars. Like Jervaulx, this was a two-storey elevation with a clerestory wall passage, but unlike Jervaulx it had a high vault. Signs of collapse and large cracks and fissures in the masonry meant that in the time of abbot John Darnton (1478–95) the high vaults were dismantled, the vault springers were cut back and a wooden roof was installed.⁹² Though ruined, Fountains is now the building that most closely resembles Beverley in design, and it makes similar use of marble, from Nidderdale, for shafts, abaci, corbels and capitals.⁹³

The presbytery aisle walls show a continuous wall arcade of trefoiled moulded arches supported by marble shafts, though only one now survives at the west end of the north arcade. The bays are divided by triple marble shafts supported on marble corbels and the single lancet windows are flanked by blind arches of curious rounded form. William St John Hope called them curious hook-shaped arches due to their shape (Fig. 24).⁹⁴ Overall, the elevation closely resembles that at Beverley, where the flanking arches are pointed in the main transept aisles. The same design was presumably employed in the south presbytery aisle before the windows were enlarged (Fig. 25).

Fig. 24. Fountains Abbey: the north presbytery aisle, showing the ‘hook’ shape of the arches flanking the lancet windows

S. Harrison

Fig. 25. Beverley Minster: north transept, west aisle

S. Harrison

The presbytery main arcades at Fountains were unfortunately lost when the central tower collapsed eastwards and brought them down. Enough survives of the responds at each end to indicate the design of the elevation, and the Chapel of the Nine Altars shows the clerestory design very clearly (Fig. 26). Parts of the elevation survive as fragments excavated in the 1840s by Richard Walbran, and a reconstruction drawing was published by J. A. Reeve of 1892.⁹⁵ There is sufficient evidence to show that the piers alternated between relatively plain octagonal piers and eight-shafted examples with unusually large core blocks, four round shafts in the angles, and large marble keeled shafts which butted up to the core on the cardinal points. A third pier type is shown by the east responds, which have filleted cardinal shafts and had round detached marble shafts in the angles. The western responds are simple half-octagons. The pier bases are very similar to those at Beverley, of octagonal design with a mid-height base moulding. Altogether the evidence indicates a form of alternating piers that we suggest was derived from the alternating piers at Beverley (Fig. 11).

Fig. 26. Fountains Abbey: one bay of the clerestory in the Chapel of the Nine Altars, showing the system of rear wall corbels and chamfered blind arches. At left of centre is the cut back high vault springer with tas-de-charge level-bedded blocks

S. Harrison

Elements of the mouldings from the main arcade arches also survive and, from their spandrels, small corbels supported single shafts dividing the bays. At the base of the clerestory were corbels below triple marble shafts and then marble capitals to support the vault springers; in essence this is the same bay-dividing arrangement of shafts employed at Beverley. This arrangement can still be seen in the Chapel of the Nine Altars. As mentioned above, the vault springers were cut back but the coursing can be followed and, just like Beverley, there are five courses of level-bedded stones before the angled jointing for the top of the tas-de-charge on the springers.

The high vaults at Beverley spring from the clerestory screen arcade capital level and those at Fountains did exactly the same, but with a clerestory arcade of three arches in each bay, not five. The Beverley vaults are kept relatively low and the ribs employ more than one centre of curvature to achieve that. By contrast, the Fountains high vault rose considerably above the clerestory eaves and the vault cell abutments sloped down from the ridge of the vault to the clerestory (Fig. 27). The vault ridge needed to be so high to butt up to the existing 12th-century eastern crossing arch, which was remodelled with an inserted wall rib to accommodate the high vault. It must have been this aspect of the vault geometry that caused the movements in the arcade and Nine Altars walls despite the provision, as at Beverley, of flying buttresses to the main arcades.⁹⁶

Fig. 27. 3D reconstruction of the bay elevation of the presbytery at Fountains Abbey presbytery. The vault cells have been omitted for clarity

S. Harrison

The design of the roof trusses at Fountains may have also contributed to the spreading of the walls and failure of the high vault. This was because it was impossible to have any roof trusses with tie-beams due to the height of the vault, which rose high into the roof space. There can only have been collars, though presumably soulaces or scissors were used as braces, as at Beverley. As the walls spread and the roof trusses started to open up at the joints they must have put additional pressure on the walls. Like Beverley, the main roofs of the new extension at Fountains also lacked parapets, oversailed the clerestory walls and had eaves corbels of similar mask design.⁹⁷

Guisborough Priory

Byland, Jervaulx and Fountains present the clearest surviving evidence of the northern style of Gothic, but others have lesser but in some cases conclusive evidence they also had similar elevations and design details. At Guisborough Priory in Cleveland, the church is largely destroyed but excavations in the 19th century and in more recent times produced evidence of a clerestory with a front screen arcade. These survive in the form of fragments from the screen arcade and from the jambs and voussoirs of windows from the clerestory rear wall that are very similar to those at Beverley, Fountains and Jervaulx (Figs 28 and 20). Other early Gothic fragments include keeled shafts and a moulded pier respond base that indicate a developed early Gothic church from around 1170 into the later 12th century. It seems to have included an arcade of syncopated design in which the arrangement of the two arcades was similar to that seen in the Beverley triforium, but with waterleaf capitals (Figs 29 and 30). These fragments collectively suggest that Guisborough may have played a major part in northern Gothic development.⁹⁸

Fig. 28. Guisborough Priory: mouldings of the surviving stones from the clerestory, showing the close similarities with Jervaulx, Beverley and Fountains. Left: front screen arcade arch. Right: clerestory window

S. Harrison

Fig. 29. Reconstruction of a syncopated wall arcade from Guisborough Priory, with waterleaf capitals

S. Harrison

Fig. 30. Guisborough Priory: waterleaf capital and keystone from a syncopated arcade

J. Phillips

Carlisle Cathedral

At Carlisle Cathedral, the presbytery was rebuilt in the early-13th century, before it was damaged by a great fire that meant the arcades were taken down and rebuilt.⁹⁹ The main arcades used a lot of salvaged voussoirs, but the elevation design appears to have been new. The aisle walls are very similar to those at Beverley and Fountains, with lancet windows flanked by blind arches and quadripartite vaults. The blind arches in particular have the same sort of curvature and appearance as those at Fountains, with the curious ‘hook’ shape (Fig. 31). Like Beverley, the aisle vault shafts start at ground level. Given these close similarities it is not unreasonable to suggest that the building which burned down had arcades of the Beverley type, even if it was more like Jervaulx in having no high vault.

Fig. 31. Carlisle Cathedral: the north aisle wall, showing ‘hook’ arches like those at Fountains Abbey, shown in Fig. 24

S. Harrison

Meaux Abbey

Wilson suggested a possible link between Beverley and Cistercian Meaux Abbey, a few miles east of Beverley, particularly because it was a daughter house of Fountains. Since then a more detailed look at the site and the small collection of loose architectural fragments at nearby Abbey Farm has shown this to be the case (Fig. 32).¹⁰⁰ Some of the fragments, such as pier shafts with fillets or round or keeled shafts, can be related to Ken Beaulah’s excavation notes and were certainly found in the abbey church area.¹⁰¹ Others are clearly stones collected from the abbey site and used as garden ornaments, including a structure known as the monument. This incorporates stones from octagonal pier bases and evidence of such piers was found in excavation by Beaulah. Put together with the large pier shafts we have evidence of very similar eight-shaft piers to those at Beverley (Fig. 11). One moulded capital from a corner position also survives and is very like those at Beverley, Fountains and Jervaulx (Fig. 33). Though slight, this evidence suggests that Wilson was correct and Meaux was another church built in this distinctive northern Gothic style.

Fig. 32. Meaux Abbey: mouldings and arches, showing similarities with Jervaulx, Beverley and Fountains

S. Harrison

Fig. 33. Meaux Abbey Farm: corner capital, similar to those at Jervaulx, Beverley and Fountains

S. Harrison

Revesby Abbey

Surviving evidence from Revesby, another Cistercian abbey in Lincolnshire, raises the same possibility. There are no standing buildings or ruins on the site, which has been reduced to earthworks and was partly excavated in the 19th century.¹⁰² It was a daughter house of Rievaulx so it had strong Yorkshire connections. Recent archaeological survey of the site, including geophysics, has clarified the plan of the church. The original 12th-century church was replaced in the 13th by a much larger building on the model of Jervaulx and Meaux with an aisled presbytery. Excavations in the 19th century established that it had octagonal piers in the nave and surviving pier fragments show large round shafts with broad fillets. Some mouldings have also survived and these are similar to those from other sites in the group. Capital fragments are also moulded like those at Beverley (Fig. 34).

Fig. 34. Reconstructed nave arcade at Revesby Abbey (Lincs.), based on excavated evidence for octagonal pier bases, pier fragments, capitals and arch soffit mouldings

S. Harrison

Bridlington Priory

Bridlington Priory was an Augustinian house and had a large church, of which only the nave and traces of the transept survive. To this physical evidence may be added some architectural fragments and a survey by Richard Pollard in 1537.¹⁰³ The nave has lancet windows and a large arcade with octagonal pier bases, piers with round and filleted shafts and moulded capitals. The triforium of the north arcade has a gallery type arrangement with round enclosing arches and four bays of trefoiled open arches within it. The clerestory windows are of geometrical tracery design. Stylistically, the arcade belongs to the third and fourth quarters of the 13th century. Pollard’s description and accounts of excavation of the presbytery and transept show that the presbytery was of six bays and had paired lancet windows in the aisles, like the choir of Carlisle Cathedral (Fig. 35). Two base sections most likely from the south transept are of octagonal form and similar to those in the nave arcades.

Fig. 35. Bridlington Priory: north nave arcade, showing similar piers and mouldings to Beverley Minster

S. Harrison

It seems clear that there was a major campaign to construct the presbytery and transept before work started on the nave. It seems likely, therefore, that the eastern parts of the building should be dated to the first and second quarters of the 13th century. A large springer and voussoir fit together and formed part of a clerestory front screen arcade on the Beverley pattern, and suggest the eastern parts had a similar clerestory with lancet windows but that the triforium was of the more conventional gallery type (Figs 36 and 20). This was replaced in the nave clerestory windows with tracery.

Fig. 36. Bridlington Priory: reconstruction of a clerestory screen arcade like those at Jervaulx, Beverley and Fountains, based on a loose springer and a voussoir in the north nave aisle

S. Harrison

Kirkham Priory

Kirkham Priory was another Augustinian house, with a magnificent seven-bay aisled presbytery built in the first half of the 13th century.¹⁰⁴ Mostly reduced to a low level, only one section of the east wall survives in sufficient detail to be able to reconstruct the whole elevation, helped by a drawing by Samuel Buck and an engraving by Thomas Atkinson.¹⁰⁵ The east wall had a lower range of heavily moulded lancet windows set between buttresses, which have detached nook shafts and foliate capitals. The capitals of the windows and the east respond were also foliate rather than moulded. Buttresses with nook shafts are a feature of Beverley (although there the shafts are coursed rather than detached), and have very similar foliate capitals to those at Kirkham. The upper buttress nook shaft capitals at Kirkham are moulded with round abaci that are also similar to many at Beverley. The remaining Kirkham east arcade respond has large round and filleted shafts standing on a round base. There is evidence to suggest that the respond was designed as a free-standing pier to support the arches of an eastern aisle or ambulatory, like that at Byland. The design was changed during construction and instead a full height façade was introduced, with two tiers of lancet windows surmounted by a large rose window, which was probably similar to those at Byland, Jervaulx, Beverley and Fountains.

Not enough of Kirkham’s presbytery arcade elevation survives to determine its form and very few architectural fragments remain at the site. One surviving detail is, however, very similar to the external elevations at Beverley: the curved rim of a small circular window that lit the north aisle roof space (Fig. 37). This unusual feature is paralleled at Beverley in the eastern and western transepts, where small circular windows fitted with a variety of plate tracery designs light the aisle roof spaces. Whether the arcade elevations followed the Beverley pattern is a moot point, given the paucity of evidence. More might emerge in the future in the form of antiquarian drawings or further loose stone fragments.

Fig. 37. Kirkham Priory: east front, showing capitals on the corners of the buttresses and the partial outline of circular window at top right to light the north presbytery aisle

J. Phillips

Other buildings in the region

Other major buildings in the northern region, such as Whitby, Rievaulx and the York Minster transepts, were built with some of the Beverley type of detailing, with octagonal pier bases and shafted piers with round and filleted shafts, moulded capitals and complex arcade mouldings. York also employed a lot of Purbeck marble. The syncopated design of the triforium stage at Beverley was not copied, nor the clerestory with its front screen arcade. Whitby, Rievaulx and York all had a triforium of paired and subdivided arches instead, with some bays with overall enclosing arches.¹⁰⁶ The Rievaulx clerestory can be seen as an advance on the Beverley clerestory design, which omitted the front screen arcade and opened up the rear wall by having a single open arch flanked by small narrow blind arches as decorative space fillers.

Beverley, Lincoln and York

Various claims have been made for the influence of Lincoln Cathedral on the design of Beverley and other northern churches. This was to some extent governed by the consensus regarding the date of Beverley, the use of Purbeck marble and decorative blind arcades of syncopated design at both churches. Yet, as we have seen, there was an alternative and earlier source for the latter at Guisborough, and possibly elsewhere in the region (Fig. 29). With a possible start date around 1189–90, work at Beverley was potentially slightly ahead of Lincoln, where construction seems to have begun after 1193/94, implying that architectural ideas could have travelled from Beverley to Lincoln, not vice versa.¹⁰⁷ The new dating of Beverley thus, theoretically, puts it before the start of Lincoln, but with a longer period of building, set back by a tower collapse. More work on Lincoln is perhaps now merited in light of the evidence presented here, but essentially we argue that Lincoln and Beverley should now be seen as contemporary constructions in very different styles that owe little to each other.

Though it was a comparatively rich institution, the resources Beverley could devote to a building fund were clearly far more modest than at Lincoln, yet the design and construction is of first-class quality. Hoey and Brieger noted the similarity of plan with double transepts to that of the 12th-century choir built by Archbishop Roger (1154–81) at York Minster.¹⁰⁸ That building was also the first in the north to make extensive use of Purbeck marble for detached shafts in the clerestory arcading and also some of the main arcade piers. Yet Beverley’s eastern termination was nothing like that planned at York, which may have resembled the tower chapel built at Lincoln, similar to the corona at Canterbury Cathedral.¹⁰⁹ In plan, Beverley’s east end instead resembles that of a Cistercian abbey church like Roche, from the 1170s, and shares with it the design of its elevation and overall massing, especially the aísleles presbytery termination.¹¹⁰

This review of northern great churches with similar design characteristics to Beverley Minster allows us to construct a timeline for some:

Jervaulx Abbey

Described in the foundation history as ‘a great church’ when started by Abbot John de Kingstan, the abbey church at Jervaulx was still unfinished in 1197.¹¹¹ Abbot John was the founding abbot, and had already retired by 1190, when he was called to account at the general chapter at Savigny, together with abbot Roger of Byland.¹¹² This surely puts construction of Jervaulx before the start of work on Beverley, and thus the earliest documented example in the group.

Beverley Minster

Burnt down in 1188, and the evidence above suggests that rebuilding began promptly, with the roofs of the eastern parts built sometime between 1188 and 1206.

Fountains Abbey

The new eastern arm and Chapel of the Nine Altars were started by Abbot John of York (1203–11) in 1203, continued by abbot John of Ely (1211–20), and finished by Abbot John of Kent (1220–47).¹¹³

Meaux Abbey

Started in 1207 and eventually consecrated in 1253. The history of construction at Meaux is complex and the church of 1207 is likely to have been a partial remodelling of an earlier building.¹¹⁴

When compared with these other buildings, Beverley stands out for its ambitious design, with double aisles to the main transept, which was of four bays with three eastern chapels in each arm. Beverley was also highly original in combining the double transept plan of York with the Cistercian transept and presbytery plan. Beverley’s proportions seem to owe much to the rebuilding over the Romanesque plan, which determined the general bay spacing of a perch (16’1” and 4.70 m). Equally, in the nave, the piers and buttresses were largely placed where their predecessors had stood. That means that the arcade bays and overall width are relatively narrow, and the aisle bays almost square in plan, and this endows the elevations with a great sense of verticality even though the height to the apex of the high vault is only about 66 ft (20.12 m).¹¹⁵ Combined with the elaborate piers and mouldings, this created a building so good that the builders of the new nave simply updated the design with Gothic tracery and kept to the Romanesque plan when they demolished and replaced the old nave. The canons were clearly determined to erect a building of sufficient quality to house the relics of St John.

ACKNOWLEDGEMENTS

We thank Ian Tyers, Alison Arnold and Robert Howard for their work on the dendrochronology, and Stefan King for his earlier work on the roofs. We also thank the Friends of Beverley Minster, who funded the dendrochronology survey over a number of years, and also made a generous grant towards the Open Access publication of this paper. Christopher Wilson kindly answered some queries and supplied copies of John Bilson’s notes on the main arcade mouldings. Janet Burton also kindly helped with documentary queries. We would also like to thank the two anonymous peer reviewers who contributed many useful comments that helped improve the text.

Notes

1 C. Norton, ‘New Light on the Gothic Minster’, Friends of Beverley Minster Annual Report, 73 (2009), 9–15.

2 J. Bilson, ‘Beverley Minster’, Architectural Review, 3 (1894–98), 197–204, 250–59.

3 L. Hoey, ‘Beverley Minster in its 13th-Century Context’, Journal of the Society of Architectural Historians, 43 (1984), 209–24; C. Wilson, ‘The Early Thirteenth-Century Architecture of Beverley Minster: Cathedral Splendours and Cistercian Austerities’, Thirteenth Century England III, ed. P. R. Coss (Woodbridge 1991), 181–95.

4 G. Coppack, Fountains Abbey (London 1993), 56 and 58, showing Harrison’s reconstruction drawings of the eastern rose window and arcade of the presbytery. Further work for the Cistercians in Yorkshire project produced a complete 3D model of the whole abbey church: S. Harrison, ‘Jervaulx Abbey and Gisborough Priory: how a study of architectural fragments can inform our understanding of these lost buildings’, Jaarboek Abdijmuseum ‘Ten Duinen 1138’ Novi Monasterii, 9 (2009), 58–73.

5 S. Harrison, ‘Jervaulx Abbey and its Relationship to Beverley Minster’, Friends of Beverley Minster Annual Report, 74 (2010), 8–12.

6 J. Phillips, ‘Of a Fair Uniforme Making’. The Building History of Beverley Minster 1188–1736 (Pickering 2016).

7 Ibid.

8 Corpus Christi College Cambridge, Parker Library, MS 298, fols 151b–52v.

9 Bede’s Ecclesiastical History of England, ed. and trans. A. M. Sellar (London 1907), Book V, ch. 2, 303.

10 R. Morris and E. Cambridge, ‘Beverley Minster Before the Early Thirteenth Century’, in Medieval Art and Architecture in the East Riding of Yorkshire, ed. C. Wilson, BAA Trans., ix (Leeds 1989), 9–32.

11 M. Johnson, ‘New Light on the Development of Beverley Minster’, JBAA, 166 (2013), 31–50. Harrison and Norton were consulted on the archaeological sequences and detail during the drafting of his paper.

12 The scale plan of Southwell was expanded slightly so that the naves of both buildings were the same width. Once that revised plan was overlaid on that of Beverley the crossing piers and bay spacings also aligned, which seems to indicate a close affinity in the overall scale of the buildings.

13 Wilson, ‘The Early Thirteenth-Century Architecture’, 183, n. 9.

14 C. Wilson, The Gothic Cathedral, The Architecture of the Great Church 1130–1530 (London 1990), 172–73. Wilson discusses the geometry of the east end, the planning of the bay dimensions and their basis in the Golden Section as allusions to the Trinity. Phillips, Building History, 78–80, discusses the setting out in relation to the use of the measurement of a perch of 16 ft 6 in.

15 Phillips, Building History, 107–08, illus. 4.2, shows what is likely to have been the abutment of the Romanesque nave roof against the western crossing arch. It shows a pitch around forty-five degrees and is set at a lower level than the present roof.

16 In fact, the whole plan of the building as well as the proportions of the elevations are based on multiples of this single measurement.

17 J. Bilson, ‘Norman Work in the Nave Triforium of Beverley Minster’, The Antiquary, 27 (1893), 18–23, at 21, shows an elevation and plan of the triforium that indicate the depth and size of the rear arches.

18 Wilson, ‘The Early Thirteenth-Century Architecture’, 185, n. 20.

19 Phillips, Building History, 80, illus. 3.32.

20 The Itinerary of John Leland In or About the Years 1535–1543, ed. L. Toulmin Smith (London 1907), 46.

21 A. F. Leach, Memorials of Beverley Minster: The Chapter Act Book of the Collegiate Church of St John of Beverley, A.D. 1286–1347, Surtees Society 98, 2 vols (Durham 1898–1903), II, 54; M. Woodworth, ‘Unnatural Ornament: Beverley Minster, Historical Consciousness and the Early English Style’, in Art and Nature: Studies in medieval art and architecture, ed. L. Cleaver and K. Gerry (London 2009), 23–38.

22 Roger of Howden, Chronica, ed. W. Stubbs (Rolls Series 1868–71), II, 354; Copies of Inscriptions in Beverley Minster, Hull History Centre, Papers of the Constable Maxwell Family, U/DDEV/5/1; W. Dugdale, The Visitation of the County of York, ed. R. Davies, Surtees Society, 36 (Durham 1859), 22.

23 S. E. Wilson, The Life and After-Life of St John of Beverley, The Evolution of the Cult of an Anglo-Saxon Saint (Aldershot and Burlington 2006), 13 and 217–18.

24 J. Raine, Historians of the Church of York and its archbishops, I (London 1879), 345–47.

25 Wilson, The Life, 217–18.

26 Wilson, ‘Early Thirteenth-Century Architecture’, 183, n. 13, suggested that it was a spire; L. F. Salzman, Building in England Down to 1540, 3rd edn (Oxford 1997), 26, suggested a ‘short conical spire’.

27 Notably, the four corner stair turrets in the eastern crossing are partly capped off with broken sections of octagonal Purbeck marble shafts, similar to those used in the clerestory arcades: Phillips, Building History, 76, illus. 3.31. The use of Purbeck marble must surely relate to the present building rather than an earlier Romanesque structure that was modified.

28 T. Rickman, Gothic Architecture—An Attempt to Discriminate the Styles of Architecture from the Conquest to the Reformation, 6th edn (Oxford 1862); T. Meadley, The History of Beverley Minster, from it Foundation: Including Its Antient Monuments etc. With a Description of St Mary’s Church, Collected from Antient Records (Hull 1804); George Poulson, BEVERLAC: or, The Antiquities and History of the Town of Beverley … etc. (Beverley and London 1829), 537; J. Britton, The Architectural Antiquities of Great Britain, 5 vols (London 1807–35), V, 22; J. Coltman and W. R. Johnson, A Short History of Beverley Minster etc. (Beverley 1835); J. L. Petit, ‘Remarks on Beverley Minster’, Memoirs Illustrative of the County and City of York Communicated to the Archaeological Institute of Great Britain and Ireland 1846 (York 1848), 7; J. J. Sheahan, An Antiquarian Ramble through the Town of Beverley (Beverley 1856), 15.

29 Petit, ‘Remarks’, 7.

30 In his footnote to Petit, Willis suggested that the quatrefoils represented part of an earlier east wall, destroyed when the east transept and east end were added to an existing four-bay presbytery. His comments suggest that he never actually saw the evidence but may have commented from a drawing. See also Wilson, ‘Early Thirteenth-Century Architecture’, 194, n. 55. The use of Purbeck marble shows it was always an internal feature as no Purbeck is used externally at Beverley.

31 Bilson, ‘Beverley Minster’, 202; Wilson, Gothic Cathedral, 170–74; P. Draper, The Formation of English Gothic: Architecture and Identity (New Haven and London 2006), 167–68; Hoey, ‘Beverley Minster’, 209–24.

32 Hoey, ‘Beverley Minster’, 212, n. 13. See also Wilson, ‘Early Thirteenth-Century Architecture’, 194 and nn. 29 and 57.

33 S. Harrison and M. Thurlby, ‘Observations on the Romanesque Crossing Tower, Transepts and Nave Aisles of Selby Abbey’, in Yorkshire Monasticism: Archaeology, Art and Architecture, ed. L. Hoey, BAA Trans., xvi (London 1995), 54–56. The medieval central tower of Beverley was taken down in 1721 and only one poor illustration of it is known. In three of the four corners are stair turrets rising from clerestory level. The spandrels of the crossing arches are decorated with quatrefoils and were clearly meant to be seen. In other words, the medieval central tower was also designed to have a lantern.

34 Bilson, ‘Beverley Minster’, 199–200.

35 O. Lehmann-Brockhaus, Lateinische schriftquellen zur Kunst in England, Wales, und Schottland vom Jahre 901 bis zum Jahre 1307, 5 vols (Munich 1955–60), I, 88, no. 336.

36 Calendar of the patent rolls preserved in the Public Record Office (London 1891–), Henry III, 3, 1232–47: Westminster, 14 May 1235, Protection with clause rogamus, for seven years from Whitsuntide, for the church of St John of Beverley, and the preachers thereof, gathering alms for the building of the said church (repeated at Woodstock, 15 June); Calendar of the Close Rolls, 47 vols (London 1900–63), Henry III, 7, 1251–53: 25 March 1252, ‘De Quercibus datis- Mandatumn est G. de Langel’, justicario foreste, quod in foresta regis de Shirewude faciat habere procutori ecclesiae Sancti Johnannis Beverlac’ xl. quercibus ad maeraemium fabricam ecclesiae supradicte, de dono regis. Teste ut supra. Per regem (Teste rege apud Westmonasterium XXV die Marcii)’.

37 Beverley Cartulary: London, British Library, Add. MS 61901, fol. 83.

38 Norton, ‘New Light’, 11.

39 Wilson, ‘Early Thirteenth-Century Architecture’, 194–95 and nn. 57–59.

40 Norton, ‘New Light’, 12–13, also interpreted a blocked opening in the south aisle of the presbytery that would look into the eastern transept as a former window in the first east end. Along with another on the north side and two others at the north and south ends of the eastern transept chapels, all looking into the east end, it is instead an opening to admit light and air into the aisle roof space. These blocked openings can clearly be seen inside the church.

41 Hoey, ‘Beverley Minster’, 212–13.

42 These have been catalogued as far as possible and the distribution of the marks can be interrogated in conjunction with other detail to obtain indications of the building sequence.

43 Phillips, Building History, 279.

44 Hoey, ‘Beverley Minster’, 214.

45 J. Bilson, ‘On the Discovery of Some Remains of the Chapter-House of Beverley Minster’, Archaeologia, 54 (1895), 425–32. Unfortunately, Bilson did not include any masons’ marks.

46 Phillips, Building History, 85–88, illustrates the buttress capital variations and where they occur.

47 Wilson, ‘Early Thirteenth-Century Gothic Architecture’, 194–95, following P. Brieger, English Art 1216–1307 (Oxford 1957), 50–51, also suggested that the eastern crossing piers had been rebuilt by cutting back and refacing the originally fully shafted piers.

48 Assessing the full extent of their area of construction is hampered by the fact the north-eastern bay of the Lady Chapel was completely reconstructed in the 18th century, and that all the lancet windows in the east wall of the Lady Chapel, the south-east transept, and those in the south choir aisle were replaced with Perpendicular tracery windows in the early part of the 15th century.

49 A. F. Leach, Memorials of Beverley Minster: The Chapter Act Book of the Collegiate Church of St John of Beverley, A.D. 1286–1347, Surtees Society 98, 2 vols (Durham 1898–1903), II, 299.

50 Ibid., 224–25.

51 Ibid., 299.

52 R. K. Morris, ‘Thomas of Witney at Exeter, Winchester and Wells’, in Medieval Art and Architecture at Exeter Cathedral, ed. F. Kelly, BAA Trans., xi (London 1991), 79, n. 14, established that the shafts were made from the same template.

53 N. Dawton, ‘The Percy Tomb at Beverley Minster: the Style of the Sculpture’, in Studies in Medieval Sculpture, ed. F. H. Thompson (London 1983), 122–50, discusses the likely date and the identification of the occupant of the tomb in detail.

54 Ibid., 124.

55 Leach, Memorials, 299.

56 Ibid.

57 W. H. Thompson, ‘The Black Death in Yorkshire (1349)’, The Antiquary, 37 (1901), 134–37.

58 There was clearly a partition which separated the east end of the building from the nave, just west of St John’s tomb. It was probably a wooden wall keyed into the main fabric. There is evidence for its removal in the spandrels of the second and third bays of the nave, where new ashlar was inserted into the walls. This can be dated to the third phase of the nave building because it shares masons’ marks with the two west towers.

59 Romanesque chevron-decorated voussoirs from several arches were used as relieving arches in the back of the triforium in this area: Bilson, ‘Norman work’.

60 A. C. Armstrong and D. Cant, ‘Carpenters of the nave roof’, in Who Built Beverley Minster, ed. P. S. Barnwell and A. Pacey (Reading 2004), 59–80.

61 A. Arnold and R. Howard, ‘Reused Timbers of the Nave Roof, Beverley Minster. Tree Ring Analysis of Timbers’ (hereafter NTRDL) (unpublished report, John Phillips Collection archive, East Riding of Yorkshire Archives, Beverley, September 2021).

62 I. Tyers, ‘The tree-ring analysis of timbers from The Minster, Beverley’ (unpublished report, John Phillips Collection archive, East Riding of Yorkshire Archives, Beverley, July 2012).

63 NTRDL.

64 Phillips, Building History, 111, illus. 4.7.

65 London, British Library, Lansdowne MS 896, part III, fol. 40v.

66 R. G. Davies, ‘Alexander Neville, Archbishop of York 1374–1388’, Yorkshire Archaeological Journal, 47 (1975), 87–101.

67 Calendar of the Close Rolls, 47 vols (London 1900–63), 1385–89, Richard II, 3, 413: ‘June 10. To John Godard escheator in Yorkshire. Order to deliver to the Westminster, chapters of St. Peter York and St. John Beverley, their attorneys or deputies, for the works of those churches divers oaks for timber in his woods of Skaholme and “Southburdonwode” which were freely given them by Alexander late archbishop of York, and were cut down before the judgment of forfeiture rendered against him, his forfeiture or the seizure of his goods and chattels notwithstanding. By C.’.

68 NTRDL.

69 A. F. Leach, Beverley Town Documents, Seldon Society, 14 (London 1900), 24.

70 Salzman, Building in England, 229–32.

71 N. Lloyd, A History of English Brickwork (Woodbridge 1925), 96, with a table of measurements.

72 F. W. Brooks, ‘A Medieval Brickyard at Hull’, JBAA, 4 (1939), 151–74.

73 A. F. Leach, ‘The Building of Beverley Bar’, Transactions of the East Riding Antiquarian Society, 4 (1896), 26–37; John Bilson, ‘The North Bar, Beverley’, Transactions of the East Riding Antiquarian Society, 4 (1896), 37–49.

74 Beverley, East Riding of Yorkshire Archives, BCIV/14//1: ‘Beverley Minster Minutes and Accounts 1718–1731’.

75 Beverley, East Riding of Yorkshire Archives, I DDBC/3/104, 1750: ‘Cash paid to workmen upon stripping and Covering, viz., the North End of the Cross Isle being the High Roof from the Dome to the North Gable End’.

76 Beverley, East Riding of Yorkshire Archives, BCIV/14/1, ‘Beverley Minster Minutes and Accounts 1718 to 1731’.

77 Phillips, Building History, 210–27.

78 Ibid., 217–18, illus 9.11 and 9.12.

79 Another aspect of the medieval appearance of the church was observed from scaffolding on the south-east transept. The corbels and cornice beneath the later parapets show extensive traces of limewash in areas that have been protected from the weather. It seems likely that the whole building had been limewashed externally in the medieval period.

80 The amount of debris lying on top of the high vaults east of the central tower testifies to the reconstruction of the roofs and also the defrassing of timbers that had been attacked by beetles. It contains wood shavings and a lot of sawdust reduced to a fine grey overall consistency with time. Recent releading and replacement of timbers in the nave roof generated a considerable amount of similar debris which the contractor had to remove.

81 Adapted from explanatory text by NTRDL.

82 Given the time sequence we have demonstrated between the western choir, south transept, north transept and nave, it seems impossible that these roofs once covered the refurbished burnt-out Romanesque building following the 1188 fire and were then transferred piecemeal to cover the new building.

83 R. R. Laxton, C. D. Litton and R. E. Howard, Timber: Dendrochronology of Roof Timbers at Lincoln Cathedral, English Heritage Research Transactions, 7 (London 2001), figs 25 and 26.

84 The stiff-leaf capitals bear a remarkable resemblance to those of the chapter-house capitals at Jervaulx Abbey, for which see below.

85 Bilson, ‘Chapter-House of Beverley Minster’, 425–32.

86 S. Harrison, ‘The Architecture of Byland Abbey’ (unpublished MA dissertation, University of York, 1988); Byland Abbey, English Heritage guidebook (London 1990); Wilson, ‘The Early Thirteenth-Century Architecture’, 191–92.

87 W. H. St John Hope and H. Brakspear, ‘Jervaulx Abbey’, Yorkshire Archaeological Journal, 21 (1911), 303–44, at 324; S. Harrison and D. Robinson, ‘Cistercian Cloisters in England and Wales’, JBAA, 159 (2006), 131–207, at 170–72 and 189–92.

88 Harrison, ‘The Architecture of Byland Abbey’, 104–07. The aisle and west responds of the nave at Byland show multiple capitals set under a single circular rim and several sections of abacus from such piers were found in excavation. One half-pier capital in the garden at Myton Hall is also of this type.

89 E. Sharpe, Architectural Parallels (London 1848).

90 Those at Fountains were smashed out at the Dissolution to recover the iron and leadwork supporting the arcading, so none remain intact. Loose stones from the arcades and in situ springers show the design.

91 It may be the case that rather than five-arch clerestory screen arcades there were three-arched ones in the wider choir and nave bays and three-arch ones with narrower flanking arches to the central opening in the narrower transept bays. All three possible versions have been shown in the reconstructions.

92 G. Coppack, Fountains Abbey (Stroud 2009), 90–91.

93 Hedon Church uses some similar details but has no marble.

94 W. H. St John Hope, ‘Fountains Abbey’, Yorkshire Archaeological Journal 15 (1900), 269–402.

95 J. A. Reeve, A Monograph on the Abbey of St Mary at Fountains (London 1892).

96 The high vaults and aisle vaults at Beverley nonetheless spread the walls of the presbytery and most notably the transepts and nave considerably. The transept chapels on the south side have a notable outward lean and the clerestory wall passages bend outwards in the middle of the transept and nave arcades.

97 Similar mask corbels survive loose at Jervaulx and must come from a similar eaves cornice.

98 S. Harrison, ‘Stonework from the Excavations’ and ‘Architectural stonework from earlier clearance’, in D. Heslop, ‘Excavation within the church at the Augustinian priory of Gisborough, Cleveland 1985–6’, Yorkshire Archaeological Journal, 67 (1995), 80–86 and 106–17.

99 J. S. Alexander, ‘The Construction of the Gothic Choir of Carlisle Cathedral, and the Evidence of the Masons’ Marks’, in Carlisle and Cumbria, Roman and Medieval Architecture, Art and Archaeology, ed. M. McCarthy and D. Weston, BAA Trans., xxvii (London 2004), 106–26.

100 G. Coppack and S. Harrison, ‘New Thoughts on an Old Chronology: Meaux Abbey, East Riding of Yorkshire (England) in a new light’, Citeaux Commentarii Cistercienses, 67/iii–iv (2016), 361–70.

101 Copies of Ken Beaulah’s notes about his excavations are kept at the English Heritage Archaeology Store at Helmsley.

102 G. Coppack and S. Harrison, ‘Revesby Abbey’, forthcoming.

103 J. Caley, ‘Copy of a Survey of the Priory of Bridlington, in Yorkshire, taken about the 32d year of Henry VIII’, Archaeologia, 12 (1821), 270–75; G. Coppack, S. Harrison and J. Phillips, ‘Reconstructing Bridlington Priory: Richard Pollard’s Suppression Survey and the Evidence of Antiquarian and Current Research’, forthcoming.

104 G. Coppack, S. Harrison and C. Hayfield, ‘Kirkham Priory: The Architecture and Archaeology of an Augustinian House’, JBAA, 148 (1995), 55–136.

105 S. Harrison, Kirkham Priory, English Heritage Guidebook (London 1999), 24–25.

106 Obvious Lincoln-derived features at York Minster include the five-part vaults with major and minor external buttresses in the transept aisles. Yet the double aisles to the York transepts must derive from Byland/Beverley and the abandoned high vault originally designed to cover the south transept owed much to that built over the presbytery at Fountains. Although it is no longer obvious, the north choir aisle at Rievaulx also had five-part vaults and most likely alternating major and minor aisle wall buttresses: P. Fergusson, G. Coppack, S. Harrison and M. Carter, Rievaulx Abbey (London 2016), plans. The evidence for the five-part vaults has never been published but exists in the form of a vault boss in the English Heritage Helmsley Archaeology Store.

107 Norton, ‘New Light’, 14–15; R. Stalley, ‘Lapides Reclamabunt: Art and Engineering at Lincoln Cathedral in the Thirteenth Century’, Antiquaries Journal, 86 (2006), 131–37, at 131, n. 1, neatly corrected the misconception that the building works commenced in 1192. For the early dendrochronological dating of the Lincoln Cathedral roofs, see Laxton, Litton and Howard, Timber: Dendrochronology of Roof Timbers at Lincoln Cathedral, 40–46.

108 Hoey, ‘New Light’, 216; Brieger, English Art, 48–51.

109 S. Harrison, ‘The Original Plan of the east End of St Hugh’s Choir at Lincoln Cathedral Reconsidered in the Light of New Evidence’, JBAA, 169 (2016), 1–36; S. Harrison and C. Norton, York Minster: an Illustrated Architectural History 627–c.1500, (York 2015), 20 and 29.

110 Peter Fergusson and Stuart Harrison, Roche Abbey (London 2013), 7 and plan.

111 J. Burton, The Foundation Histories of the Abbeys of Byland and Jervaulx, Borthwick Texts and Studies, 35 (York 2006), 56.

112 Twelfth-Century Statutes from the Cistercian General Chapter, ed. C. Waddell (Brecht 2002), 209 (1190/53).

113 A. W. Oxford, The Ruins of Fountains Abbey (London 1910), 229–30, gives a complete English translation of the Latin text of the Fountains chronicle.

114 G. Coppack and S. Harrison, ‘New Thoughts’, 361–70.

115 Bilson, ‘Beverley Minster’, 200–01, discusses the plan proportions in some detail, as does Wilson, Gothic Cathedral, 172–73.