The three towers of the church, which were the last additions to the 12th century building, are very similar in character, allowing for the difference in size between the central and the two western towers. In all three cases the stage below the belfry is surrounded by a wall arcade, which is continued round the buttresses, and is composed, in the central and north-west towers, of intersecting round arches, in the south-west tower, of pointed arches, which seems to indicate that this tower was the last of the three to be finished, as the central tower was certainly the first.1 The uppermost stages are also arcaded with round arches, continuous in the central tower, but broken by the flat angle buttresses of the western towers. The three middle arches in the central tower are pierced to form belfry window-openings. In the more slender western towers the middle arch in each face is pierced. The openings in all three towers were each sub-divided into two lights by a shaft. The multiplication of shafting in the jambs of these tower windows is a noticeable feature.2 All the towers must originally have been finished with pyramidal caps or low spires of timber covered with lead, to match the high pitch of the church roofs. In 1711 the church was much injured by fire: up to that time spires of uncertain date had crowned the western towers, but the capping of the central tower appears to have been superseded by a flat roof, probably in the 15th century, when the pitch of the roofs of nave, quire, and transept were lowered. New spires were added after 1711 to the western towers, and remained there for nearly a hundred years, until they were removed, on a hint of danger to the northern tower. The present spires were added by Mr. Ewan Christian on the lines of their predecessors, and fairly represent the intention of the original builders. The central tower remains as it was restored after 1711: the parapet is said to have been made out of fragments taken from the transept gables at the time when the roofs of the transepts were lowered, while the circular turrets with pinnacles, at the angles of the tower, came from the same portion of the fabric, where they originally stood on the kneelers at the foot of the gables.3
On the west face of the north tower are the marks of the roof of a building—probably an outhouse—which formed no part of the medieval minster. Its character and the date of its removal are difficult to ascertain. It covered a three-light window which had been inserted in this wall4 in the 14th century. There was a corresponding window in the south tower, which was removed in more recent times, and is now in a garden of the vicars' court.5 These windows must have given much needed light to the dark interiors of the ground-floors of the towers, and to the chapels which probably occupied the western part of the aisles beyond. Their place is taken by modern "Norman" window openings on what is, no doubt, the pattern of the openings designed by the 12th century builders for this position.
To recapitulate what has been said of the Norman minster, its plan was steadily continued upon lines laid down in the first quarter of the 12th century; but the progress of the work was gradual, and it was not wholly executed until about half-a-century after the death of Thomas II. No further work, so far as is known, was taken in hand uutil the second quarter of the 13th century. An indulgence of twenty days for contributions to the fabric was published by archbishop Walter Gray in 1234, which gives us an unquestionable date for the building of the quire.6 This beautiful addition to the church must have been well aided by the funds of the faithful, as the uniformity of the design shows that it was built without intermission, and was probably ready for use by 1240. The Norman presbytery and aisles were destroyed, the new building being nearly three times their length.
It need hardly be said that this enlargement of the eastern arm of the building was in keeping with the general tendency of the age, in parish churches as well as in the churches of monasteries and collegiate bodies. In 1232 archbishop Gray had issued a similar indulgence with regard to the rebuilding of the eastern portion of Beverley minster.7 The quires of York and Ripon had already been rebuilt on an enlarged scale in the time of archbishop Roger. Norman churches with their short presbyteries, and a mere ritual quire which encroached upon the nave, were by this time too cramped for the services of the church. In many instances, the devotion paid to the shrine of a local saint was a powerful inducement towards the enlargement of the eastern portion of the church—for example, at Canterbury (1175-1184), at Ely (1235-1252), at Durham (1242-1280), and at Lincoln (1255-1280). Thus at Ripon the shrine of St. Wilfrid probably occupied the place of honour behind the high altar.8 At Beverley the shrine of St. John was made early in the 14th century, and the new high altar was dedicated in his honour in 1308, the church itself being dedicated to St. John the Evangelist.9 The quire of Beverley had been finished, however, long before this date; and, both at Beverley and at Southwell, where there was no great shrine to attract pilgrims, the first object of the new eastern arm was to provide more room for additional altars, and a special chapel for the services of the quire. Between 1241 and 1280, no less than eight chantry services were founded at six altars, one of which is known to have been in the quire, and two in the nave, of Southwell minster.10 A similar growth of chantries at York led, not to the immediate re-building of the quire, but to the construction of new aisled transepts on a vast scale, during the pontificate of archbishop Gray.11 The extension of the western transepts at Lincoln was due to the same reason. At Southwell, however, the transepts were left without re-building, save for one alteration to be noted presently. The plan of the new eastern arm consisted of eight bays, of which the six western were provided with aisles. Two small transeptal chapels, nearly square on plan, projected from the second bay from the east end of the aisles. The east end of the church and of the aisles, following what had now become the habitual English use, were rectangular. On plan, the arrangement bears an apparent resemblance to some of the secular cathedrals of the west of England, especially to Exeter and Wells. At Wells, it is true, the plan ends in a polygonal chapel; but the transeptal chapels, mere excre-sences from the aisles, occur in the same position as, and are similar in plan to those at Southwell. The elevation, however, shows the real dissimilarity in character between the plans. At Exeter and Wells the aisles are returned to form an ambulatory round the high altar, and the aisleless extension to the east is a chapel, the roof of which is no higher, or only slightly higher than that of the aisles and ambulatory; while the clerestory and vault of the quire stop above the high altar. At Southwell, on the other hand, the high vault of the quire is continued to the extreme east end of the church. This was the case at the sister church at Beverley; but at Beverley, where the plan was probably influenced to some extent by St. Hugh's quire at Lincoln, eastern transepts were provided, which were the whole height of the main building, and were furnished with an eastern aisle, and the aisleless extension at the east end was only a projection of one bay. It is evident that, both at Southwell and Beverley, the plan was derived from one which was much used in the churches of Augustinian canons, and was probably in its origin a development from the unaisled rectangular presbytery of Cistercian churches. A long rectangular presbytery was flanked by shorter chapels, which either, as at Haughmond,12 were divided from it by solid walls, or, as at St. Frideswide's at Oxford, at Repton, and at Lanercost, formed aisles divided from the presbytery by arcades. The church of secular canons at Wimborne has arcades of three bays, and an aisleless eastern bay13: this is almost exactly the plan of the eastern arm of the Augustinian church at Oxford, where, however, the arcades are of four bays. In the noble 14th century Augustinian quire at Bristol, there are arcades of five bays, and, as at Southwell, an eastern extension of two. It may be remarked, however, that the Southwell plan, apart from any Augustinian influence, was, after all, merely an extended edition of the original plan of the church. Eight bays were taken for its length instead of three. The rectangular presbytery of two bays was translated into the current terms of architectural expression: the new aisles were built without apses, leaving the presbytery walls free to be pierced with windows in the lower storey of their western bays, and additional room for altars was provided by the new northern and southern transeptal chapels.
The levels of the quire have been much altered,14 and the present position of the high altar against the east wall was certainly not the medieval arrangement. In all our larger churches, where the roof of the quire was continued at one height to the extreme east end, whether they were aisled throughout their whole length or not, a clear space for processions was provided behind the high altar. There is no reason to suppose that any departure from the normal arrangement was made at Southwell; and the medieval disposition of the plan can be restored from that at Beverley, where it has remained until our own day, to say nothing of that at Lincoln and York. The three bays immediately east of the tower arch and its abutments were occupied by the stalls; for it must be remembered that provision had to be made, not only for the canons, of whom very few were resident at a time, but for their vicars choral, for chantry priests, choristers, deacons, incense bearers, and others, who were on the foundation, and were expected to attend the quire services.15 The fourth bay was clear of stalls: in the screen on either side were the ostia chori, or upper doorways of the quire. In the fifth bay, the ritual chancel, was the high altar, with a screen at the back. The sixth bay was left clear as an ambulatory for processions—and how necessary this was would appear, not merely at that important function, the Sunday procession before high mass, when each altar was visited in turn and sprinkled with holy water, but at the great Whitsuntide procession, when free circulation was indispensable for the vast crowd which came from every part of Nottinghamshire.16 The aisleless presbytery formed a separate eastern chapel, probably screened from the ambulatory There was an altar at the end of each aisle, and in each of the transeptal chapels, so that the eastern arm contained at least five altars in addition to the high altar.
Although no structural alteration was made to the central tower and crossing, and the present pulpitum or stone screen is nearly a century later than the fabric of the quire, a pulpitum must have been made when the quire was built. The doorway in the south transept was the chief entrance for members of the foundation.17 In the western arch of the tower was, no doubt, another screen, above which was the rood-beam ; and against its western face was the parochial altar, which we know to have been dedicated to St. Vincent.18 The nave aisles would also be screened off from the transepts, with doors through which the chantry priests could go to their masses at the altars of the western part of the church. The whole nave thus became the parish church of Southwell: the crossing and transepts formed the vestibule of the quire ; while the eastern arm was the church of the college. No passion for a vista or objection to screens as interfering with the view ever disturbed the medieval mind.
Choir—Southwell
The progress of the work of the quire can be traced with approximate certainty. The first necessity being to complete the ritual quire and get the high altar into position, the arcades were built first, beginning at the east end, and it would seem that the side walls of the eastern chapel were raised only so far as they would afford abutment to the new work.19 The three eastern arches on the south side of the nave were the earliest work completed: they are without dogtooth ornament, and the capitals and bases of the piers are of a less advanced type than those to the west.20 The piers of the northern arcade were also begun, but were finished less rapidly21; and, before these were completed, the aisle walls and transeptal chapels seem to have been built. The arcades were then finished, the aisles were vaulted, and the upper stage of the quire, which shows a distinct advance in richness of ornament, was executed. The quire was now ready for service, and was doubtless covered by a temporary wooden roof, while the open space, a bay east of the high altar, was boarded up. The side walls of the eastern chapel were now raised to their full height, and the east wall was finished. This would probably be left to the last, because a clear space at the east end gave the builders the opportunity of carrying materials for their work into the new building without hindrance. When this was achieved the high vault was taken in hand, and the finishing touch was given to the whole design.
One very significant feature of the progress of the work is the fact that the third arch of the south arcade from the east is lower and slightly wider than the rest. This arch is immediately east of the site of the south-east angle of the 12th century chancel, and the capital and base of the pier from which it springs on the west are of the more elaborate, while those on the east are of the plainer of the two types which are seen in this part of the building. It is probable that the old chancel and quire were left standing until the new south arcade had been brought down to this point. A slight pause was then made in the work, while the older building was cleared away; and it may be that, either by a slight miscalculation of distance, due to the old masonry which was encumbering the site, or by deliberate design, for the same reason, the site of the new pier on the west was shifted a little, with the consequence that the proportions of the arch were altered. The arch, however, is of the earlier and plainer type, and it may therefore have been turned first and left to rest upon the older wall to the west, until the new pier was built for it. In any case the presence of the older chancel until some time after the beginning of the new quire accounts for the discrepancy. A small blank space of wall was naturally left between the crown of this arch and the triforium string; and this was covered by a carved trefoil ornament of rather unusual design, to make the difference in height less noticeable. No such discrepancy occurs in the north arcade, which was not finished until the whole of the older masonry had disappeared.
The clusters of eight filleted shafts, which form the columns of the quire arcade, are set closely together, after the fashion of the north of England and the north-east Midlands, in which, for the main piers, marble was dispensed with, or, as at Beverley, used sparingly. The Southwell columns, like those at Beverley, have plain moulded capitals and bases, but are infinitely less graceful in proportion and design: they bear a strong likeness to the earlier columns of the parish church of Great Grimsby, and to those of the contemporary work at Thurgarton priory church—a church whose prior, though not a canon of the church, had a stall in quire at Southwell. The capitals of the respond and two earlier piers on the south have no ornament: in all the rest, with the exception of the more elaborately treated capital of the north eastern respond, the two groups of mouldings are divided by a band of nail-head. The character of the bases has been noted already: the later bases are composed of convex rolls without hollow mouldings. The arches of the arcade are of the equilateral pointed type prevalent after the first quarter of the 13th century. Their mouldings consist of an elaborate succession of rolls, those of the outer order filleted on the face, while those of the inner have lateral fillets only. A band of dog-tooth ornament divides the two orders in all the arches, except in the three eastern arches of the south arcade.22 The tendency of the mouldings of the arches, capitals, and bases, is to group together convex forms, and leave little space for the hollows between. This tendency is also very strongly marked in the grouping of the shafts of the piers, which is the same in every case throughout the quire.
The design of the upper part of the quire walls, above the main arcade, although by no means unique, is at any rate curious, and was not at this time characteristic of the architectural district to which Southwell belonged. This arrangement is obtained by combining the clerestory and triforium stages internally into one story. The clerestory is composed of a pair of small lancet-shaped openings in each bay, and below these is the usual passage between the vaulting of the aisle and its outer roof. But there is no clerestory passage : on the inner side of the wall on each bay are a pair of tall lancet arches, equal in height to the clerestory and triforium stages taken together; and, at the back of the passage at the triforium level, the wall is pierced in each bay by two plain pointed arches which give access to the passage between the roofs. This internal combination of the two upper stories of a building of moderate elevation into one was not uncommon in the west of England, where, from the end of the 12th century onwards, there was a tendency to suppress the triforium stage.23 The reason of this has been stated as arising from a design to give more height to the main arcade and the aisle walls, in which larger windows could thus be made. The buildings in which it is most noticeable, such as the quire of Pershore abbey, which is contemporary with that of Southwell,24 have arcades which are not remarkable for height; and it seems probable that the design was first adopted in buildings where, from the necessity of adapting new work to the scale of old, a full-sized triforium was out of the question. The quire at Southwell, if built higher than the nave and transepts, would have interfered with the eastern elevation of the central tower, no alteration in which was intended. The builders were, however, at liberty to give more height to their nave arcades and aisle walls, provided that they kept their roof sufficiently low. In this case, a triforium on the scale of that in the nave, which had reduced the clerestory to a minimum, would have blotted out the clerestory altogether; while, if the clerestory was to be given its full share in the design, a separate triforium stage would have been inconveniently cramped.25 The expedient which we see was therefore adopted; and the suggestion may be hazarded that the master mason in whose mind it took root may have come from the archbishop of York's estates in north Gloucestershire,26 and may have been acquainted with the somewhat earlier and, on the whole, more successful design at Pershore. At Southwell the two stages of the interior elevation are too nearly alike in height, with the consequence that an effect of lowness is produced. This might have been diminished if, as at Pershore, the upper arcade in each bay had been made with three arches, the central one higher than those at the sides. As it is, there are only two, which are necessarily of the same elevation, and therefore slightly monotonous in effect. Each pair of arches springs from clustered responds, and is divided by a slender pier of four clustered shafts. The capitals of these are in some cases carved with foliage, in other cases plain,27 but a band of dog-tooth occupies the hollow moulding between the orders of each arch, and is continued through the shafting of the jambs.
(1) The detail of the central tower is, on the whole, remarkably plain. Hollar's engraving in Thoroton's Antiquities of Nottinghamshire, 1677, shows intersecting arcading in both the western towers; but the details are untrustworthy.
(2) The stages of the western towers immediately below the continuous wall arcades also have arcades of three arches in each face, with a plain central opening. These arches are surrounded by chevron ornament, while the belfry arches have a series of pellets. Both towers were probably constructed as high as the pitch of the nave roof, and finished somewhat later by the two additional stages.
(3) Their general likeness to the flanking turrets of the north porch will be recognised. One of them is illustrated in Mr. Arthur Dimock's Cathedral Church of Southwell, 1898, p. 27. They have the interlacing ornament characteristic of the transept gables. The gables were restored to their original pitch by Mr. Christian. The pitch to which they had been reduced is marked by the difference in the colour of the old and the modern stonework.
(4) The engraving in Dickinson's History of Southwell, 1801, p. 46, shows no building against the wall of the tower. The three-light window is clearly shown: it was removed before 1850, and the present window inserted.
(5) Dimock, op. cit., p. 38. Hollar's engraving in Thoroton, ut sup., shows no window at this point in the south tower.
(6) See York Epis. Reg. Gray (Surt. Soc.), p. 64 (no. ccclxxvii). The date is 23 November, anno 19, i.e. 1234. The date is sometimes given as 1233 or 1235, but Gray's pontifical year was reckoned from 27 March, 1216 (Hist. Ch. York, ii., 402). The indulgence was for thirty days: we learn from it that a papal indulgence for twenty days had also been issued. An indulgence for work at Ripon—probably the west front of the church—was issued on 27 November in similar terms (Reg. Gray, p. 65).
(7) Reg. Gray, p. 64.
(8) See J. T. Fowler, F.S.A., Memorials of Ripon, (Surt. Soc.), vol. iii., p. xx.
(9) 21 June, 1308: Beverley Chapter Act Book, (Surt. Soc.), i., 218-9.
(10) See the article on the Southwell chantries which follows.
(11) An approximate date for the construction of the south transept at York is given by archbishop Gray's ordination of a chantry of three chaplains at St. Michael's altar, 22 March, 1240-1. (Hist. Ch. York, iii., 153-5).
(12) The influence of the Cistercian plan on that of the churches of canons regular is very noticeable in the plans of both the earlier and later church at Haughmond (see Archaeol. Journal, lxvi., 281-310), and of the church of Lilleshall abbey, also in Shropshire.
(13) The approximate date for the quire of Wimborne is 1220, some fifteen years earlier than the work at Southwell.
(14) i.e. east of the fourth bay, counting from the east of the aisled portion. The floor of the aisleless chapel at the east end seems to have been always at much its present level.
(15) Some information about these will be found in the certificates printed in the following article.
(16) In York Epis. Reg, Zouche, f. 228, there is a most vivid and interesting description of a forcible attempt made at Whitsuntide, 1347, and apparently on previous occasions, by certain "sons of iniquity" who, acting bestiali rabie, hindered the procession and carried off the offerings. On 11 May, 1348, Zouche issued sentence of excommunication, to prevent the recurrence of this digraceful scene.
(17) There is a blocked doorway, inserted in the 14th century, in the south aisle of the quire, a bay west of the ostium chori on that side. A round-headed blocked doorway in the north aisle, east of the present entrance to the chapter-house, possibly gave access to an earlier chapterhouse now destroyed.
(18) See evidence as to the dedication of altars in the following article.
(19) This may be noticed in the detail of the western bay on either side in this part of the church. On the south side the western jamb-shaft and the cluster which separates the two lights of the double lancet window opening in the lower stage have plainly moulded capitals, while throughout the rest of the eastern chapel the capitals are carved with foliage. On this side, too, the corbel which carries the western vaulting-shaft is plainly moulded. On the north the corresponding corbel, like the corbels which carry the vaulting-ribs of the aisles, has a row of nail-head; and the corresponding capitals are foliated. This may be taken as evidence that this part of the south wall is earlier than the north, and was begun when the south arcade was taken in hand.
(20) A slight change in design is noticeable in the base of the third pier from the east.
(21) The water-mouldings of the four eastern bases (including that of the respond) on this side are better than those of the two eastern bases on the south; while the third and fourth bases on the south show two successive departures from this feature of the design. The northern bases were probably finished and laid in place first, but the piers and arches were not taken in hand at once.
(22) Another point which differentiates the earlier from the later arches is that these three arches have, on the side next the south aisle, hood-mouldings ending in small carved heads. These are omitted from the aisleward faces of all the other arches of the quire.
(23) Instances from the west of England are given by F. Bond, Gothic Architecture in England, 1905, p. 536.
(24) The design at Pershore is said to have been executed 1225-39.
(25) It hardly need be said that the triforium stage must be taken into account, so long as the roofs above the aisles retain their high pitch.
(26) Aldred, the last Saxon archbishop of York, on resigning the bishopric of Worcester, which he held for some time with the archbishopric, kept back twelve manors for himself (Hist. Ch. York, ii. 348). Churchdown, between Gloucester and Cheltenham, and Oddington, near Stow-on-the-Wold, were favourite residences of the medieval archbishops. All this district was within the architectural influence of the great abbeys of the lower valley of the Severn.
(27) There is some variety of treatment in this respect in the three eastern bays; but in the western bays carved foliage is general. The westernmost bay, however, is treated more plainly than the rest.