Part 2

II. TRANSITIONAL.—About 1145 innovations began to be made in Anglo-Norman Romanesque. Among other things the pointed arch was introduced; at first chiefly in arches supporting central towers or vaults. In 1155 the founder received a grant of the forfeited estates of Roger de Berkeley; and, owing to this vast accession of wealth, was able to finish his Norman and Transitional work with exceptional richness. (1) In the _Vestibule_ to the Chapter-house, the bays being oblong, pointed arches were used on the short sides of each bay, so that their crowns might rise to the same height as those of the semicircular arches on the longer sides, as the builders thought was demanded by the requirements of vaulting. In the same way, beneath the central towers of Oxford Cathedral and Bolton Priory, pointed arches were built over the narrow transepts; while over the broader nave and choir the semicircular arch was employed. (2) The _Chapter-house_, as in most monastic establishments, _e.g._, Gloucester, Canterbury, Oxford, Chester, was oblong; the more beautiful polygonal form being more in favour in cathedrals served by Secular Canons. It was originally 71 by 25 feet. The vast scale of these conventual arrangements, here and elsewhere, is exceedingly striking. The monks of St. Albans thought that their church required a nave 300 feet long; the six canons of Bristol must have been lost in a Chapter-house 71 feet long. The recessed arcades round the walls here and at Worcester are interesting; simple as they are, they are the germs of the beautiful arcades of canopied niches round the Chapter-houses of York and Wells and the Lady Chapel of Ely. Here is preserved a fine piece of archaic sculpture (see CHICHESTER). (3) Exceedingly rich, too, is the lower part of the _Great Gateway_ in College Green; “its courses, however, are nearly double the height of Norman masonry; the hood-moulds of all the arches are perpendicular, and at the crown of each arch are mitred into a perpendicular string-course; which, with the high finish of the ornament, points to the gateway having been wholly rebuilt in Perpendicular times”; in which case, like the eastern bay of the triforium of Rochester, it is an early example of “architectural forgery.” Fitzhardinge’s nave and aisles seem to have been pulled down in the sixteenth century with a view to rebuilding; which, however, was not carried out till 1877.

III. LANCET.—Early in the thirteenth century a beautiful Lady chapel was built, projecting eastward from the north transept, and separated by a few feet from the north wall of the choir. The same position was adopted later for the Lady chapels of Peterborough and Ely. This chapel is the artistic gem of the cathedral; it is surrounded by arcades of the greatest beauty, with sculptured grotesques interspersed among the foliage that remind one of the rich sculptured work in the retro-choir of Worcester, which is possibly a few years earlier. Later on it received the name of the elder Lady chapel; another Lady chapel having been built in the more normal position at the extreme east of the choir.

IV. Fitzhardinge’s cathedral, supplemented by the Lady chapel, seems to have satisfied the Canons very well for the first half of the thirteenth century. But in the EARLY GEOMETRICAL period (1245-1280), preparations were made to improve the lighting of the _north transept_ by the insertion of a big window in its north wall. The tracery of the present window is later; but its inner jambs, mouldings and shafts, and the external sill and string, as well as a great part of the buttresses, seem to have been executed _c._ 1250. About this date is the fine _doorway_, with a cleverly-contrived lintel, at the south end of the east walk of the _Cloister_.

V. In the MIDDLE of the GEOMETRICAL period, _c._ 1280, the east end of the Lady chapel was rebuilt, and a window with Geometrical tracery inserted; the tracery is of early character, containing nothing but foliated circles. At the same time it received a simple quadripartite vault; and to resist the thrust of this vault, the northern buttresses were reconstructed and weighted with pinnacles. Of these pinnacles only one, at the north-east corner, remains; the others seem to have been reconstructed at the time when the pinnacles of the choir were built.

VI. CURVILINEAR.—But the great building period at Bristol was between 1315 and 1349—that short but brilliant period when English mediæval design was at its best—which culminated in the Octagon and Lady chapel of Ely, the choir-screens of Southwell and Lincoln, and the Percy monument at Beverley. Everything east of the nave was pulled down and rebuilt _de novo_. And a very remarkable design it is; in fact, quite unique among our cathedrals. All other cathedral authorities had agreed long ago that the cardinal fault in all Romanesque design was its bad system of lighting, but that the remedy was to be found mainly in improving the top-lighting—_i.e._, in increasing the dimensions of the clerestory. Beverley clerestory had taller windows than Durham; Salisbury clerestory had three windows for every one of Beverley; Exeter spread out its windows in increasing breadth till they touched the buttresses on either side; the clerestories of the choir of Gloucester, now in course of erection, were a vast, lofty, continuous sheet of glass. But there was an alternative system of improving the lighting, which in many large churches, such as Grantham, Ledbury, and Leominster, was the result of fortuitous growth, but which in the choir of the Temple Church, London, and in Patrington Church now in course of erection, was the result of deliberate design. It was to magnify the aisles at the expense of the nave, to lift them up so high that windows of vast height could be placed in their walls, to dispense with a clerestory altogether, and to give to the pier-arcade of the nave the additional height gained by the suppression of the clerestory. It was to substitute side-lighting for top-lighting; to rely exclusively on the flood of light passing from vast, lofty aisle windows into the nave through its elevated arches. Hence the big windows of Bristol choir, each representing a pair of windows; the lower half the usual small window of an aisle, the upper half the larger window stolen from the clerestory.

But the new design had another merit, which probably weighed still more with the Bristol builders. The cardinal difficulty of the mediæval builders was how to keep up on the top of lofty clerestory walls a heavy stone vault which was always striving to push them asunder. They succeeded at length in keeping the clerestory walls from being thrust out by propping them up with flying-buttresses, perilously exposed, however, to all the vicissitudes of English weather. But there was another solution of the problem, which had been worked out in the thirteenth century in the Temple Church, and in the twelfth century in many a church in central and southern France, such as the old cathedral of Carcassonne. It was to stop the outward thrusts of the nave vault not by the inert resistance of buttress, pinnacle, and flying-buttress, but by bringing into play opposing thrusts—_i.e._, the inward thrusts of vaults built over the aisles. But to make these outward and inward thrusts balance and neutralise one another, the aisle vaults must be of pretty much the same height and span as those of the nave. The nave must be lowered or the aisles must be raised, or both. At Bristol the architect has preferred to raise the aisles. Then, the stability of the nave vault being secured, all the builder has to do is to stop the outward thrusts of the vaults of the aisles. This it is easy enough to do by a row of buttresses weighted with pinnacles. So Bristol Cathedral, to eyes accustomed to contemporary cathedrals, presents the strange solecism of having neither clerestory nor flying-buttresses. The whole design may well have been based on that of the English cathedral at Poitiers.

The Bristol architect was an architectural Radical. When it came, however, to the question of vaulting the aisles he had not the courage of his convictions. Having constructed aisles unusually lofty and impressive, he immediately sets to work to make them look low and ordinary by a complicated, costly and unnecessary system of skeleton vaulting. He seems to have thought that without some method of reducing the apparent height of the aisles, the church would look like no other church; it would look not as if it had the ordinary aisled nave, but as if it had three naves side by side. He ought to have been thankful to have the means of making it look unlike other churches; and, as a matter of fact, a design of parallel naves has a most noble effect, as may well be seen in the nave of Warwick Church, built in the most debased Gothic of the seventeenth century, but one of the most impressive designs in the country, full of light and atmosphere.

A little later, but still in the Curvilinear period, the Berkeley chapel was built, projecting southward from the easternmost bay of the south aisle of the choir, and a sacristy by which it is approached. Here is more skeleton-vault and much elaborate carving. The Newton chapel was also built, projecting eastward from the south transept; and perhaps at the same time the simple vault of the adjacent bay of the choir aisle was put up. A curious and by no means pleasing characteristic of the Bristol fourteenth-century architecture is the series of sepulchral recesses in the walls; they remind one of the cloister doorway at Norwich.

VII. PERPENDICULAR.—Early in the fifteenth century a central tower was added. Here again one is struck by the originality of the Bristol people; it is as beautiful as it is original. Fifteenth-century tower design rather runs in grooves. Gloucester, Evesham, Wrexham, Taunton, and the Somerset towers are cast very much in a mould. It is not so with the Bristol tower. The designer had noticed how beautiful is the effect of a close-packed range of tall clerestory windows, such as those of Leighton Buzzard Church. So, instead of restricting himself on each side of the tower to one or two windows, he inserts no less than three. The range of clerestory windows which the fourteenth-century builder refused to the choir becomes the special ornament and glory of the tower.

VIII. In 1888 the new nave and western towers were completed by Mr. Street. He can hardly be said to have made the most of his opportunities. The new work is copied closely from the old, and so is somewhat uninteresting. He had but to remedy some of the faults of the original design: to omit the ugly skeleton vaulting of the aisles, to simplify the vaulting of the choir, to discard the sepulchral recesses, to improve on the design of the windows of the choir—in which, as at Ely, flowing tracery is not seen at its best; and we should have had something original and interesting. The new nave might have been a criticism of the choir.

The Cathedral of Christ Church, Canterbury.

The metropolitan cathedral of Canterbury owes its enthralling interest to its vastness of scale, its wealth of monuments, its treasures of early glass, the great historical scenes that have been enacted within its walls—above all, to that greatest of all historical tragedies to the mind of the mediæval Englishman, the murder of Becket. It does not owe its distinction to its architecture. Whole building periods are almost wholly unrepresented; for the century and a half when English design was at its best, the Canterbury authorities slumbered and slept. What we have is the result of two periods only, with some scraps incorporated from earlier Norman work. What is there is not of the best: the Perpendicular work can be bettered at Gloucester, Winchester, and York; the work in the choir, a foreign importation, is not equal to that of its prototype, the French cathedral of Sens. We have many heterogeneous cathedrals in England. In the rest there is ever an attempt, usually a successful attempt, as at Hereford, and Gloucester, and Wells, to weld the conflicting elements of the design into symmetry and harmony. Canterbury scornfully declines any attempt at composition. Transepts and turrets and pinnacles are plumped down anyhow and anywhere; to the east it finishes abruptly in the ruined crags of a vast round tower; to the west the towers of its façade were, till lately, as incongruous in character as in date. Externally, the lofty central tower alone gives some unity to the scattered masses; internally it is an assemblage of distinct and discordant buildings.

I. NORMAN.—Of the pre-Conquest cathedrals of Canterbury nothing remains, unless it be fragments of rude masonry in crypt and cloister. Of Lanfranc’s cathedral, built, together with the Benedictine monastery, between 1070 and 1077, there remains the plinth of the walls of nave and transept. In the north transept some of his small square blocks of Caen stone are well seen just above the site of the martyrdom, as well as his turret in the north-west corner. His nave was allowed to stand till the fifteenth century. The present nave and central transept are built on Lanfranc’s foundations. The Norman work has had a most deleterious effect on the later design. To preserve the Norman north-west tower of the façade, the fifteenth century nave was built too short; to preserve the lines of his transept, it was rebuilt with a projection of only one bay to north and south, and without aisles, and appears woefully shrunken when compared with the vast transept of York, or even with the twelfth century transepts of Winchester and Ely. Lanfranc’s cathedral was an unambitious building, built in a hurry; closely copied, to save time probably, both in plan and dimensions, from William the Conqueror’s abbey-church at Caen, from which Lanfranc came to rule at Canterbury.

His choir consisted of but two bays and an apse. This was altogether inadequate for the church of a big monastery, and the seat of the Primate of all England. In 1096 it was pulled down and replaced by an enormous apsidal choir, some ten bays long, with a square Saint’s chapel to the east as at Rochester, with northern and southern towers flanking the main eastern apse, and crossed midway by a big eastern transept, from each arm of which two apsidal chapels projected to the east. What was the object of this vast eastward extension? It was probably due to the increasing tendency towards sacerdotalism, to the increase of veneration for the sacrament of the Lord’s Supper, and to the increase of Saint-worship. (1) In the monastic churches the monks were accustomed to sit in the crossing and in the western bays of the nave, where their stalls still remain at Norwich, Gloucester, and St. Albans. They wanted to be segregated and screened off from the laity, the sheep from the goats; the very stones of the church must preach the dignity of the priestly function. So they removed to the east, and shut themselves up in a choir of their own. (2) Secondly, they wanted a clear space for the high altar, a sanctuary for it. (3) They wanted space for many chapels and altars; not only those of the greater saints of the church, St. Mary, St. Peter, and the rest, but those saints—and they were very numerous at Canterbury—whose bodies or relics of whom were among the treasures of the cathedral—such as St. Dunstan, the destroyer of Secular Canons, and the martyred Archbishop Alphege. Crowds of pilgrims flocked to Canterbury to these shrines; room had to be found for them to pass round, to gaze for a moment at the holy relics, to say one prayer. Whatever the object of the extension, it set a precedent which was followed in the great majority of the cathedrals of England. In the latter days of the twelfth century Hereford and Chichester extended their choirs; in the thirteenth century the choirs of Lincoln, Lichfield, Worcester, Ely, and Exeter were rebuilt, and those of Winchester, Durham, and Lincoln were enlarged. In the fourteenth century those of Lichfield and Wells were enlarged, and that of York was rebuilt on a magnificent scale. Old St. Paul’s had a choir twelve bays long. Thus, till the end of the fourteenth century, the history of the English cathedrals was largely a history of the rebuilding or enlarging, or, as at Gloucester and Norwich, of the remodelling of choirs.

Of “Conrad’s glorious choir” (it was commenced by Prior Ernulph _c._ 1096 and finished _c._ 1115 by Prior Conrad) a considerable amount remains. The round-arched work in the crypt is nearly all of this date, except the carving of many of the capitals, which was executed later; and from the extent of his crypt one can plot out the exact shape and dimensions of the Norman choir. Much of it is seen outside, especially in and near the south-east transept with its intersecting semicircular arcades, and the most charming little Norman tower imaginable. In the interior many Norman stones, “cross-hatched,” may be seen in the aisle-wall immediately after entering the choir-aisle by the flight of steps; the lower part of the vaulting-shaft in this wall, built of several stones and not of solid drums, as it is higher up, is also Norman. In the eastern transept the triforium occurs twice over; the upper of the two was Conrad’s clerestory. Much of Conrad’s semicircular arcade also remains on the aisle-walls.

Conrad’s choir was not only far longer than Lanfranc’s, but it had the curious peculiarity (preserved in the French choir) that it was broader than the nave, and moreover widened out as it proceeded to the east. The double apses of each of his transepts were copied by St. Hugh at Lincoln a century later. A very noble feature of Canterbury choir is its elevation, necessitated by the construction of the crypt below it. The raising of the floor adds great dignity to the choir (one misses it painfully at Bristol); and was still further added to by the French architect, so that now at Canterbury one goes eastward from height to height. We climb from the nave to the choir, from the choir to the sanctuary, from the sanctuary to the eastern chapels. One gets a bathos at Durham and Worcester, where at the east one plunges into a hole.

II. TRANSITIONAL.—But Conrad’s glorious choir was destroyed by a great fire in the year 1174, amid much mediæval cursing and swearing, and the tears of all the people of Canterbury. Then the monks did an abominable thing. Instead of being satisfied with our home-bred English architecture, of which such a beautiful example was just being completed at Ripon, they sent for a foreigner. The present choir of Canterbury, like that of Westminster, was “made in France.” The only consolation one has is the fact—which is a fact—that with that stolid insularity which from the twelfth century has insisted on working out its own salvation in its own way—English architects ignored them both. The new French choir was to be a rock on which the main current of English art struck and parted asunder only to meet again on the other side. English design passed on, as if Canterbury choir had never existed, from Ripon and Chichester and Abbey Dore and Wells to Lincoln Minster. The coupled columns, the French arch-moulds, the Corinthianesque capitals of Canterbury were un-English; no one would have anything to do with them anywhere.

The choir, as rebuilt, was even longer than Conrad’s long choir. It has an elongated aisled apse beyond, and a curious circular chapel east of that. The former goes by the name of Trinity chapel, the latter of Becket’s corona. Becket’s first mass had been said in an older Trinity chapel; his body lay from 1170 to 1220 in the crypt below it; in 1220 he was translated to a magnificent shrine in the present Trinity chapel. The corona may perhaps have been erected to cover another shrine placed here and containing a fragment of Becket’s scalp. Sens seems to have had a similar corona.

The design of the choir is a close copy of the work at Sens, Noyon, Senlis, and the neighbouring cathedrals. Columns almost classical in proportion replace the heavy English cylinder. The coupled columns and Corinthianesque capitals of Sens are faithfully reproduced in the Trinity chapel. The choir, as at Sens, is arranged in coupled bays with sexpartite vaulting; while principal and intermediate piers, single and compound vaulting-shafts occur alternately in either choir. In unstable French fashion the vaulting-shaft is perched on the abacus. The abacus is square, except in the eastern part of the crypt. The capitals of the choir are foliated; the English moulded capital occurs only in the crypt. Each bay of the triforium in both cathedrals contains a couple of arches, each arch subdivided by a central shaft. Both cathedrals have round transverse arches in the vaulting of the aisles. The windows are not the tall slender lancets of England, but the broad squat lancets of France. The pointed arches of the apse of Trinity chapel on their tall stilts have a thoroughly French look. French, too, is the wish to dispense with a hood-mould round the pier-arches. And, as at Noyon, flying-buttresses emerge from the gloom of the triforium into the open air.

But there is another factor besides the personality of William of Sens. Having engaged a foreigner to do the work, the next step of the monks probably was to distrust him. The choir, as it appears now, bears unmistakable marks that not only William of Sens was at work, but also a British building committee. Being a Frenchman, William of Sens must have been an iconoclast, and would have liked to clear away the ruins and start _de novo_. His British employers, here, as in nearly all our cathedrals, parsimonious in the extreme, insisted on retaining every inch of old wall or pillar that could be utilised. The towers of St. Anselm and St. Andrew still stood, more or less stable, to the east. William therefore had to constrict his choir to pass between them. The result is the awkward twisting of the arcade and clerestory of the choir. The choir starts with becoming wider as it advances from the east; then it suddenly contracts to pass between St. Anselm’s and St. Andrew’s chapels; then it expands once more into the Trinity chapel. No Frenchman, if he had had his way, would have permitted his cathedral to be so distorted. Again, the building committee insisted on utilising the piers of Conrad’s crypt as far as possible as supports to the arcade of the choir above. The result is that the piers of the choir are not equally spaced; and the narrow arches are pointed, while the broad ones are semicircular and stilted. William must have been ashamed, too, of all the zigzag and billet ornament. But the building committee had probably a large stock of it collected from the debris after the fire, and wanted to work it off. One beautiful feature, however, is to be placed to English credit—viz., the profuse use of Purbeck marble shafts. On the whole we may take it that what we have is not William’s design as he would have wished it, but his design criticised, fettered, altered and ignored by a clerical building committee.