CANTERBURY CATHEDRAL FROM A DRAWING BY A. BRUNET-DEBAINES

GOTHIC
ARCHITECTURE

BY

ÉDOUARD CORROYER

ARCHITECT TO THE FRENCH GOVERNMENT AND INSPECTOR
OF DIOCESAN EDIFICES

EDITED BY

WALTER ARMSTRONG

DIRECTOR OF THE NATIONAL GALLERY OF IRELAND

With Two Hundred and Thirty-Six Illustrations

New York

MACMILLAN AND CO.

1893

EDITOR'S PREFACE

The following pages, which have been translated under my supervision by Miss Florence Simmonds, give such an account of the birth and evolution of Gothic Architecture as may be considered sufficient for a handbook. Mons. Corroyer writes, indeed, from a thoroughly French standpoint. He is apt to believe that everything admirable in Gothic architecture had a Gallic origin. Vexed questions of priority, such as that attaching to the choir of Lincoln, he dismisses with a phrase, while the larger question of French influence generally in these islands of ours, he solves by the simple process of referring every creation which takes his fancy either to a French master or a French example, here coming, be it said, into occasional collision with his own stock authority, the late Mons. Viollet-le-duc. The Chauvinistic tone thus given to his pages may be regretted, but, when all is said, it does not greatly affect their value as a picture of Gothic development. Mons. Corroyer confines himself in the main to broad principles. He travels along the line of evolution, pointing out how material conditions and discoveries, and their consequent social changes, brought about one development after another in the forms and methods of the architect. In a treatise so conceived, the fact that the field of observation is practically restricted to France, the few excursions beyond her frontier being made rather with a view to displaying the extent of her influence than with any desire for catholicity of grasp, is of no great moment. The English reader for whom this translation is intended, will get as clear a notion of how Gothic, as he knows it, came into being, as he would from a more universal survey, while he has the advantage of some echo, at least, of the vivacity, which inspires a Frenchman when his theme is "one of the Glories of France."

W. A.

CONTENTS

ILLUSTRATIONS

GOTHIC ARCHITECTURE

INTRODUCTION

The term Gothic, as applied to the architectural period dating from the middle of the twelfth to the end of the fifteenth century, is purely conventional.

The expression is clearly misleading as indicating the architecture of the Goths or Visigoths; for these tribes were vanquished by Clovis in the sixth century, and left no monumental trace of their invasion. Hence, their influence upon art was nil. The term is radically false both from the historical and the archæological point of view, and originates in an error which demands the strenuous opposition due to persistent fallacies. By a strange irony of fate the term Gothic, used in the last century merely as the opprobrious synonym of barbaric, has been specialised within the last sixty years in connection with that polished epoch of the Middle Ages which sheds most lustre upon our national art. And this, in spite of its Germanic origin.

Romanesque architecture, or to be exact, that architecture which, by virtue of the archæologic convention of 1825, we agree to label Romanesque, undoubtedly borrowed its essential elements from the Romans and Byzantines, modifying and perfecting them by the genius of Western Europe; but the architectural period which began in the middle of the twelfth century, and is so unjustly dubbed Gothic, was of purely French birth; its cradle was the nucleus of modern France. Aquitaine, Anjou, and Maine were the provinces in which it first took root. The royal domain, and notably the Ile-de-France, witnessed its most marvellous developments, and it was from the very heart of France that its splendour radiated throughout Europe.

But the tyranny of usage leaves us no choice as to the title of this volume. We are compelled to style it Gothic Architecture, though we would gladly have registered our protest by naming it French Mediæval Architecture.[1]

[1] This idea, which has recently found support in quarters which might have been considered free from such chauvinism, is based upon a narrow and peculiarly modern view of art. Art activities in the Middle Ages were as instinctive and unconscious as speech. The forms of architecture were invented and elaborated much in the same way as language. For the purpose of the historian of architecture, the northern half of France, the three southern quarters of Great Britain, and the districts threaded by the Rhine, form a single country, a single foyer of art. They all pressed on from similar starting-points to similar goals; and if the French went ahead in one direction, they fell astern in another. It may be allowed that, on the whole, the architects of the Ile-de-France did better than their rivals. Gothic architecture is pre-eminently logical, and logic is pre-eminently the artistic gift of the Frenchman. So that its more scientific development in the "French royal domain" was only to be expected. That success of this kind gives a right to call the whole development "French mediæval architecture" cannot be allowed.—Ed.

The term Gothic is, however, purely arbitrary, as is also that of pointed, which has been introduced by writers who admit the principle of the broken arch as the characteristic of so-called Gothic architecture.

The broken or pointed arch, which is formed by the intersection of two opposite curves at an angle more or less acute, was known to architects long before its systematic application. It occurs in buildings of the ninth century in Cairo, and was used prior to this in Armenia, and still earlier in Persia, where indeed it superseded all other forms of span from the times of the last of the Sassanides onwards. It is an expedient which gives increased power of resistance to the arch by diminishing its lateral thrusts.

The pointed arch is a form which admits of infinite variations. The one law which governs its construction is expediency. It frankly abandons those rules of classic proportion which are the canons, so to speak, of the round-headed arch. Thus we shall find the pointed approximating to the round-headed form in the twelfth century, only to diverge from it more widely than before, till, towards the close of the thirteenth and throughout the fourteenth century, it took on the acute proportions necessitated by a perilous disposition to prefer loftiness to solidity.

Fundamentally, it is of little moment whether the architecture of the twelfth to the sixteenth century be termed Gothic or pointed, when we recognise these terms as equally inexact. The point to be really insisted upon is that the filiation we have already demonstrated in our book on Romanesque Architecture continued slowly, but surely, in the wake of civilisation, of which architecture is ever one of the most striking manifestations.

So-called Gothic architecture was not the product of a single generation; it was the continuous logical development of the Romanesque movement, just as the latter in its time had been the outcome of a gradual adaptation of old traditions to new-born exigencies. Thus our Aquitainian forbears, by their successful translation into stone of the eastern cupola, prepared the way for the groined vault, the embryo of which is clearly traceable in the pendentives of the dome at St. Front.

The great churches which, towards the middle of the twelfth century, rose throughout the rich Western provinces that cluster about Aquitaine, were all constructed with groined vaults. In these examples we can discern no halting, tentative application of newly adopted principles. The work is that of consummate architects, who brought to their labours the assurance born of experienced skill, and in the later part of the twelfth century, the new system had replaced all others for the construction of vaults throughout Western Europe.

The architects of the royal domain, and notably those of the Ile-de-France, had been the first to adopt the groined vault. Towards the close of the twelfth century their assimilation of the new principles, their native ingenuity, and professional hardihood alike urged them to its further development. They became the inventors of the flying buttress.

The substitution of the groined vault for its parent, the cupola, was the direct consequence of the old tradition. The development was merely a stage in the march of ideas, a consummation logically arrived at in the track which the Romans, constructors not less bold though more prudent than their artistic progeny, had marked out for them. The groined vault, in short, is simply the growth of Roman principles perfected by continuous experiment. But the flying buttress, or rather the system of construction based on its use, caused a radical change in the art of building of the twelfth century. Stability, which in the ancient buildings was ensured by solid masses at the impost of vaults and arches, was replaced by the balance of parts. From this daring system some of the most marvellous of architectural effects have been won; but the innovation had a dangerous inherent weakness, inasmuch as it involved the exterior position of those essential vital organs for whose preservation the ancients had wisely provided, by keeping them within the building.

It is therefore not surprising that though fifty years after its introduction the groined vault was generally adopted throughout Western Europe, and even in the East, the success of the flying buttress was infinitely more gradual and restricted. Thus, in the North, the multiplication of great religious monuments built, or even rebuilt on the new lines, was simultaneous with the construction in the South of vast churches on the old principles. The adventurous builders of the North had eagerly adopted the new division of churches into several aisles, all with groined vaults, the vault of the great central nave relying upon exterior flying buttresses for resistance to its thrust.

In the South, on the other hand, architects were prudent, either through instinctive resistance to, or deliberate reaction from, the innovating influence, or by way of fidelity to an ancient tradition. They built with a single aisle, wide and lofty; the vaults were indeed supported by ribs, but their thrusts were received by powerful buttresses inside the walls, the projections thus formed being further utilised for the construction of chapels in the intervals.

This latter system, which has the incontestable merit of perfect solidity, recalls the construction of the Basilica of Constantine, or of the tepidarium in the Baths of Caracalla. The stability of the edifice was ensured by the resistance of masses at the imposts, and the whole principle of construction formed, as it were, a protest against the miracles of equilibrium so much in favour among the Northerners.

The new system of vaults supported by flying buttresses made very slight way in the South. It appears but rarely, and in the few instances where it is used has entirely the air of a foreign importation. Even in the cradle of its origin, it took root slowly and with difficulty, for its first applications were not without disaster. Lacking that mathematical knowledge which is the mainstay of the modern architect, the experimental skill shown by the thirteenth-century builder in constructing his vaults, and then in neutralising their thrusts by flying buttresses reduced to the legitimate function of permanent struts, was little short of miraculous. For it must be borne in mind that the thrust of these vaults, and the strength of the flying buttresses, varied of necessity according to their span, and the resisting powers of their materials. It was only by dint of long gropings in the dark that the necessarily empirical formulæ of the innovators were gradually transformed into recognised rules, and this knotty problem of construction received no positive solution till the last years of the thirteenth, or more emphatically, the first years of the fourteenth century. While even then the solution could claim no universal acceptance, for what was comparatively easy in countries where stone abounds became difficult, if not impossible, in districts where such a material as brick was the sole resource of builders.

Nevertheless, the growth of Gothic architecture was rapid, so rapid that even in the fourteenth century it began to show symptoms of that swift decadence which is the Nemesis of facile success. The abuse of equilibrium, the excessive diminution of points of support—defects often aggravated by insecurity of foundation and exaggerated loftiness of structure—the poor quality of materials, and the faulty setting thereof due to empirical methods, the over-rapidity of execution caused by mistaken emulation, the dearth of funds consequent on social and political convulsions complicated by the miseries of war,—all these things joined hands for the extinction of a once resplendent art. But the initial cause of its ruin must be sought in its abandonment of antique traditions. These traditions had persisted uninterruptedly throughout the so-called Romanesque period, only to pave the way for a seductive art in novel form, which, casting aside the trammels of the past in obedience to the dictates of the moment, fell on decay as rapidly as it had risen to eminence. Dawning in the France of Louis the Fat, it reached its apogee under St. Louis, and was in full decadence before the close of the fifteenth century.

The narrow limits assigned to us forbid not only detailed discussion of our great monuments, but even a summary of the most famous. We must be content to work out that theory of evolution already put forward by us in L'Architecture Romane. We propose merely to offer a synthesis of that architectural development which succeeded the so-called Romanesque epoch, from its birth in the twelfth to its extinction in the fifteenth century.

And as the groined vault is, broadly speaking, the essential characteristic of so-called Gothic architecture, and the flying buttress one of its most interesting manifestations, we shall make a special study of their origin, their modifications, and their principal applications in connection with religious, monastic, military, and civil architecture. We shall dwell more particularly upon religious architecture as presenting the grandest and most obvious evidences of artistic progress, not in its admirable buildings alone, but in those masterpieces of painting and sculpture to which it gave birth in France.

PART I
RELIGIOUS ARCHITECTURE

CHAPTER I
THE INFLUENCE OF THE CUPOLA UPON SO-CALLED GOTHIC ARCHITECTURE

The cupola, in its symbolic aspect, was the germ, whence sprang an architectural system the revolutionary action of which upon art can scarcely be over-estimated.[2]

[2] L'Architecture Romane, by Ed. Corroyer; Quantin, Paris, 1888.

So-called Gothic architecture was no spontaneous and miraculous manifestation. Like all human activities, its end is easy to determine; but it is difficult to fix even an approximate date for its beginning. The traces of its origin are lost in that period of architectural activity which preceded it, and prepared its way by a train of unbroken evolution.

The cupola of St. Front, which we may reasonably call the mother cupola of France, was not an imitation of that of St. Mark at Venice, for both were based upon the church built by Justinian at Constantinople, in honour of the Holy Apostles. But the form thus imported into Aquitaine received such modification and development, as to make it virtually an original achievement. One of the knottiest of architectural problems was solved in the process, and that admirable constructive principle was established which consists in concentrating the thrust of a vault upon four points of support strengthened by pendentives.

The construction of such a cupola as that of St. Front in dressed stone was an event of great moment in a district which still preserved the Gallo-Roman tradition in its integrity, and was commonly reputed the fatherland of our architecture. Its immediate consequences were shown before the close of the eleventh century by the erection of large abbey churches on the model of St. Front in various neighbouring provinces.

But while accepting the new principle, the architects of the period directed their energies to its perfectibility. Their efforts, and even their successes, in this direction are manifest so early as the first years of the twelfth century. The churches of Angoulême and of Fontevrault may be cited in proof. "We here recognise the main preoccupation of the Romanesque builders—namely, how best to reduce the immense masses of churches built with the primitive cupola by a more deliberate and judicious distribution of thrust and resistance. We further see how the adoption of these principles led to the emphasising of critical points by buttresses, which now began to project from the exterior walls."[3]

[3] L'Architecture Romane, by Ed. Corroyer; Quantin, Paris, 1888.

The new system spread rapidly, notably in Anjou and Maine, its growth being marked by an ever-increasing refinement and perfection. The architects of the rich abbeys of these provinces, the importance of which was aggrandised by their strong attachments to the all-powerful religious organisation of the period, gave a further development to the Aquitainian method. They transformed the pendentives of the cupolas into independent arches which performed exactly the same functions, thus logically working out an architectonic principle of amazing simplicity, the success of which was so rapid that, by the middle of the twelfth century, it was systematically applied to the construction of great churches at Angers, Laval, and Poitiers.

The works of the Angevin architects were of course known to their Northern brethren, who, in common with all the builders of the day, had long been seeking the final solution of the great problem of the vault. The architects of the Ile-de-France at once appropriated the Angevin system with that special professional ingenuity which characterised them, and applied it to the construction of innumerable churches, large and small, all of them built on the basilican plan—that is to say, with three, or even five aisles.

Thus the Aquitainian cupola of dressed stone exercised an absolutely direct influence upon Gothic architecture, since it gave birth to the intersecting arch, which is the main feature of so-called Gothic. This influence was first manifested in the general arrangement of single-aisled churches vaulted upon intersecting ribs, the earliest departure from the original cupola. It was then more grandiosely demonstrated in vast abbey or cathedral churches, built in accordance with the basilican tradition, and all vaulted on the new principle.

Angers and Laval are primitive examples of churches whose square compartments carry groined vaults, which thenceforth took the place of cupolas with pendentives.

The abbey church of Noyon shows the application of this principle, novel in the twelfth century, to the several-aisled churches of the Northern architects. The original vaults of Noyon[4] were planned in square. The intersecting arches united the principal piers diagonally, the strain being relieved by a subordinate or auxiliary arch which rested upon secondary piers, indicated on the exterior by buttresses less salient than those of the main piers, and on the interior by a column receiving the lateral archivolts which united the chief piers.

[4] The original disposition of the vaults built about 1160 is indicated by the spring of the arches above the capitals, and by the base plan of the principal piers. The present vaults on rectangular plan were built after the fire of 1238, in accordance with prevailing fashions.

This system of construction, the principle of which was logically developed at Noyon, for instance, no longer exists, save in its traditional state in the great churches of Laon, and in the cathedrals of Paris, Sens, and Bourges, to name but the principal, without regard to the innumerable churches built on these principles throughout Western Europe. In these great buildings the vaults were all square on plan down to the adoption in the first half of the thirteenth century of equal bays, vaulted on a rectangular plan, and marked inside and out by equal piers and projections, as at Amiens, Rheims, and many other churches of the period.

Hence we see how incontestable was the influence of the cupola upon so-called Gothic architecture. This truth is demonstrated by monuments yet in existence, lapidary documents above suspicion. It cannot be insisted upon too strongly, not merely for the satisfaction of archæeologic accuracy, but more especially as yet another proof that the filiation between the art of the ancients and that of the so-called Romanesque architects is no less evident than that which links together the Romanesque and the so-called Gothic. Of this latter filiation we have a direct proof in the Aquitainian cupola, the parent of those of Angoumois, which in their turn gave birth to the Angevin intersecting arch, and so prepared the way for the flying buttress, which again was to mark a new departure.

CHAPTER II
THE ORIGIN OF THE INTERSECTING VAULT

So early as the eleventh century churches were built with one or several aisles, and in this latter case the side aisles only had ribbed vaults, the nave being covered by a timber roof. The next step was to vault all three aisles, buttressing the barrel-vaulted nave by continuous half-barrel vaults or ribbed vaults over the aisles, and further strengthening it by projecting transverse arches, or arcs doubleaux, the whole being crowned by a roof which embraced the side aisles. These cumbrous and timidly constructed buildings were merely imitations of the Roman basilicas. To ensure their solidity they had perforce to be narrow; and the necessary abolition of top lighting made them gloomy. We find then that, before the appearance of the cupola, mediæval architects were perfectly acquainted both with the barrel vault and the ribbed vault, the latter formed, on traditional principles, by the interpenetration of two demi-cylinders. They had even attempted to improve upon the construction by strengthening the line of penetration with a salient rib, giving an elliptic arch. But this rib was purely decorative, for in the Roman vault the stones at the line of intersection, whether ribbed or not, were in complete solidarity with the filling on either side in which they were buried.

It follows that we shall seek in vain in the Roman ribbed vault the germ of the intersecting arch, with its essentially active functions.

For the origin of the intersecting arch we must turn to the eleventh century. We shall find it in the dressed stone cupola of St. Front, and more especially in its pendentives.

[Fig. 1] gives the plan of one of the cupolas of St. Front. It is composed of four massive transverse arches, the thrusts of which are received upon four piers united by pendentives (Figs. [2] and [3]) passing from the re-entering angles at the spring of the arches to the base of the circular dome itself, each of the concentric courses bearing upon the keys of the arcs-doubleaux, and transmitting to them, and therefore to the piers by which they are supported, the weight of the cupola itself.

1. PLAN OF A CUPOLA OF THE ABBEY CHURCH OF ST. FRONT AT PÉRIGUEUX

2. PENDENTIVE (MARKED A) OF A CUPOLA OF THE ABBEY CHURCH OF ST. FRONT

[Fig. 3] is a section through one of the pendentives of St. Front, following the line A B in [Fig. 1]. It shows that the first six courses are cut so as to make what is called a tas de chargé; the upper surfaces are horizontal, the faces curved to the radius of the dome itself. After the sixth course the voussoirs are cut normally to the curve of the arch. The vaulting of religious buildings having long been the crux of mediæval architects, the construction of the St. Front cupolas must have been an event much noised abroad, for towards the close of the eleventh century a large number of churches with cupolas were built in imitation of the mother church at Périgueux.

3. SECTION OF A PENDENTIVE ON THE DIAGONAL A TO B IN PLAN, FIG. 1

The construction of the churches of Angoulême and Fontevrault in the first years of the twelfth century shows that the architects were attempting to cover spaces of ever-increasing span on the Aquitainian model, while at the same time they set themselves to lighten their vaults, and consequently to reduce their points of support.

[Fig. 4] gives the plan of one of the cupolas of Angoulême or of Fontevrault, both being built on precisely similar plan, with the exception of the number of bays to the nave.

4. PLAN OF A CUPOLA OF ANGOULÊME OR FONTEVRAULT

[Fig. 5] gives the section of a bay in one of these churches, and illustrates the considerable difference already existing between the mother cupola of St. Front and its offspring. The cupola on pendentives begins to show a certain attenuation, and we shall presently note a fresh step forward towards the solution of that problem so persistently grappled with by the mediæval architect—how to reduce the weight of the vault.

5. SECTION OF A BAY OF THE CUPOLAS OF ANGOULÊME

The Church of St. Avit-Sénieur furnishes a most instructive example.

The cupola of this building is strengthened by stiffening ribs. It becomes an annular vault, formed of almost horizontal keyed courses, sustained by transverse and diagonal ribs, which act the part of a permanent centering.

The Church of St. Pierre at Saumur marks a further step onwards in the construction of vaults derived from the cupola.[5]

[5] L'Architecture Romane, by Ed. Corroyer.

6. SECTION OF A BAY IN THE CHURCH OF ST. AVIT-SÉNIEUR

Finally, the architects of Maine and Anjou achieved the long-desired consummation. Under their treatment the pendentives resolved themselves into their actively useful elements, the visible signs of which were diagonal or intersecting arches, salient and independent, set in precisely the same manner as the pendentives of the cupola ([Fig. 3]), and performing identical functions ([Fig. 8]).

7. PLAN OF VAULT ON INTERSECTING ARCHES

The vault proper is no longer formed of concentric courses, as in the mother cupola. It consists thenceforward of voussoirs cut normally to the curve, and filling the triangles (A, B, C, D, [Fig. 7]) determined by the longitudinal, the diagonal or intersecting, and the transverse arches. These arches form a stone skeleton, no less solid though far less ponderous than the cupola pendentives, and sustain the vault by distributing its thrusts over four points of support.

The triangular fillings no longer imprison the ribs, or, more exactly speaking, the intersecting arches, nor do they any longer neutralise their active functions. These fillings, on the other hand, have, like the intersecting arch, gained a new independence. They now contribute to the elasticity of the divers organs of the vault, a most essential element in its solidity. The peculiar arrangement of the intersecting arches in the nave of Angers gives incontrovertible proof of the direct filiation of this building to the Aquitainian cupola. The voussoirs of the intersecting arches are about equal in horizontal section to those of the transverse arches, while their vertical section equals the thickness of the filling plus the internal salience which marks their function. They look in fact like slices cut from the pendentives of a cupola (A, [Fig. 8]). It must be remarked, too, that at Angers the stones of the filling do not yet rest upon the extrados of the ribs, in the fashion adopted some years later in the Ile-de-France and elsewhere (see B, [Fig. 8]), but embrace them (as at A).

8. SECTION OF AN INTERSECTING ARCH

The identity of function in the pendentive and in the Gothic intersecting arch, both constructed, as they are, of stones dressed normally to their curves, shows that they sprang from a common origin, which is as much as to say that the Aquitainian cupola begat the intersecting vault.

CHAPTER III
THE FIRST GROINED VAULTS

The first application of the system of intersecting vaults appears in the great churches of Angers and Laval.

It is probable that the new methods propagated by the religious architects of Aquitaine and neighbouring provinces had excited the emulation of the Northern builders, more especially those of the Ile-de-France. Evidences to this effect are to be found in certain subordinate portions of their buildings at this period, such as side aisles or apsidal chapels. Their timid arrangement seems, however, reminiscent of the Roman system of ribbed vaulting, with a slightly increased prominence of the ribs superadded, rather than of the revolution that had been effected in church vaulting generally.

9. PLAN OF A BAY IN THE NAVE OF ST. MAURICE AT ANGERS

10. TRANSVERSE SECTION OF THE NAVE OF ST. MAURICE AT ANGERS

But, if we except perhaps Laval, nowhere shall we find the new system of vaulting upon intersecting arches more mightily demonstrated than at Angers, the aisles of which measure 54 feet across. The grandeur of the architectural composition, no less than the admirable technical skill shown in the details, gives proof of the consummate mastery arrived at by the builders of these noble structures so early as the middle of the twelfth century. The plan of these churches resembles that of Angoulême and Fontevrault. It is in no way allied to the Northern buildings.

11. PLAN OF A BAY OF THE NAVE IN THE CHURCH OF LA STE. TRINITÉ AT LAVAL

They are constructed with single aisles, like the cupola churches, with a series of bays, square on plan; but the arrangement of the vaults has been perfected by the logical use of intersecting arches in the place of pendentives, the architects of the day having realised by this time the progress we have explained and demonstrated in the preceding chapter.

These vast aisles, vaulted on intersecting arches, are of course allied to the cupolas; they recall their general outline, but the arrangement of the vaulting is different. The intersecting ribs are no longer merely decorative features; they have taken on all the active functions of the arc-doubleau and the formeret. Their union constitutes an elastic ossature, the weight being concentrated upon four points of support, which receive the impost of the arches, and compose a stone skeleton, each unit of which has been cut and dressed to fill the exact place it occupies in the whole.

12. LONGITUDINAL SECTION OF TWO BAYS IN THE NAVE OF LA STE. TRINITÉ AT LAVAL

If we compare the sections (Figs. [13] and [14]) of the churches of Angoulême and Angers, we may clearly trace the filiation between these buildings, the one dating from the first years of the twelfth century, the other from some thirty or even forty years later. We shall also note the advance made by the Angevin architects in the construction of groined vaults in the
place of domes with pendentives, a development worked out by the more perfect and reasoned application of the same architectural principle.

13 AND 14. COMPARATIVE SECTIONS OF THE CHURCHES OF ANGOULÊME AND ANGERS

15. VIEW IN PERSPECTIVE OF THE NAVE VAULT OF ST. MAURICE AT ANGERS

The Church of Laval, built simultaneously with that of Angers, or only a few years later, shows a further advance, not merely in the matter of form, but in the increased science and ingenuity of combinations, and the methodical accuracy of the execution.

16. PLAN OF A SUMMER OF THE NAVE VAULT OF LA STE. TRINITÉ AT LAVAL

17. PLAN OF ONE OF THE PIERS OF THE NAVE OF LA STE. TRINITÉ AT LAVAL

The arches which compose the ossature of the vaults become independent in their functions, as at Angers, immediately upon leaving the abacus, an essential characteristic of the new system. The lateral points of support are composed of piers proper and of clustered columns, crowned by corbelled capitals, which, by prolonging them, mark the formerets, the diagonal, and the transverse arches as they fall upon the abaci. It is easy to see in this arrangement the origin of those clustered shafts so generally and even excessively used in the thirteenth and fourteenth centuries, the main object of which was to conceal as far as possible the points of support.

These details, and the section ([Fig. 12]) showing the mode of construction in the vaults, demonstrate sufficiently that at Laval, no less than at Angers, a direct filiation exists between the dome upon pendentives and the groined and ribbed vault.

CHAPTER IV
BUILDINGS VAULTED UPON INTERSECTING ARCHES

The new system derived from the domes upon pendentives, so brilliantly applied in Anjou and Maine in the first half of the twelfth century, was thenceforth the normal method of the religious architect. The admirable simplicity of the new method and its adaptability to every class of building, from the great abbey church to the modest chapel, sufficiently accounts for its rapid dissemination throughout Western Europe, where religious bodies had founded innumerable abbeys, large and small, of varying rules and orders, but all welded together by one mighty organisation.

A long array of churches on the Angevin model rose, not only in the neighbouring provinces—as Ste. Radegonde at Poitiers, Notre Dame de la Coulture and the nave of St. Julien at Mans,—but farther afield towards the south. To name only the most important—the charming Church of Thor, dedicated to Ste. Marie du Lac, between Avignon and the fountain of Vaucluse; that of St. Sauveur at St. Macaire, near Bordeaux; the nave of St. André at Bordeaux, begun in 1252 on the cupola plan, but modified and finally crowned with a groined and ribbed vault; St. Caprais at Agen, which shows the same modifications, and lastly, the immense brick nave of St. Étienne at Toulouse, which measures 64 feet—all demonstrate the progression of the new principles in the second half of the twelfth century.

18. PLAN OF THE NAVE, ST. MAURICE AT ANGERS

Towards the North the advance was no less general. Various buildings show to what excellent account contemporary architects had turned the system of vaults on intersecting arches, recognising its admirable adaptability to different climates, and to the most diverse materials. But it was reserved for Angers, the cradle of its birth, to give an added perfection to this ingenious system.

The Church of the Ste. Trinité, on the right bank of the Maine, built by the sons or pupils of those architects who had planned St. Maurice for the hill on the opposite shore, marks a fresh advance in the construction of these vaults. Like St. Maurice, it has but a single aisle, which is divided into three bays, each as nearly as possible square on plan. The system of vaulting takes on a greater elegance by the insertion of a transverse arch, with its supporting shafts, in the centre of each bay. This divides the bay into two equal parts, and, cutting the diagonal ribs at their intersection, supports them at the critical point.

[Fig. 19] gives the plan of these vaults, the system of which was eagerly seized upon by the Northern architects, and the great abbey church of Noyon
appears to have been the first-fruits of this new development of the Angevin idea.

19. PLAN OF THE CHURCH OF LA STE. TRINITÉ AT ANGERS

20. LONGITUDINAL SECTION OF A BAY OF LA STE. TRINITÉ AT ANGERS

The great abbey churches and immense cathedrals which were built from the second half of the twelfth to the middle of the thirteenth century attest the importance of the development carried out at Angers by the arrangement of their own vaults in square compartments. For we now find this system adopted in the construction of the churches or cathedrals of Noyon, Laon, Notre Dame at Paris, Sens, and Bourges, to name only acknowledged masterpieces of so-called Gothic.

21. TRANSVERSE SECTION OF A BAY OF LA STE. TRINITÉ AT ANGERS

The influence of the cupola, which we established in our first chapter, was both direct and consecutive. It was direct in churches built with one aisle and vaulted on intersecting arches, and consecutive in the so-called Romanesque churches, which were either completed or modified on the new lines by the substitution of vaults on intersecting arches of dressed stone for timber roofs. A large number of buildings in England, Normandy, Germany, Northern Italy, Switzerland, the Rhine Provinces, and those of Northern France bear testimony of the highest interest to the transformations consequent on the invention of the groined vault and its universal application.

Architects who had been trained in the great abbey schools, emboldened by the successes of their forerunners and their own individual experience, raised on every hand vast cathedrals, in which every known development of the system was essayed with unequalled daring. Going on from strength to strength, they eventually abandoned the antique traditions, and disregarding the statical conditions which ensured the solidity of the ancient buildings, they invented a system of construction which is, as it were, merely a skeleton in stone, a stone version of the timbered roof; its characteristic expression was
the permanent strut known as the flying buttress; its governing idea was equilibrium, for which it provided by architectural stratagems ingenious in the highest degree, but also extremely precarious. Its existence or stability depends for the most part on the quality of the materials and their degrees of resisting power, the essential organs, by which I mean those vital weight-carrying portions, the failure of which would involve the ruin of the whole, being outside the building, and therefore exposed to all those deteriorating influences from which the load they bear, that is to say, the vaults, are protected by walls and roof.

22. SECTION OF A SINGLE-AISLED CHURCH VAULTED ON INTERSECTING ARCHES WITH BUTTRESSES

23. SECTION OF A THREE-AISLED CHURCH VAULTED ON INTERSECTING ARCHES WITH FLYING BUTTRESSES

The great buildings constructed on these new principles consisted of a central nave with two, or even four side aisles. The huge structure depended for its light first upon low windows in the collateral portions, secondly, upon windows at a much higher level. Hence it became necessary to raise the vault of the central nave, and to give it an abutment in the form of detached semi-arches or flying buttresses. The crowns of these semi-arches impinged the piers at the planes of greatest pressure and received the collective thrust of all the ribs, formerets, transverse and diagonal arches. Their bases rested upon abutments, the strength of which was calculated according to the thrust they had to meet.

CHAPTER V
ORIGIN OF THE FLYING BUTTRESS

The primitive method of vaulting adopted in the central provinces of France in the construction of churches with three aisles rendered such buildings of necessity low and heavy. The main aisle being covered by a barrel vault, supported on either side by a continuous half-barrel vault, the sole means of lighting lay in the windows of the side aisles, so that the nave was always gloomy in the extreme. The Norman architects had avoided this difficulty, first in their native province, and afterwards in England, by vaulting the subordinate aisles only, and by raising the lateral walls of the nave high enough to allow a line of windows to be introduced between the lean-to roofs of the side aisles and the nave roof, the latter being an open timber construction instead of a vault.

The lateral gallery in the first story of Norman churches built on the basilican model is merely a development of the ancient tradition.[6] It bears the name of triforium because—or so we are told—each compartment of such an interior gallery between the main piers of the nave was originally divided into three by pillars supporting lintels or by small columns supporting an arcade.

[6] See L'Architecture Romane, by Ed. Corroyer; Maison Quantin, Paris, 1888, chaps. i. iii. and iv.

Towards the close of the eleventh century Norman architects on both sides of the Channel were raising vast churches, the side aisles of which bore above their ribbed vaults galleries after the fashion of the primitive basilicas. These galleries in their turn were covered by open timber roofs like that of the nave. The bays were emphasised in the nave and in the side aisles by transverse arches, or arcs-doubleaux, which served as buttresses to those of the main vault. But after the adoption, towards the middle of the twelfth century, of the Angevin method of vaulting for religious buildings, the functions of the lateral walls and of the supporting arches became better defined, for these walls and arches had now to meet the thrusts of the transverse as well as that of the diagonal arches, which, meeting in bundles, as it were, at each pier, gathered their energies at well-marked points.

It was thus that the cross walls or arcs-doubleaux of the side aisles were gradually modified till they became detached semi-arches concealed beneath the outer roof of the side aisles.

We have traced this modification in the Abbaye aux Dames at Caen.[7]

[7] L'Architecture Romane, by Ed. Corroyer; Paris, Maison Quantin, 88, chap. xvii.

[Fig. 24] shows us an English example. It may be followed out in a number of other churches in England, at Pavia in Italy, at Zurich in Switzerland, and at Basle on the Rhine, to name but a few of the churches in which the modification of the vaults was long posterior to the construction of the building itself.

24. DURHAM CATHEDRAL. TRANSVERSE SECTIONS

In France we shall find no example more deeply interesting than Noyon, which at the date of its construction (the last quarter of the twelfth century) formed, as it were, an epitome of the advance so far made by the architects of the Ile-de-France. In this curious building we find a fusion of the antique tradition developed by the Normans in their triforiums, and of the Angevin methods, as manifested in the groined vaults derived from domes: methods further perfected by the example of La Ste. Trinité at Angers; in other words, by the adoption of intersecting arches planned on a square, the thrusts of all being received on the main piers, reinforced by an intermediate transverse arch. And we note the appearance of the detached semi-arch beneath the roofing of the inferior aisles merging at its springing into the lateral arc-doubleau, and so resisting the thrust of the intersecting arches and transverse arches of the nave.

25. ABBEY CHURCH AT NOYON. PLAN

26. TRANSVERSE SECTION OF NOYON CHURCH

It has been said that Noyon was suggested by Tournai, doubtless on account of their superficial affinities. But the likeness is merely in general aspect, the methods of construction being wholly different. At Tournai the apsidal transepts are vaulted upon transverse arches of great strength, and upon radiating semi-arches united where they meet by a ring of voussoirs set horizontally, and at their springing by vaults keyed into their mass, an ingenious arrangement which recalls the vaulting of the Salle des Capitaines over the porch of the monastery church at Moissac.

The combination of these arcs-doubleaux, which, in addition to the solidity of their independent structure, are strongly reinforced by the massive circular courses of the walls, is very peculiar, for it dispenses altogether both with auxiliary arches and with abutments. Tournai, therefore, cannot be held to have begotten Noyon, for here we have groined vaults, the intersecting arches of which demand the reinforcement of abutments either concealed or apparent to sustain the thrust of these vaults over the lateral arcs-doubleaux. The ingenious arrangement above cited had in no sense modified the methods of abutment followed by the architects of the twelfth century even after the adoption of the vault on intersecting arches. These, as will be remembered, consisted in buttressing the walls and piers of the nave by cross walls or by arches concealed beneath the roofing of the side aisles.

27. CHURCH OF TOURNAI, BELGIUM. EXTERIOR VIEW OF THE NORTH TRANSEPT TOWARDS THE SCHELDT

28. MONASTERY CHURCH AT MOISSAC. VAULT OF THE HALL KNOWN AS THE HALL OF THE CAPTAINS ABOVE THE PORCH

29. CHURCH OF TOURNAI, BELGIUM. INTERIOR OF THE NORTH TRANSEPT

We find at Soissons the first application of an architectural system, the special feature of which is the flying buttress.

The south transept of Soissons Cathedral was evidently suggested by Noyon. This is apparent in the adoption of the two-storied side aisle and in the semi-circular plan. But the method of vaulting common to both churches has a greater refinement at Soissons. Reduced to its simplest expression of strength by the attenuation of its skeleton, the vault still exercises its full thrust on those parts which rise above the upper gallery.

The architect of Soissons was not content, like his brother of Noyon, to support the vault laterally by interior arches collaborating with the arcs-doubleaux of the triforium, and reinforced by an abutment impinging on the wall of the central nave. To him the idea occurred of detached semi-arches in open air, springing from above the roof of the triforium and its buttresses and marking each bay. Thus was born the flying buttress, a feature frankly emphasising its special aim and function, namely, to meet the thrust of the main vault at its points of concentration.

30. SOISSONS CATHEDRAL. SOUTH TRANSEPT. SECTION OF FLYING BUTTRESS

31. PERSPECTIVE VIEW OF SOUTH TRANSEPT, SOISSONS CATHEDRAL[8]

[8] These flying buttresses, in themselves insufficient for the task laid upon them, and worn by the destructive action of the weather, were pushed entirely out of shape by the constant pressure from within, the thrust of the vault being aggravated by the circular plan of the building, while the vaults themselves became dislocated by reason of their insufficient abutments. It became necessary to reconstruct the buttresses in 1880, to avert the total collapse of the south transept.

The reconstruction of these flying buttresses, and of many others of the same period, furnishes us with a criticism ad hominem upon the system.

The flying buttress, in combination with the intersecting arch, gave birth to a new system of construction, a system on which were raised vast buildings which compel our admiration and demand our careful study, but should not invite our imitation. They are monuments to the ingenuity of the twelfth and thirteenth century architect, but no less are they beacons warning against the perils of a rationalism—more apparent than real—which their authors carried to its extreme limits, casting to the winds all traditional principles, and consequently all authority.

It would seem as though the architects of this period, emboldened by such achievements as the churches of Noyon, Soissons, Laon, Paris, Sens, and Bourges, and spurred by professional emulation, went on from one feat of daring to another, passing from the triumphs of Rheims, Amiens, and Mans to the supreme architectural folly of Beauvais, and creating monuments no less amazing in dimension than in the statical problems grappled with, if not always solved.

CHAPTER VI
CHURCHES AND CATHEDRALS OF THE TWELFTH AND THIRTEENTH CENTURIES

The study of mediæval architecture is one of the most fascinating of pursuits, but it is one beset with difficulties. The obscurity in which the origin of our great monuments is buried is profound and often impenetrable.

A fertile cause of error is the confusion which in many cases has arisen between the dates of foundation and of consecration. Very often a church was built and afterwards considerably modified, rather than actually reconstructed, on the same consecrated site.

Lightning was the most frequent cause of the destruction, total or partial, of mediæval churches. Striking the steeple, the tower, or the roof, it fired the timber superstructure of the nave. This in itself would not have been an irreparable disaster; but as the timbers gave way the calcined beams charred the piers, and so prepared the downfall of the whole building, which was then either restored or reconstructed in the fashion of the day. Hence, whether we base our deductions upon more or less trustworthy records or upon contemporary readings of existing data, the result is too often a confusion among vanished monuments, or a contradiction between the buildings as they now exist and the historic records which relate to them.

32. CATHEDRAL OF LAON. PLAN

33. CATHEDRAL OF LAON. INTERIOR OF THE NAVE

Nothing is easier for interested theorists than to post-or ante-date the structure of a building. They have nothing to fear from the testimony of writers, and, with very few exceptions, it is difficult to assign a precise date to the construction of great churches and cathedrals or to point with certainty to their architects. The obscurity of these great artists is perhaps to be accounted for by the fact that they were ecclesiastics. As such the honour of their achievements belonged not to the individual, but to the corporate body, the order of which they were members, and members moreover who had, in most cases, taken the vow of humility.

Modern science, architectural and archæological, has failed to throw much positive light on this subject. It contents itself for the most part with ingenious hypotheses and learned deductions which leave us still in doubt as to precise dates. But we shall at least find some sort of foothold in a careful architectural survey of buildings themselves. This should be, of course, supplemented by study of historic records, and such a study will convince us that art in the Middle Ages, as in all epochs, obeyed the immutable laws of filiation and transformation. We shall follow the artist step by step, observing his research, his hesitation, his errors, and even his corrections.

These are trustworthy documents in which to study the origin of a building and to note its successive transformations, which latter were far more frequent than total reconstructions. For it was not until the beginning of the thirteenth century that great cathedral churches in any considerable numbers were conceived and continuously executed.[9]

[9] It is possible, if not easy, to trace the architectural development of the Middle Ages in a good many cathedrals and churches of the twelfth and thirteenth centuries. We have, however, confined ourselves, for the purposes of our present synthesis, to the churches and cathedrals of the royal domain, and more especially of the Ile-de-France, not only because they served as models for the architects of their day, but because they illustrate in a remarkable degree the various transitions we desire to study.

34. CATHEDRAL OF LAON. MAIN FAÇADE

The great abbey churches founded towards the close of the twelfth century in the royal domain, but continued and finished in the early years of the thirteenth, still preserved a more ancient tradition.

Laon, which is derived from Noyon and from the south transept of Soissons, consists of a nave with transepts, and of two-storied side aisles vaulted upon intersecting arches, above which, as at Soissons, rise flying buttresses, which meet the thrust of the main vault.

This arrangement of the side aisles proves the continuity of the Norman formulæ, just as the method of construction adopted in the main vault demonstrates the persistent influence of the dome.[10]

[10] See chap. i., "The Influence of the Cupola on Gothic Architecture."

The admirably constructed main vault is square on plan, each square containing two transverse compartments, after the Angevin method as derived from the Aquitainian dome. Here we find indications that, if the builders of the Church of Laon had fully assimilated this method, their minds were nevertheless not altogether at rest as to the functions of the flying buttress. This was, of course, essential to the piers which received the united thrust of both transverse and diagonal arches. But it was far from logical to reinforce the intermediate piers supporting nothing but the auxiliary transverse arches by abutments identical with those of the main piers.

The illogicality so striking at Laon is absent from Noyon. There, on the contrary, the architects—of the original construction—had emphasised the functions of the main piers by buttresses of greater projection and solidity than those accorded to the secondary piers.

35. CATHEDRAL OF LAON. THE EAST END

36. CATHEDRAL OF LAON. SECTION OF THE NAVE

Notre Dame de Paris was begun towards the close of the twelfth century, and finished, save for the chapels, in the first half of the thirteenth. As at Laon, the Norman tradition is observed in the arrangement of the upper galleries of the side aisles, while the influence of the dome is again to be traced in the sex-partite groining. The same illogical system of abutments obtains as at Laon.

This vast building, consisting of a nave and double side aisles of equal height sweeping round the semi-circular choir, seems to be one of the first five-aisled cathedrals; its grandiose arrangement, the boldness of its combinations, and the perfection of its detail mark the considerable progress made by the architects of the Ile-de-France.

37. NOTRE DAME DE PARIS. PLAN

The method of construction here adopted has a peculiar significance. The upper internal galleries, vaulted on diagonal arches, and raised considerably above the level of the second side aisle, the boldness of the flying buttress, which at one span embraces the two side aisles and forms the abutments of the main vault—alike prove that the architects of Notre Dame de Paris had adopted the newly discovered systems even to excess, and were applying them with unparalleled skill and ingenuity.

38. NOTRE DAME DE PARIS. SECTION OF THE NAVE

The Norman tradition which had obtained in the Ile-de-France passed away in the first years of the thirteenth century. At Châlons-sur-Marne the nave is flanked by two-storied side aisles. But the upper gallery, vaulted and greatly reduced in size, shows that the conventional arrangement was fast dying out.

The influence of the dome was longer lived, as is shown in the construction of vaults at this period. We may still trace it at Langres in the domed form of the vaults, which, in spite of their rectangular plan, seem to be a reduced copy of the Angevin naves.

39. NOTRE DAME DE PARIS. FLYING BUTTRESSES AND SOUTH TOWER

The naves of Sens and of Bourges are also vaulted in square compartments. The thrust of the vaults is carried by the diagonal arches to each alternate pier, the intermediate one receiving only the auxiliary transverse arch already fully described. Yet here again the exterior flying buttresses are all of equal solidity in spite of the varying strain. This arrangement, prudent if illogical, shows once more with what distrust architects had adopted that system of exterior abutment, the characteristic of which is a detached arch exposed to all the vicissitudes of weather, and yet responsible for the stability of the whole edifice.

The Cathedral of Sens marks a new phase of development by its suppression of the upper gallery over the side aisles. These are now vaulted and covered by a lean-to roof; a flying buttress of single span receives the thrust of the main vault. The building is perfectly solid; its construction shows research, though it is as illogical as that of Laon or of Paris; for the exterior flying buttresses are all of equal strength, and so fail to proclaim their true functions, the interior thrusts varying considerably.

40. SENS CATHEDRAL. PLAN OF A BAY. VAULT IN SQUARE COMPARTMENTS OF TWO BAYS

41. SENS CATHEDRAL. SECTION OF A BAY OF THE NAVE

42. SENS CATHEDRAL. INTERIOR VIEW OF LATERAL BAYS

The arrangement at Bourges, which appears to have been mainly built, if not actually finished, in the first half of the thirteenth century, differs from that of Sens. The structure is one of five aisles, and in plan recalls Notre Dame de Paris, but the details are very dissimilar. The inner side aisles no longer support a gallery, nor are they of equal height with the outer aisles; they are raised so as to afford space for lighting (see [Fig. 43]). The main vault is sex-partite planned on squares; but the same illogicality exists here which we have already pointed out, and in connection with which we will risk appearing somewhat insistent, in the hope of directing special attention to it. It is more glaring here than elsewhere, the flying buttresses themselves being of exaggerated dimensions and of double span, embracing the two side aisles.

43. BOURGES CATHEDRAL. SECTION OF THE NAVE

Both at Bourges and Sens the space between the summit of the archivolts and the bases of the upper windows, known as the frieze, or, in modern parlance, the triforium, becomes a purely decorative feature. It consists of a narrow arcaded corridor, occupying in the interior of the building that portion of the wall space which in the exterior has been appropriated by the lean-to roof of the side aisles. At Sens there is merely a single gallery; at Bourges it becomes double, through the stepped arrangement of the side aisles (see [Fig. 43]), a variation in which we may trace an ingenious blending of the systems of Anjou and Poitiers with those of the Ile-de-France.

CHAPTER VII
THE CATHEDRALS OF THE THIRTEENTH CENTURY

The Cathedral of Rheims, which was begun soon after the destruction of the original building by the fire of 1211, is a supreme expression of the fusion of the three systems—those of Aquitaine, of Anjou, and of the Ile-de-France. It may be taken as the most perfect manifestation of persistent efforts to establish a method of construction based on equilibrium—the equilibrium, that is to say, of a building vaulted on intersecting arches, the thrusts of which are received by exterior flying buttresses.

The temerity, and even the dangers of such a system, are sufficiently demonstrated in the wonderful works of the thirteenth-century architects themselves. For, notwithstanding the skill and beauty of their many admirable combinations, they were unable to reduce their methods to scientific formulæ. The statical power of their structures remained an uncertain quantity, determined by the durability of the material and its exposure or non-exposure to the weather, the interior skeleton being formed of the same material as the exterior.

44. RHEIMS CATHEDRAL. PLAN

The perils inherent in such a system are more apparent at Rheims than elsewhere, because of the colossal proportions of the building. The arrangement of the flying buttresses, however, is more logical than at Laon, Paris, Sens, and Bourges, by reason of the quadripartite arrangement of the main vault. The thrusts being equally distributed among the supporting piers, each flying buttress performs an identical office; their equal strength and solidity is therefore perfectly appropriate and logical. But though theoretically correct in its disposition of flying buttresses of equal strength to meet thrusts of equal strength, the method is vitiated by its inherent weakness as a system of abutment. The fragility of the flying buttress exposed it to two grave dangers, active and passive; active, taking into account the constant strain upon it as an abutment; passive, in regard to the gradual reduction of its solidity by exposure to weather. In support of this statement, it is only necessary to refer to the restorations which it has been found necessary to make within the last few years, to preserve the nave. The flying buttresses have been strengthened from below, a proceeding without which the collapse of the huge building would have been inevitable.

But we shall find much to call for unqualified admiration at Rheims in the grandiose conception of the work and in its powerful execution, in the magnificent arrangement of its eastern façade, and in the perfect harmony of the ornamentation, where sculpture, capitals, friezes, crockets, and floriations are so many types of mediæval decorative art at its best.

The Cathedral of Amiens, which dates from about 1220, and is one of the largest as well as one of the most admired of Gothic masterpieces, is directly founded upon that of Rheims. The plan is on the same lines, with this exception, that at Amiens the choir is of greater importance relatively to the nave, and that the piers and points of support are weaker and much more lofty.

45. RHEIMS CATHEDRAL. SECTION OF THE NAVE

46. RHEIMS CATHEDRAL. FLYING BUTTRESSES OF THE CHOIR

The Rémois architects, while exercised by the problems of equilibrium which their system involved, sought to minimise its dangers, which they recognised no less fully than their predecessors, by prudently avoiding all false bearings. It will be easily seen by a comparison of the two sections (Figs. 45 and 48) that the builders of Amiens were troubled by no such misgivings, or that they were at least more venturesome if not more accomplished. They did not hesitate to base the columns which received the crowns of the flying buttresses on a corbel arrangement which had no solid bearing, as may be seen by following the direction of the dotted line X in [Fig. 48]. The boldness, or rather the imprudence of such
an arrangement is patent, for the failure of anyone of the courses, or the decay of any part of the pier into which the corbels are keyed, would necessarily involve a rupture in the flying buttresses, on which the stability of the main vault depends. The disintegration of the whole building and its total ruin could be the only result. The perils of such combinations, or rather such tours de force of equilibrium, are exemplified at Beauvais. The architects who built the choir, about the year 1225, basing it on that of Amiens, determined to raise a monument which should surpass, both in plan and elevation, all the structures of their epoch. They increased the breadth of the choir and of its bays, raising, in the latter, intermediate piers on the crowns of the lower archivolts, thus dividing the upper bays, and at the same time strengthening the vault by auxiliary transverse arches. They exaggerated the height of the archivolts and of the large windows, and diminished their thickness, in order to give greater elegance and lightness, and the main vault rose to a height of more than 160 feet above the ground level. This tremendous elevation, the exaggeration of which in proportion to the width of the nave is striking, necessitated a complicated system of flying buttresses surpassing in boldness all that had gone before. The section in [Fig. 51] will give some idea of what has been justly described as an architectural folly. It is astonishing that the structure should have stood as it has done, taking into account the false bearings of the intermediate piers, here again shown by the dotted line X ([Fig. 51]).

These rest for half of their thickness on off-sets from the piers, which, proving unequal to the strain, have been temporarily stayed, and must eventually be consolidated.

47. AMIENS CATHEDRAL. PLAN

48. AMIENS CATHEDRAL. SECTION THROUGH THE NAVE

The choir, however, was finished about 1270, and stood for several years. But dislocations then declared themselves. The forces so elaborately balanced lost their equilibrium, and on the 29th November 1284 the vault fell, dragging down with it the flying buttresses, and carrying havoc through the rest of the building. In the reconstruction which followed it was thought imperative to double the points of support in the arcades both of the main and side aisles, and to reinforce the flying buttresses by iron chains.

During the thirteenth century a number of cathedrals were raised all over Europe on the model of the great buildings of Northern France, and more especially of Amiens, which seems to have roused a great enthusiasm; these were, however, of far more modest dimensions. They had neither the exaggerated height nor the structural audacities of their exemplars. Few of these churches and cathedrals, the reconstruction of which on the new system generally began with the choir, which was
added to the primitive nave, were completed by those who initiated their erection. The most highly favoured in this respect were finished in the course of the fourteenth century; but in the greater number of cases the work dragged slowly on, and reached its end some two centuries after its inauguration. Reconstructive undertakings were constantly impeded by wars or social convulsions, which either hampered or entirely cut off the resources of bishops and architects, their promoters. Such interruptions were of great service to modern archæological study, offering as they do distinct evidence of the various transformations which were successively accomplished from the so-called Romanesque period to the Gothic.

49. BEAUVAIS CATHEDRAL. APSE

50. BEAUVAIS CATHEDRAL. NORTH FRONT

51. BEAUVAIS CATHEDRAL. TRANSVERSE SECTION

52. CHARTRES CATHEDRAL. ROSE WINDOW OF NORTH TRANSEPT

The majority of these great buildings, which show traces of the vicissitudes through which they passed, bear a strong likeness to each other, and vary only in detail, according to the skill of their constructors.

The peculiar interest of Chartres centres in its remarkable statuary; it has, however, other features which command attention, such as the rose window of the north, transept and the design of the flying buttresses. These consist of three arches, one above the other, the two lower ones being connected by colonnettes, radiating from a centre, so that the lower arch is related to the upper, as the nave of a wheel is to the felloes, the colonnettes forming the spokes.

At Mans the arrangement of the choir is so far more remarkable in that it is extremely unusual, or indeed, in its way unique. The flying buttresses are planned in the form of a Y (see A on the plan [Fig. 53]), thus affording space for windows in the exterior wall, to light the vast circular ambulatory, which at Mans is of unusual importance, and surrounds the choir with a double aisle. The flying buttresses which rise above the arcs-doubleaux, bi-furcated (B on the plan), are over-attenuated in section; their exaggerated height and proportionate slenderness threaten to make them spring, so that it has been found necessary to bind them together by ties and
iron chains. Such expedients are a sufficient criticism of the ingenious but precarious system adopted by the architects of Mans.

53. MANS CATHEDRAL. PLAN

54. MANS CATHEDRAL. FLYING BUTTRESSES OF THE APSE

55. MANS CATHEDRAL. SECTION OF THE CHOIR

The influence of the Ile-de-France in Normandy is manifest in the arrangement of choirs and apsidal chapels in Norman cathedrals of the thirteenth century. The Cathedral of Coutances, a monument of the eleventh century, was rebuilt in the early
years of the thirteenth century under the impulse given by Northern France to the architecture of the period. It is in the choir that we clearly trace this influence, in the double columns of the apse, and the ingenious disposition of its collateral vaults. But the façade is purely Norman, not merely in general design, but in the details of the composition, facsimiles of which may be found in England.

56. COUTANCES CATHEDRAL. NORTH TOWER

The Cathedral of Dol in Brittany, one of the great churches of the thirteenth century, seems to have escaped the influences of the Northern innovation. Its general plan, its square apse lighted by large windows, the details of its architecture and ornamentation, all proclaim its affinity to the great churches which rose contemporaneously with it on either side of the Channel, in Normandy, and in England. It is very probable that it was built by the same architects or their immediate disciples, working on the more ancient methods of the Norman schools founded by Lanfranc at Canterbury towards the close of the eleventh century, on the model of those he had established in France at the famous Abbaye du Bec.