CHARACTER OF
RENAISSANCE ARCHITECTURE

CHARACTER OF
Renaissance Architecture

BY

CHARLES HERBERT MOORE

AUTHOR OF “DEVELOPMENT AND CHARACTER OF
GOTHIC ARCHITECTURE”

_WITH TWELVE PLATES IN PHOTOGRAVURE AND ONE HUNDRED
AND THIRTY-NINE ILLUSTRATIONS IN THE TEXT_

New York
THE MACMILLAN COMPANY
LONDON: MACMILLAN & CO., Ltd.
1905

_All rights reserved_

Copyright, 1905,
By THE MACMILLAN COMPANY.

Set up and electrotyped. Published October, 1905.

Norwood Press
J. S. Cushing & Co.—Berwick & Smith Co.
Norwood, Mass., U.S.A.

TO MY DAUGHTER

E. H. M.

PREFACE

In the following attempt to set forth the true character of the architecture of the Renaissance I have endeavoured to reduce mere descriptions of buildings to a minimum, and to give graphic illustrations enough to make the discussions clear. The illustrations in the text are mainly from my own drawings, for the most part from photographs: but in a few cases I have reproduced woodcuts from the works of old writers, indicating, in each case, the source from which the cut is derived. The photogravure plates are from photographs by Alinari, Moscioni, Naya, Wilson, and Valentine. The right to reproduce and publish them has been obtained by purchase.

With the best intentions and the greatest care, it is almost inevitable that a writer on such a subject should make some mistakes, and I cannot affirm that no inexact statements will be found in these pages, but I believe that no fundamental errors occur.

I am again indebted to my almost life-long friend, Professor Charles Eliot Norton, for valuable criticism, and painstaking revision; but Professor Norton is not responsible for anything that I have said. I am indebted, also, to my publishers for their courteous compliance with my wishes as to the style and manufacture of the book, and to Mrs. Grace Walden for the care and thoroughness with which she has prepared the index.

Cambridge, Mass.,

October, 1905.

CONTENTS

ILLUSTRATIONS IN THE TEXT

LIST OF PLATES

CHAPTER I
INTRODUCTION

The great change in ideas and ideals which, after the remarkable intellectual and artistic life of the Middle Ages, was manifested in the so-called Renaissance, is not always correctly conceived or fairly stated; and the character and merits of the Fine Arts of the Renaissance, as compared with those of mediæval times, have not, I think, been often set forth in an entirely true light. Of the merits of the best Italian art of the fifteenth and sixteenth centuries there can be no question, but the belief that this art is altogether superior to that of the Middle Ages will not bear examination in the light of impartial comparison.

The Fine Arts are always an expression of the historical antecedents, the intellectual, moral, and material conditions, and the religious beliefs of the peoples and epochs to which they belong. They derive their whole character from these antecedents and conditions, and cannot be rightly understood or appreciated without reference to them. Thus a brief consideration of these conditions in the Middle Ages on the one hand, and in the period of the Renaissance on the other, may help us to understand the nature of the above-mentioned change, and to gain a more discriminating appreciation of the real character of the artistic productions of the latter epoch.

During the Middle Ages ideas and imagination were governed by a religious faith which, though in many ways mistaken and misguided, was for the most part firm and unquestioning. Mediæval Christianity was a living power with the masses, and an inspiration to men of genius. The mediæval Christian mythology was well fitted to stimulate artistic invention, and the ideals which it maintained were full of beauty. It is true, indeed, that human conduct was not wholly governed by this faith; but the precepts of the Christian religion, as defined and interpreted by the Roman church, were generally held as of supreme authority, and to them most people acknowledged that they ought to conform. This Christianity gave the chief motive power for the best activities of the time, and the social relations of men were, in theory at least, based upon its teachings. The history of the Middle Ages abounds in evidence that popular habits of life were in many ways exemplary. Villani tells us that the citizens of Florence lived in sobriety and frugality, that they had loyal hearts, were faithful to one another, and that they required the same fidelity in the administration of public affairs.[1] Florence in the fourteenth century was alive with industry, and the open country around the city was prosperous with agriculture. Of such conditions her Fine Arts were an outgrowth and expression.

But the mediæval faith began at length to weaken. The church, as an ecclesiastical establishment, had grown corrupt and oppressive, so that men of spirit were moved to reject its dogmas and to resist its intellectual tyranny. Independent thought began to widen the range of ideas, and the reading of ancient authors gave a fresh incentive to philosophical speculation, and awakened a spirit of scientific investigation, as well as a taste for ancient poetry and mythology. The desire for intellectual freedom, and the thirst for new knowledge, which were thus stimulated in the fifteenth century constitute the good side of the Renaissance movement, the side which has hitherto been most emphasized by writers, and to which the modern world is indebted for a strong stimulus in the direction of some of its most fruitful activities.

But there were other conditions that must not be ignored if we would rightly understand the spirit of the Renaissance, by which the ideals and aims of this brilliant epoch were materially qualified and weakened. Influences were at the same time at work that were not in harmony with what was best. The humanist learning bred a Neo-pagan spirit which favoured and strengthened a growing indifference to moral principles and religious beliefs. The strong feeling of opposition to the church was in part due to this. In fact, the Renaissance was by no means an entirely noble movement in the interest of spiritual and intellectual emancipation, or an unqualified advance in ideas and attainments beyond those of the Middle Ages. With all of its abuses the church still stood for moral order and spiritual aspirations. The revolt against it was in part a revolt against both religion and morals. The animating spirit of the movement contained much that was unchristian and destructive of high ideals.

It is true that noble, and even pious, feelings survived in the minds of many men, especially during the early Renaissance time. Generous acts were still common among the merchant princes of Florence. In the early part of the fifteenth century the lives of Florentine patricians continued to be simple, and many of them recognized the responsibilities which their wealth imposed.[2] But toward the close of that century a different spirit prevailed. Luxury and extravagance took the place of plainer living, the pursuit of pleasure without regard to justice or morality engrossed the minds of men, and vice and crime flourished in high places until the prophetic denunciations of Savonarola were called down upon the wickedness and vanity of the upper classes.

Into the service of this luxurious and immoral life the Fine Arts were now called, and of the motives which animate such life they become largely an expression. The mediæval endeavour to embody the beauty of Christian ideals in works of art gave place to the desire to make the Fine Arts minister to sensuous pleasure and to mundane pride. In the height of its splendour the vicious life of Florence, the chief centre of literary and artistic productions, was appalling. Men now not only sought to escape from all forms of ecclesiastical and ascetic restraint; they went further, and freely proclaimed the sufficiency of intellectual, æsthetic, and sensuous enjoyments to satisfy the whole of man’s nature. They mistook the illusive pleasures of self-indulgence for the true joys of life. In abandoning himself to mundane pursuits and gratifications, the man of the Renaissance fancied that he got the utmost good out of this life, and took little thought of any other.

In a corresponding spirit the architect now set himself to the task of producing a luxurious and specious style of palatial architecture, drawing his inspiration from the monuments of imperial Rome, and the sculptor and the painter sought to portray physical beauty as the primary and sufficient end of their art. Their conceptions of this beauty were in part drawn from the remains of the art of classic antiquity that were then accessible. But the ancient works of art known at that time were not those of the best periods of ancient artistic culture. They were, for the most part, works of the decadent Greek schools as represented in Roman copies. Many of these have, indeed, a great deal of sensuous charm, and display much technical refinement; but they are wanting in the nobler qualities that characterize the finest arts of Greece. From the Roman copies of fauns, Apollos, and Venuses that had been preserved in Italy, it was impossible that high inspiration and true guidance should be drawn.

The Fine Arts of the Renaissance are in part a reflection of this decadent art of classic antiquity, and in part an expression of something quite different which was peculiar to the Italian genius at this time. To the man of the Renaissance the classic inspiration was necessarily different from what it had been to the man of antiquity. To the ancient Greek and Roman the pagan ideals had been real, and their inspiration was genuine; but to the Italian of the fifteenth century these ideals could not have the same meaning, or supply a true incentive. After the intervening centuries of Christian thought and experience it was impossible for men to approach the ancient themes in the spirit of the ancients. Thus the Neo-pagan Art of the Renaissance is not wholly spontaneous and sincere. It contains elements that are foreign to the pagan spirit, and not compatible with it. The art of the Renaissance is, in fact, an embodiment of heterogeneous ideas and conflicting aims.

Much has been said of the importance of the Renaissance movement in developing the individual man, and it is true that one of the most marked characteristics of the artistic productions of this time, as contrasted with those of the Middle Ages, is a distinctly individual, or personal, stamp. This is especially marked in architecture. Whereas before, and during, the Middle Ages in particular, architecture had been a communal art, the joint product of companies of men working together on traditional lines, with common aims and aspirations, it was now become very largely an expression of the personal tastes of individuals working independently of each other. The architects of the Renaissance were scholars and artists, newly acquainted with the Roman antique, animated with desire to appropriate what they apprehended of its principles, and at the same time ambitious to achieve personal fame. A building of the Renaissance is thus always the product of the fancy of a particular designer, as a building of the Middle Ages is not. But architecture of the highest excellence can hardly be produced by an individual working independently. The noblest architecture of the past has always been an evolution of a people, the joint product of many minds, and the natural expression of many conditions. The importance of the opportunity for the development of the individual opened by the Renaissance has been exaggerated, and the conditions conducive to such development which had existed before have been too much overlooked. We are apt to forget that the mediæval communal life stimulated the faculties of the individual in many noble ways, and we do not always enough consider that individuality may be exercised in harmful as well as in salutary directions. The individuality that had been developed by the institutions and the intellectual life of the Middle Ages was vastly different from that which was produced by the influences of the Renaissance, and it was in many ways more excellent. The individuality of the Middle Ages was obedient to the demands of corporate and coöperative life, while that of the Renaissance was independent and capricious. Conditions favourable to individual development had arisen early in the Middle Ages in connection with organized monastic life. The cultivation of literature, philosophy, and the Fine Arts in the monasteries had given considerable range to the exercise of individual powers,[3] though in limited directions, and the rise of the great communal organizations tended still further to stimulate an admirable individual development. But the individual of the Middle Ages felt himself a part of an organized body from which he derived moral support, and with which he felt that he must coöperate. It was the strong communal spirit, giving unity of purpose to the varied faculties of individuals, that made possible the production of the noble arts of the Middle Ages; and it is as the expression of this unity of purpose coördinating the fine artistic energies of the time, that these arts are preëminently notable. In so far as the development of the individual in the period of the Renaissance differed from that of the Middle Ages, it did so mainly in favouring individual caprice at the expense of harmonious collective effort. The capricious and irresponsible individuality of the time, together with the confused complexity of ideas and aims, gave rise to most of that which is open to criticism in the Fine Arts of the Renaissance.

Nearly all of the architects of this epoch were sculptors and painters. Few of them had ever had a thorough training in architectural design and construction, such as had been general with the members of the great mediæval building corporations; and hardly any of them were endowed with a natural aptitude for logical construction. The artistic genius of the Italian people has, in fact, always been essentially a genius for painting, and the painter’s habits of mind are constantly manifested in the Italian architecture of all epochs. This is especially noticeable in their use of the Orders, which is rarely based on any structural need, but is governed only by the fancy of the designer in seeking to produce a pleasant surface composition. Columns and pilasters, answering to nothing in the real structural scheme of a building, are disposed with no thought save for agreeable lines and rhythmical spacings. Thus they soon came to be used in many novel ways. They were set in pairs, stretched through several stories, embraced by pediments, and varied in countless fanciful ways. In this way the architecture of the Renaissance even more than that of imperial Rome, became a mere surface architecture differing fundamentally from all of the great architectural systems of ancient times, and of the Middle Ages. This is a consideration of capital importance of which too little account has been taken. The unqualified and short-sighted laudation of this architecture by the sophisticated writers of the sixteenth century has been too readily accepted, and a more discriminating judgment cannot fail to alter materially the esteem in which it has been held.

In surveying the history of architectural design with attention to its fundamental principles we shall find that there have thus far existed in Europe but three entirely consistent and distinctive styles; namely, the Greek, the Byzantine, and the Gothic. All other varieties of architecture may be broadly divided into two classes, the one consisting of buildings of transitional character, and comprising all organic and progressive types of Romanesque, and the other composed of styles made up of mixed elements not in process of organic fusion. The first architecture of the second class is that of imperial Rome with its off-shoots, the Christian Roman and the numerous subsequent forms of the basilican type, and the second is the architecture of the Renaissance. When, after studying the architecture of Greece, we come to examine that of Rome, we are at once struck by the incongruous mixture of elements which it exhibits; and although we may be impressed by its grandeur, we are unable to give it our unqualified admiration. In Byzantine art we find Greek, Roman, and Oriental elements, logically modified in adaptation to new uses, and fused into a radically new and distinctive style of entire consistency and great nobility.[4] In the transitional art of western Europe we see the creative genius of Northern races gradually evolving the Gothic style, in which elements derived from the older systems are wholly recreated and assimilated in a wonderful manner, and when we turn from the beauty, and the structural logic, of the consummate Gothic Art[5] to the architecture of the Renaissance, a similar contrast is again apparent.

In one branch of art, however, the best achievements of the Renaissance period command our unqualified admiration; namely, the art of painting. As before remarked, the Italian genius appears to have been primarily a genius for painting, and in this field the conditions all conspired to produce results that were without precedent for excellence, and that still remain unrivalled. Yet here, too, we shall need to discriminate. Italian painting of the sixteenth century presents a variety of phases that are by no means of equal merit, and the noblest forms of it show the least of the essentially Renaissance spirit. The Christian painters of the fourteenth century had laid a foundation on which their successors could build, and this gave a character to much of the art of the Early Renaissance which the dominant influences of the time itself could not give.[6] But the spirit of the sixteenth century was unfavourable to the highest ideals and the most exemplary practice, and, save for the works of a few exceptional men, there were no high achievements in painting after about 1520, except in Venice, where more than elsewhere natural and wholesome conditions had been maintained.

Among the many influences that were stirring the artistic minds of the Renaissance there were two of chief importance, the Neo-pagan revival, and the true intellectual life of the people which was independent of the retrospective movement, and had been growing up through the Middle Ages. The most sterling qualities of the artistic products of the period are due to this intellectual life, and Florentine and Venetian painting, the two most admirable phases of the supreme art of Italy, are the finest expression of this. In other words, it was not the revival of interest in ancient thought and feeling, nor the influence of classic models, so much as the ripened development of the native Italian genius itself, that produced what is most excellent in the Fine Arts of the Renaissance. A consciously retrospective motive can hardly be a vital force in artistic development, and the direct attempt, in so far as such attempt was made, to shape the arts after classic models was an unmixed evil. The native traditions and innate tendencies of the Italian people were enough of themselves to give a strong classic quality to their art. In architecture what of classic feeling was natural to them needed only in the fifteenth century to be freed from the elements which had been misappropriated from the mediæval art of the North to allow it true expression in forms adapted to their needs. In normal human progress each successive stage of development creates its own appropriate forms; but peoples, like individuals, sometimes pass through periods of partial aberration, and while genius may still find scope enough, as in the Renaissance, to produce much that is admirable, the noblest forms of art are not an outgrowth of such conditions.

CHAPTER II
THE DOME OF FLORENCE

The great dome of the cathedral of Florence marks the beginning of the Renaissance movement in architecture, though in its general form and structural character it has no likeness to ancient domes, and has few details drawn from the Roman classic source. It exhibits a wide departure from any previous forms of dome construction, and is an expression of the creative genius of a remarkably gifted man of great independence, working under inspiration drawn in part from ancient sources, in part from mediæval building traditions, and in still larger part from the new motives that were beginning to animate the artistic ambitions of the fifteenth century.

The dome of the Pantheon and the dome of St. Sophia, the two greatest domes of former times, had been built on principles that did not admit of much external effect, and the numerous smaller ones of the Middle Ages, in western Europe, had been equally inconspicuous externally, if not entirely hidden from view, in consequence of rising from within a drum which reached far above the springing level. In most cases the whole construction was covered with a timber roof, so that from the outside the existence of a dome would not be suspected. This was a secure mode of construction, and one that for stability could not be improved; but it did not give the imposing external effect that Brunelleschi sought.

Fig. 1.—Hagia Theotokos.

Fig. 2.—Aachen.

Attempts to make the dome a conspicuous external feature had indeed been made before Brunelleschi’s time. The later Byzantine builders had raised small domes on drums resting on pendentives, and rising above the main roof of the building, but they had still carried these drums up somewhat above the springing of the dome, and had further fortified them with buttresses built over the supporting piers, as in Hagia Theotokos of Constantinople ([Fig. 1]). Thus in such designs the dome still remains partly hidden from view, the drum being the most conspicuous part of the composition. Among the early domes of western Europe is that of Aachen ([Fig. 2]). In this case the drum is carried up far beyond the springing, and is covered with a timber roof which completely hides the dome from external view. The same adjustment of the dome to its drum is, with minor variations of form (the dome being in some cases polygonal on plan, as at Aachen, and in some cases hemispherical) found in most other mediæval domes, and the timber roof over all is likewise common. But in a few cases a different scheme was adopted in which the dome is set on the top of the drum instead of within it. In such cases, however, the drum is low, not rising above the ridge of the timber roof of the nave, and the dome, being unprovided with abutment, is insecure except in so far as it may have a form that is self-sustaining as to thrusts (which removes it from the true dome shape, or may be secured by some kind of binding chain.[7] An example of such a dome occurs on a small scale over the crossing of the cathedral of Pisa ([Fig. 3]). This dome is not hemispherical, its sides rise steeply, and with such moderate curvature as to render it measurably self-sustaining as to thrust.[8] Another instance of a similar scheme, and on a larger scale, is that which appears to have formed a part of Arnolfo’s design for the cathedral of Florence. This dome was never executed, and our knowledge of it is derived from the well-known fresco in the Spanish chapel of Santa Maria Novella.[9] Here both the dome and the drum are octagonal in conformity with the plan of the part of the building which it covers. The outline [(Fig. 4]) is slightly pointed, but the sides are nevertheless so much curved in elevation that a structure of this form would not stand without strong cinctures. It is, however, not unlikely that the fresco painter has given it a more bulging shape than Arnolfo intended. But domes of this character were exceptional in the Middle Ages. The builders of that epoch confined their practice for the most part to the safer form in which the vault is made to spring from within the drum, and is thus necessarily, either in part or entirely, hidden from external view.

Fig. 3.—Dome of Pisa.

Fig. 4.—Dome of Arnolfo.

Fig. 5.—Section of Baptistery.

A remarkable dome of this latter class is that of the Baptistery of Florence, which, though the building has undergone various superficial transformations since its original construction at an early, though uncertain, epoch, has come down to us in essential integrity. This building on plan is in the form of an octagon, and the dome is of corresponding shape, and sprung from a level far below the top of the enclosing walls. In elevation the dome ([Fig. 5]) has a pointed outline, and is covered by a pyramidal roof of stone the upper part of which is incorporated with the dome itself, while beneath the lower portion is a void between the dome and the enclosing wall. The structure has an internal anatomy that is both ingenious and admirable. The span is about 25 metres, and the wall at the level of the springing is over 3 metres thick. Above this the wall (_a_, Fig. 5) rises to a height of about 8 metres. The dome at its base is about 1 metre thick, and its extrados rises vertically to a height of about 2½ metres, leaving an open space between it and the wall of the enclosing drum of 1.26 metres in width. Above this vertical portion the extrados is stepped by several courses of masonry, somewhat after the manner of the dome of the Pantheon. From the reëntrant angles of the octagon (_a_, Fig. 6) solid abutments are built up against the salient angles of the vault, and, between these, two secondary abutments (_b_) are carried up against each of its sides. These buttresses are in the form of cross walls dividing the space on each side of the octagon into three compartments, and over each of these compartments a barrel vault, on an axis inclined in conformity with the slope of the roof, is turned. The upper ends of these vaults intersect on the surface of the dome, as shown in Figures 5 and 6. The voids between the crowns of these vaults and the buttresses are filled in with masonry so as to form the sloping planes of the roof below where it is incorporated with the dome, and on these are laid the slabs that form the external covering. With such an effective buttress system as is here provided it is hard to find a reason for the chain of timbers which is inserted at the haunch of the dome. The constructive principle embodied in this monument is altogether sound,[10] and its architectural character is in keeping with the construction.[11]

Fig. 6.—Dissection of the vault of the Baptistery.

Such were the models of mediæval dome building accessible to Brunelleschi when he was forming his great scheme for the covering of the octagon of the cathedral of Florence. But the idea of a low dome, or a hidden dome, could not meet the wishes of the Florentines of the fifteenth century. Their civic pride and large resources called for an imposing design which should make the dome a dominant architectural feature of their city. It was decided that it should be raised upon the top of a high drum, and the task to which Brunelleschi applied himself was to fulfil this requirement.

Of the vast and soaring dome which he succeeded in erecting many opinions have been held, but all beholders are impressed with its grandeur. It has been common to speak as if the master had been chiefly inspired by the ancient monuments of Rome, and had taken the Pantheon as his principal model.[12] But although he came to his task fresh from the study of the ancient Roman monuments, and undoubtedly had the Pantheon much in mind, yet the dome which he produced has little in common with that great achievement of imperial Roman constructive skill. In general it follows, though with great improvements as to outline and proportions, the scheme of Arnolfo as illustrated in the fresco of the Spanish Chapel; but the model to which it most closely conforms, notwithstanding the obvious and essential points of difference, is that of the Baptistery just described. There can, I think, be little question that this monument supplied the chief inspiration and guidance to both Arnolfo and Brunelleschi. A comparison will show that the dome of the cathedral, with its supporting drum, is, in fact, little other than a reproduction of the Baptistery of San Giovanni in a modified form, and enlarged proportions, raised over the crossing.

Plate I

DOME OF BRUNELLESCHI

Florence

Fig. 7.—System of the dome.

But while taking the scheme of the Baptistery as the basis of his own scheme, Brunelleschi was obliged to make some daring changes in order to give his design the external character which he sought. This great dome (Plate [I]), like that of the Baptistery, is octagonal in plan and pointed in elevation. It rises from the top of the octagonal drum, and consists of two nearly concentric shells of masonry, with an interval between them. Eight vast ribs of stone rise from the angles of the drum and converge on the curb of an opening at the crown. These ribs extend in depth through the whole thickness of the double vault and unite its two shells. Between each pair of these great ribs two lesser ones are inserted within the interval that divides the two shells, and nine arches of masonry, lying in planes normal to the curve, are sprung between the great ribs and pass through the lesser ones on each side of the polygon (Figs. [7] and [8]), while a chain of heavy timbers (_a_, Fig. [8], and Fig. [9]), in twenty-four sections clamped together at the ends with plates of iron, binds the whole system between the haunch and the springing. So much of the internal structure can be seen in the monument itself, but further details are described in Brunelleschi’s own account of what he intended to do.[13] From this we learn that the base of the dome, which was to be built solid to the height of 5¼ braccia, was to consist of six courses of long blocks of hard stone (_macigno_) clamped with tinned iron and upon this were to be chains of iron.[14] Mention is also made of a chain of iron over the timber chain (“in su dette quercie una catena di ferro”); but no such chain is visible in the monument, and if it exists, it must be embedded in the masonry of the vault, like the chains at the base.

Fig. 8.—Section.

Fig. 9.—Part Plan.

It will thus be seen that while Brunelleschi’s scheme is essentially different from that of the Baptistery, its structural system is little more than an ingenious modification of it. The parts of the one answer to those of the other with singular completeness. The attic wall and pyramidal roof of San Giovanni are transformed into the external shell of the cathedral dome, the angle buttresses of the older monument become the great angle ribs of Brunelleschi’s vault. The intermediate abutments of the Baptistery are changed into the intermediate ribs of the great dome, and the inclined barrel vaults of the Baptistery scheme are represented in the cathedral dome by the arches sprung between the great angle ribs.

It has been thought by some writers that the rib system of the dome of Florence gives the structure a somewhat Gothic character, and it is sometimes called a Gothic dome.[15] But there can be no such thing as a Gothic dome. It is impossible for a dome of any kind to have the character of a Gothic vault. The difference between the two is fundamental. A Gothic vault is a vault of concentrated thrusts, and it requires effective concentrated abutments. A dome is a vault of continuous thrust, and for sound construction it requires continuous abutments, as in the Pantheon. Whatever use the ribs of Brunelleschi’s vault may have, they do not, and cannot, perform the function of the ribs in Gothic vaulting. Their use is to strengthen the angles of the dome, and to augment its power of resistance to the weight of the lantern which crowns it. They do not support the vault as the ribs of a Gothic vault do. Being composed of very deep voussoirs, they have more strength to withstand thrusts, as well as to bear crushing weight, than the enclosing shells have, and thus to some extent they may hold these shells in. But it appears plain that the architect did not feel confidence in their power to perform this function without reënforcement by a chain, or chains, which, in his own words, “bind the ribs and hold the vault in” (che leghino i detti sproni e cingano la volta dentro). However this may be, the ribs of a dome cannot have any function like that of the ribs in Gothic vaulting. The shell of a Gothic vault is not held in by the ribs, nor is it in any way incorporated with them. Both shell and ribs are held in by the buttresses. This point will be considered further in connection with the dome of St. Peter’s.

The whole scheme of this dome was a daring innovation of one man, and in this it differs from former architectural innovations, which were the comparatively slow outcome of corporate endeavour, progressive changes being so gradual that no wide or sudden departures from habitual modes of building were made at any one time, or by any one person.

It was a prodigious undertaking. The span of the dome is nearly a hundred and forty feet, the springing level is a hundred and seventy-five feet above the pavement, and the height of the dome itself, exclusive of the lantern, is about a hundred and twenty feet. Such a project might well appall the most courageous of building committees, and we need not wonder that the Board of Works drew back in dismay when it was first laid before them.[16]

The successful accomplishment of the work, and the stability which it has thus far maintained, show that the architect was a constructor of great ability,[17] and the fact that he managed to raise the vast fabric without the use of the ponderous and costly kind of centring that had been commonly employed in vaulting, makes the achievement still more remarkable. The precise manner in which he did this is not clear, but of the fact there appears no question.[18]

The dome of Florence is indeed a remarkable piece of construction, and it is no less remarkable as a work of art. In beauty of outline it has not, I think, been approached by any of the later elevated domes of which it is the parent. Yet with all of its mechanical and artistic merit, the scheme is fundamentally false in principle, since it involves a departure from sound methods of dome construction. A bulging thin shell of masonry on a large scale cannot be made secure without abutment, much less can such a shell sustain the weight of a heavy stone structure like the lantern of this monument, without resort to the extraneous means of binding chains. A builder having proper regard for true principles of construction in stone masonry would not undertake such a work. For although it may be possible to give the dome a shape that will be measurably self-sustaining as to thrusts, as Brunelleschi clearly strove to do,[19] it is not possible to make it entirely so, and therefore if deprived of abutment it must be bound with chains. But a structure of masonry which depends for stability on binding chains is one of inherent weakness, and thus of false character.[20]

From these considerations it appears to me that Brunelleschi led the way in a wrong direction, notwithstanding the nobility of his achievement from many points of view. And in following his example modern designers of elevated domes have wandered still farther, as we shall see, from the true path of monumental art.

Moreover, when we consider that a dome set within its drum is not only stronger, but that it is also much better for interior effect, the dome of the Pantheon still remaining the grandest and most impressive arched ceiling of its kind in the world, the unbuttressed modern domes, with their manifold extraneous and hidden devices for security, appear still less defensible.

But in the architectural thought of the Renaissance little heed was given to structural propriety or structural expression, and the Italian writers, who have largely shaped our modern architectural ideas, have not only failed to recognize the inherent weakness of such a building as the dome of Florence, but have even considered the work praiseworthy on account of those very characteristics which make it weak. Thus Sgrilli lauds Brunelleschi for having had the “hardihood to raise to such a height the greatest cupola which until its time had ever been seen, upon a base without any abutments, a thing that had not before been done by any one.”[21] And Milizia says, “It is worthy of special notice that in the construction of this cupola there are no visible abutments.”[22]

As to the permanent stability of this dome various opinions have been held by the experts among the older writers.[23] Its form is, as we have seen, as favourable to stability as it would be possible to make that of any vault which could be properly called a dome. It appears to the inexperienced eye as stable as a crest of the Apennines. Every precaution as to material and careful workmanship seems to have been taken to make it secure. The wall of the drum on which it rests is five metres in thickness, and the solid base of the dome itself is built, if the architect’s scheme was carried out as he had stated it before the Board of Works, of large blocks of hard stone, thoroughly bonded and clamped with iron. The lower system is sufficiently strong, and appears to rest on a solid foundation. But nevertheless there are ruptures in various parts of the structure which have caused apprehensions of danger,[24] and its future duration must be regarded as uncertain. The writers who have maintained that it is secure have argued on the assumption that the parts of a dome all tend toward the centre.[25] These writers overlook the fact that the force of gravity above, especially when the dome is heavily weighed by a lantern, neutralizes the inward tendency of the lower parts and causes a tendency in those parts to movement in the opposite direction. This neutralizing force is lessened by giving the dome a pointed form, as Brunelleschi has done, but, as before remarked (p. [22]), it can hardly be overcome entirely so long as any real dome shape is preserved.[26]

It may be thought that the object which Brunelleschi had in view, of producing a vast dome that should be an imposing feature of the cathedral externally, justifies the unsound method of construction to which he resorted (the only method by which the effect that he sought could be attained). But structural integrity is, I think, so fundamental a prerequisite of good architecture that in so far as this gifted Florentine was obliged to ignore sound principles of construction in order to attain an end not compatible with such principles, the result cannot be properly considered as an entirely noble and exemplary work of art, however much beauty and impressiveness it may have.

The example set by Brunelleschi was, in point of construction, a pernicious one, and bore fruit of a still more objectionable character in the works of other gifted men less scrupulous than he, and less endowed with mechanical ingenuity, as we shall see farther on.

Though there is nothing whatever of classic Roman character in this great dome, the lantern which crowns it, built from Brunelleschi’s design after his death, has classic details curiously mingled with mediæval forms. Its eight piers are adorned with fluted Corinthian pilasters surmounted by an entablature, while the jambs of the openings have engaged columns carrying arches beneath the entablature in ancient Roman fashion. From the entablature rises a low spire with finials set about its base, and flying buttresses, adorned with classic details, are set against the piers. None of the classic details have any true classic character, nor has the ornamental carving, with which the composition is enriched, any particular excellence either of design or execution. But these details are invisible from the ground, and in its general form and proportions the lantern makes an admirable crowning feature of this finest of Renaissance domes.

CHAPTER III
CHURCH ARCHITECTURE OF THE FLORENTINE RENAISSANCE

No other work by Brunelleschi is comparable in merit to the great dome of the cathedral. None of his other opportunities were such as to call forth his best powers, which appear to have required great magnitude to bring them into full play. In his other works the influence of his Roman studies is more manifest, and his own genius is less apparent. In these other works he revives the use of the orders, and employs them in modes which for incongruity surpass anything that imperial Roman taste had devised.

Fig. 10.—Plan of the chapel of the Pazzi.

Fig. 11.—Section of vault of the Pazzi chapel.

The first of these works is the small chapel of the Pazzi in the cloister of Santa Croce. It is a simple rectangle on plan (Fig. [10]), with a square sanctuary on the short axis, and a porch across the front. The central area is covered with a circular vault which by most writers is called a dome, but it is not a dome; it is a vault of essentially Gothic form, like two early Gothic apse vaults joined together (Fig. [11]). It rests on pendentives, and is enclosed by a cylindrical drum, which forms an effective, though not a logical, abutment to its thrusts, and is covered with a low-pitched roof of masonry having a slightly curved outline. Whether this external covering is connected with the vaulting in any way above where it parts from the crowns of the vault cells it is impossible to discover, because there is no way of access to the open space between the two parts. Through a small opening in the outer shell, near its crown, the hand may be thrust into the void, but nothing can be reached. It is a curious form of double vault, and differs fundamentally from the great double dome of the cathedral. The scheme as a whole is structurally inconsistent; for while the inner vault has the concentrated thrusts of Gothic construction, these thrusts are met by the enclosing drum, and not by the isolated abutments that the vault logically calls for. The sanctuary has a small hemispherical dome on pendentives, and the portico is covered with a barrel vault bisected by another small dome on pendentives.

Fig. 12.—Interior of the Pazzi chapel.

The architectural treatment of the interior (Fig. [12]) exhibits a wide departure from that of any previous type of design. The form of the building is mediæval, being, with exception of the central vault, essentially Byzantine,[27] but the details are classic Roman, and consist of a shallow order of fluted Corinthian pilasters with the entablature at the level of the vaulting imposts. In such a building, however, and used in this way, a classic order is out of place; for an order is a structural system designed for structural use, but the order here has no more structural function than if it were merely painted on the walls. It is used, of course, with a purely ornamental motive, but as ornament it is inappropriate. A proper ornamental treatment of such an interior would be either by marble incrusting, mosaic, or fresco, or else by pilaster strips, or colonnettes, and blind arches, which would break the monotony of the broad wall surfaces without suggesting an architectural system foreign to the character of the building. Such arcading would have an appropriate structural suggestiveness, if not an actual structural use; but a classic order is unsuitable for a building of mediæval character. The mediæval pilaster strip and blind arcade were designed for this use, and they have the further advantage that their proportions may be indefinitely varied to meet varied needs, as the proportions of the classic orders may not. But in their lack of a true sense of structural expression, and in their eagerness to revive the use of classic forms, the designers of the Renaissance failed to consider these things.

A particularly awkward result of this improper use of an order is that the entablature passes through the arch imposts, making an irrational structural combination. This scheme was, however, extensively followed in the subsequent architecture of the Renaissance, but it is a barbarism for which no authority can, I believe, be found in ancient Roman design.[28] The nearest approach to it in Roman art is the entablature block resting on the capital (as in the great hall of the Baths of Caracalla), which was a blundering device of the later Roman architects. The complete entablature running through the impost, as in the chapel of the Pazzi, sometimes, indeed, occurs in the early churches of Rome and elsewhere,[29] as a result of unsettled conditions of design, while the architects were struggling with the traditional use of the entablature and the introduction of the arch sprung from the columns. But after the admirable logic of the mediæval arched systems of construction had been reached it appears strange that any designer should go back to this irrational combination.

Fig. 13.—Façade of the Pazzi chapel.

In the portico (Fig. [13]) the incongruities of design are of a still graver nature because they involve weakness of construction. The order of the interior is, as we have just seen, but a simulated order, and has no structural function, but in the portico a real Corinthian order is made to carry the barrel vault and dome above mentioned, and an attic wall which encloses the vaulting. But a classic order was never intended for such use, and cannot properly perform it. Such an order is adapted only to the support of crushing weight, and has no power of resistance to the thrusts of vaulting. The weight of the attic wall tends, indeed, in some measure to neutralize the force of the vault thrust, but this is not enough to render the structure secure, and unless the order were effectively steadied by some extraneous means the attic load would constitute a source of danger, as with any disturbance of its equilibrium by thrust its weight would hasten the overthrow of the system. How it is actually maintained is not apparent. No tie rods are visible beneath the vault, such as are common in Italian vaulted structures, which are rarely buttressed in an effective manner. Ties or clamps may, however, be concealed within the attic, though they would be less effective so placed. But in whatever way the system is held together, it is bad architecture, because the parts have no proper adaptation to their functions.

The ornamental treatment of the attic wall is worthy of notice. The surface is divided into panels by diminutive pilasters, and these panels are subdivided by mouldings in a manner which recalls the treatment of the attic of the Baptistery. The coupling of the pilasters was an innovation in the use of classic members, but it enabled the architect to avoid unpleasant proportions in these details. Single pilasters of the same magnitude would be too slender for the deep entablature over them, or to harmonize with the great Corinthian order below, while wider single ones would be stumpy and inelegant. The pair give good proportion in the total composition, while each pilaster is well proportioned in itself. Another noticeable point is the manner in which the central archivolt and the archivolts spanning the ends of the porch intersect the pilasters at the springing. This could not be avoided, because the pilasters cover the whole space on the entablature over the capitals of the columns, and leave no place for the archivolts. Thus the mediæval principle of interpenetration is carried over into the neo-classic design.

It should be observed that the details of this attic are wrought in stucco, so that we have with the beginning of the Renaissance a revival of a common ancient Roman practice of architectural deceit. The great order, however, is necessarily of stone, and its general proportions are good, though the details are poor in design, and coarse in execution.[30]

Fig. 14.—Badia of Fiesole.

The façade of the Pazzi has been considered as showing noteworthy originality of design. But there are older buildings in the neighbourhood to which it bears enough likeness to suggest its derivation from them. The façade of the Badia of Fiesole (Fig. [14]) is one of these. By substituting a free-standing colonnade for the blind arcade of this front, and breaking its entablature and attic wall with an arch, we should get the leading features of the Pazzi front. Sant’ Jacopo Soprarno, with its attic surmounting an open portico having an arcade on Corinthian columns, is also strongly suggestive of the same scheme. The features that are peculiar to the Pazzi, the arch breaking the entablature, the barrel vault sprung from the order, and the dome bisecting this vault, do little credit to the architect as a consistent designer.

Fig. 15.—Impost of San Lorenzo.

Two more important examples of church architecture in Florence, which appear to be mainly by Brunelleschi, are San Lorenzo and Santo Spirito. What part Brunelleschi had in the design of San Lorenzo is not perfectly clear,[31] but the main scheme was probably his, though the work was not completed until after his death. In the old sacristy of this church, which appears to be the part that was first built, the interior design of the Pazzi chapel is reproduced with some modifications of proportions and details, including the celled vault on a system of ribs, resting on pendentives. The church itself exhibits a frank return to primitive basilican forms and methods of construction, though with modifications and some additions. The nave has a flat wooden ceiling, but the aisles are covered with domical vaulting on salient transverse ribs, and over the crossing is a hemispherical dome on pendentives. In the arcades, which are carried on Corinthian columns like those of the portico of the Pazzi, the entablature blocks of late Roman design are reproduced in the impost (Fig. [15]). The revival of this meaningless feature shows again how little impression the logic of mediæval art had made on the Italian mind, and what lack of discrimination in their borrowings from the antique the designers of the Renaissance often show. Whatever features the Roman models displayed were looked upon as authoritative, and copied without question; and the frequency with which this superfluous detail was reproduced in the subsequent architecture of the Renaissance has given it wide acceptance in more recent times. Notwithstanding the intention of the designer to revive the ancient style, mediæval features are conspicuous in San Lorenzo, and something of the mediæval logic of structural adjustment occurs in some details. Not only is the dome over the crossing supported on pendentives, which, in their developed form, are mediæval features and thus foreign to classic Roman design, but the piers sustaining this dome are compound, and consist of members of different proportions adjusted in the organic mediæval manner. The members which take part in the support of the aisle arcades are necessarily short, while those which carry the great pendentive arches are lengthened to reach the higher level from which those arches spring. But all of these members have the form of fluted Corinthian pilasters (Fig. [16]). Thus were classic members used in ways that are foreign to classic principles, and their proportions altered with as much disregard for the rules of Vitruvius as the mediæval builders had shown.

Fig. 16.—Crossing pier of San Lorenzo.

The church of Santo Spirito, built after the architect’s death, closely resembles San Lorenzo in its architectural character, though it is larger in scale. The entablature blocks occur in the arcades here also, but instead of a dome over the crossing as in San Lorenzo, there is a circular celled vault on converging ribs, like the vault of the Pazzi chapel. The interior is spacious and finely proportioned, but it presents no features that afford further illustration of the progress of neo-classic design.

The retrospective movement was carried further by the Florentine scholar and architect Leon-Batista Alberti, who, says Milizia, is justly regarded as one of the principal restorers of the architecture of antiquity.[32] His chief designs in church architecture are found in Santa Maria Novella of Florence, in San Francesco of Rimini, and in Sant’Andrea of Mantua. The first two of these are mediæval structures in which Alberti’s work is confined to the remodelling of the exteriors, but the last was wholly designed by him, though the work was not completed within his lifetime, and the dome over the crossing is the work of another architect of a later time.

How much Alberti did to the façade of Santa Maria Novella, the part of the building to which his work is confined is not very clear. Vasari speaks vaguely as if the whole front were by him,[33] but from a foot-note by Milanesi it would appear that he merely completed a part which had been left unfinished by an older architect, and the work remaining by the older architect is said to include all below the first cornice except the central portal, which is attributed to Alberti. Milizia says[34] that although it is common to attribute the whole façade to Alberti, it has too much Gothic character to be entirely by him, and that therefore a part of it may, with more probability of correctness, be assigned to Giovanni Bettini, an older architect; but he adds that the central portal is undoubtedly by Alberti.

Fig. 17.—Façade of Santa Maria Novella.

An examination of the monument itself would seem to show that the part below the first entablature, with exception of the great Corinthian columns and the central portal, is mediæval work (Fig. [17]). The whole Corinthian order, with the angle pilasters and the pedestals on which the order is raised, look like neo-classic work, and are probably by Alberti. This order is wholly different in character from mediæval design, and quite foreign to the mixture of Pisan Romanesque and Italian Gothic features of the distinctly mediæval part with which it is associated. The columns of the order are, however, of mediæval proportions, being eleven or twelve diameters in height, and they are built of small stones in a common mediæval manner. But these proportions were necessitated by the older work to which the order had to be adjusted, and the small masonry of which they are composed makes them harmonize with the older parts. The central portal has a round arch on fluted Corinthian pilasters framing in a deeply recessed rectangular opening with a classic lintel and jamb mouldings. It is noticeable that the arch does not spring directly from the capitals of the pilasters, but that the entablature block is interposed, as in Brunelleschi’s arcades of San Lorenzo and Santo Spirito. Milizia, speaking of this feature in another work by the same architect, says: “In these arcades Alberti observed a rule always followed in the good ancient times, but since universally disregarded. The arches are not sprung from the columns, because this would be incorrect, but architraves [_sic_] are interposed. It would now be ridiculous to inculcate the importance of this rule, which is familiar to children.”[35] This, like other notions to which the Renaissance gave currency, is a mistake. In inserting the entablature block at the arch impost Alberti did not follow a rule always observed in the ancient times. This feature is uncommon in ancient Roman art. It was, as before remarked, introduced by the late Roman architects, who, being accustomed to the use of the entablature over the column in the trabeate system which they had borrowed from the Greeks, did not see, when they began to spring arches from columns, that the entablature had no longer any reason for existence. The radical nature of the change wrought in architecture by the introduction of the arch was never grasped by the imperial Roman designers. First framing the arch with an order, thus uniting two contradictory systems, they afterwards, when, as in the basilica of Maxentius, springing the arches of vaulting from columns, thought that the rules required them to crown these columns with bits of entablature.

This façade appears to have been originally designed in the Pisan Romanesque style, with a tall, shallow blind arcade on pilaster-strips reaching across the ground story. But the Romanesque character was modified in some details, the portals having pointed arches, pointed arched niches sheltering tombs being ranged in the intervals between the pilaster-strips. How far the upper part of this façade had been left incomplete until Alberti took it in hand we have no means of knowing; but no mediæval features occur in it as it now stands, except the circular opening in the central compartment. Upon this front, then, Alberti appears to have ingrafted the great Corinthian order, placing a wide pilaster paired with a column on each angle, breaking the entablature into ressauts to cover the columns, which have nothing else to support, and replacing the central portal with the existing one in the revived classic style. The preservation of the greater part of the mediæval work in the ground story made it impossible to get in more than the four columns in the great order, and these are necessarily spaced in an unclassic way, with a narrow interval in the middle and very wide ones on either side. To the upper compartment the architect has given an order of pilasters surmounted by a classic pediment, and flanked by screen walls over the aisle compartments in the form of gigantic reversed consoles, apparently the first of these features which became common thereafter in Renaissance fronts. The pilasters of this order are again four in number, and are set in pairs on either side of the circular opening, the width of this opening making it impossible to space them otherwise. We thus have in the clerestory compartment of this façade a forced arrangement of pilasters, which may have led to that alternation of wide and narrow intervals that became very common in the subsequent architecture of the Renaissance. The attic over the ground story, which extends across the entire front and answers to nothing in the interior of the building, is presumably also by Alberti.

The front of Santa Maria Novella is notable as the first mediæval one which was worked over by a Renaissance architect, and as a whole, notwithstanding that it is a patchwork of incongruous elements, it exhibits a remarkable unity of effect. The merit of Alberti’s work here consists in its quietness. The applied orders are in low relief, their details are unobtrusive, and the mellowing effect of age on the beautiful marble incrusting has fused the whole front into an exquisite colour harmony that is almost unmatched elsewhere.

Very different is the west front of San Francesco of Rimini, in which Alberti has introduced a Roman composition without any admixture of mediæval elements. It is substantially a reproduction of the arch of Septimius Severus. The details are in higher relief here in conformity with the ancient model, and the ressauts of the entablature become correspondingly more salient. A ressaut of this kind is another feature of Roman art which has no justification on structural grounds, and to which there is nothing analogous in any reasonable style of architecture. To set a useless column in advance of an entablature and then make a ressaut to cover it, is irrational.

Plate II

SANT’ANDREA

Mantua

Alberti’s capital work in church architecture is Sant’Andrea of Mantua, begun in 1472, the year of the architect’s death, in which he made a frank return to Roman models in the structural forms of the whole edifice, as well as in the ornamental details—a thing that was rarely done by the architects of the Renaissance. The plan (Fig. [18]) is, however, cruciform, and the dome over the crossing is supported in the Byzantine manner on pendentives. The nave (Plate [II]) has a barrel vault on massive square piers connected by arches, the intervals between the piers forming side chapels, and the lower part of each pier having a small square chamber within it, so that it does not look as massive on the plan as it does in elevation. The east end has the strictly Roman form of a semicircular apse with a half-dome vault. The details of the interior consist of a single order of pilasters, on high pedestals, set on the angles of the piers, and of rich Roman coffering on the surfaces of the vaulting. The piers closely resemble those of the so-called arch of the Silversmiths in Rome, which it is not unlikely that Alberti had in mind in designing them, inasmuch as he was a devoted student of Roman architectural antiquities. This interior is, I think, one of the very finest that the Renaissance produced. The justness of its proportions, the simplicity of the structural scheme, and the quietness of the ornamental details are all admirable. With the given elements it is hard to see how a better composition could be made; but the incongruity between the structural and ornamental systems, the entirely superficial use of the order, and its unfitness as ornament where it has no structural meaning, are fundamental defects of this as of most other Renaissance designs. The scheme of St. Andrea foreshadows that of St. Peter’s, and was undoubtedly in the mind of Bramante when he was preparing his colossal project for Pope Julius II.

Fig. 18.—Sant’Andrea, Mantua.

Fig. 19.—Façade of Sant’Andrea, Mantua.

The west front of this church (Fig. [19]) is again an adaptation of a Roman triumphal arch design. It is, in fact, as the plan (Fig. [18]) shows, a great porch set against the true front, and has no correspondence in its parts with those of the building itself. In outline it is an unbroken rectangle crowned with a pediment. A very shallow order of Corinthian pilasters divides it into a wide central bay and two narrow ones. A great arch over a smaller order opens into a barrel-vaulted recess, on the three sides of which the entablature is returned. A rectangular portal, with square jambs and a cornice, opens into the nave, and an arch reaching to the entablature opens into the lateral compartment on each side, and each of these compartments has a barrel vault with its axis perpendicular to that of the great central one. The entablature of the small order is carried across the front of each lateral bay, dividing it into two stages, and the great order rises through it, embracing both stages, and forming an early instance of the so-called colossal order that became common in the later Renaissance. The great order is raised on pedestals, and both pedestals and pilasters of this order are panelled, while the small order rests directly on the pavement and its pilasters are fluted. It is noticeable that the design of the central arch is almost exactly like that of the central portal of Santa Maria Novella in Florence, the smaller entablature being broken into shallow ressauts over the pilasters, giving the same character to the imposts. The front as a whole has the quiet and refined character that distinguishes this architect’s work in general.

Fig. 20.—Arch of Septimius Severus.

That Alberti derived all of these façades, and especially that of St. Andrea, from the Roman triumphal arch scheme a direct comparison will show; and the arch of Septimius Severus (Fig. [20]) may, I think, be taken as the model that he had chiefly in mind. In Santa Maria Novella the mediæval scheme upon which he had to fit his work prevented such a disposition of the columns, and such general proportions as this model exhibits. He was obliged to make the lateral intercolumniations much wider than the central one, and to make the whole rectangle of the composition more oblong than that of the ancient monument; but in most other points he has followed the arch of Septimius Severus closely. As in the Roman design the entablature crowns the wall instead of the order, so that ressauts have to be formed to cover the columns. The insertion of the angle pilaster is a departure from the Roman scheme, and the placing of the stumpy pilaster of the attic over the great pilaster, instead of on the column, is another point of difference. But the general scheme of the ground story and attic will be seen to resemble that of the Roman design about as closely as the mediæval edifice on which it is ingrafted would allow. In San Francesco at Rimini the architect had a freer hand, and the order is treated in closer conformity with the ancient model as to the spacing of the columns and other details. The angles are treated precisely as they are in the arch of Septimius Severus, the pilasters being omitted, and the entablature at each end extending beyond the ressaut. The attic is omitted here, and the unfinished upper part of the façade is necessarily of different design.

In St. Andrea at Mantua the use of pilasters instead of columns, and the absence of ressauts in the great order, as well as the substitution of a pediment for the attic, make a great difference in the general character of the design; and yet the triumphal arch idea is even more strongly marked in this case, because it is not confined to the mere façade but extends to the form of the whole porch. The great barrel-vaulted recess is an exact reproduction of the central passageway of the Roman arch, and so are the lesser arches which open out of this recess on either side.

The triumphal arch idea applied to church fronts appears to be peculiar to Alberti. Most other architects among his contemporaries and immediate successors limit themselves to an application of the orders variously proportioned and disposed. In some cases the mediæval scheme of buttresses dividing the front into bays is retained, and this scheme is enriched with pilasters, or columns, and mouldings of classic profiling, as in the façade of the Duomo of Pienza by the Florentine architect Rossellino. In the later Renaissance façades, as we shall see, there is frequently no organic division of the whole front into bays by continuous members embracing its whole height, but superimposed pilasters and entablatures are variously disposed upon the surface without any suggestion, in the composition as a whole, of the triumphal arch idea (as in Vignola’s fronts, Figs. 49 and 50). But in the characteristic Palladian scheme an organic division is formed by a great order of columns reaching to the top of the nave compartment, and overlapping a smaller order of pilasters extending across the whole front as in Figure 54.

The foregoing examples are enough to illustrate the character of Florentine church architecture, and that which was wrought elsewhere under Florentine influence, in the fifteenth century. These examples show us that the designers, while ostensibly striving to revive the antique forms, were in reality working more or less unconsciously on a foundation of mediæval ideas from which they could not free themselves. The inconsistencies of their work are largely due to the irreconcilable nature of the elements which they sought to unite, not appreciating the logic of mediæval art on the one hand, nor the true principles of the best art of antiquity on the other. The classic orders were entirely unsuited to the buildings to which they affixed them. They properly belong to a very different type of architecture which had been developed by the Greeks in ancient times, and the Greeks alone have used them with propriety. The Romans misapplied and deformed them, and the Italians of the Renaissance now surpassed the Romans in their misapplication and distortion. Many further illustrations of this will appear as we go on.

Early in the sixteenth century this architecture began to assume another phase in which the mediæval elements became less conspicuous, though they were not eliminated, and the imperial Roman features were more rigorously reproduced, yet they were never used with strict conformity to ancient models. This phase of the art was inaugurated by the architect Bramante after his settlement in Rome. We shall consider the Roman work of Bramante, in the following chapter.

CHAPTER IV
THE DOME OF ST. PETER’S

When in the year 1503 Pope Julius II came to the papal chair, the architect Bramante had recently settled in Rome. Born in Urbino, he had spent his early manhood in the North of Italy, where he had come under the influence of the Florentine architect Alberti at Mantua, and of the early Renaissance masters at Milan and elsewhere. Under these influences he had acquired a style that was peculiar to the North at that time. But since coming to Rome he had begun to form a new manner under the more direct influence of the Roman antique,[36] and he soon developed a style in which the ancient Roman forms were reproduced with stricter conformity to the ancient usage, and with smaller admixture of mediæval features than had before prevailed.

An early work in Rome in which he exhibits this more rigorous classic tendency is the small building known as the Tempietto in the cloister of San Pietro in Montorio. It consists of a circular cella with shallow pilasters, surrounded by a colonnade of the Roman Doric order, and surmounted by a hemispherical dome on a high drum. It is thus in form like a Roman temple of Vesta with its dome raised out of the abutting drum and set upon its top without abutment. A glance at Figures 21 and 22, a part section and part elevation of the temple of Vesta at Tivoli, and an elevation of the Tempietto, respectively,[37] will show how great a change Bramante made in the adjustment of the vault to the supporting drum, while it will show also the essential likeness in other points between the two monuments. In Figure 21 it will be seen that the vault is well abutted by the roof of the portico, and by stepped rings of masonry over the haunch, while in Figure 22 the drum is raised high above the encircling portico, and the vault is sprung from its top, and has no abutting rings. The architect appears to have realized that such a scheme would be unsafe on a large scale, for in the one which he prepared for the dome of St. Peter’s he took care, as we shall see, to provide strong abutment.

Fig. 21.—Temple of Vesta, Tivoli, from Serlio.

The Tempietto is but a modified copy of an ancient model, and in no true sense an original design. The changes wrought by the copyist are not of a creative kind consistent with true principles of building. The pilasters, and the balustrade with which the order of the portico is crowned, are superfluous, and the work as a whole shows little of Bramante’s real ability as an architect. Such merit as it has is primarily due to the ancient model which he would have done better to have reproduced more exactly.

Fig. 22.—San Pietro in Montorio, from Serlio.

But Bramante manifested his real powers in his project for the great church of St. Peter, his capital work, which, however, was never carried to completion. It is well known how Pope Julius II had conceived the idea of erecting a vast tomb for himself, and had employed Michael Angelo to prepare the design. We are told by Vasari[38] that the project submitted by this great artist so pleased the Pope that he determined to rebuild the church of St. Peter in order to make it more worthy to enshrine so magnificent a monument. Under Pope Nicholas V, half a century before, the grand old basilica, that had stood since the time of Constantine, had been partially demolished, and a new edifice on a larger scale begun by the Florentine architect, Rossellino. This work had not progressed very far when it was suspended on the death of this Pope, and operations had not been resumed until now, when Pope Julius resolved to demolish Rossellino’s beginning along with what remained of the old structure, and to make a fresh start with a still grander scheme, which was prepared by Bramante, who began the new work in the year 1506.

There is much uncertainty as to the exact nature of Bramante’s design for the building as a whole. No authentic drawings embodying the definitive project are known to exist, and in the monument itself Bramante did not go far enough to show his whole intention. Even what he actually did cannot be wholly made out with clearness, because so many other hands were employed after his death. The exact form of his plan is uncertain, though there appears little question that it was to be in the form of the Greek cross with towers set in the external angles, and it is certain that a vast dome was to rise over the crossing.[39] The work, though considerably advanced, was not nearly completed when, in the year 1514, the master died. He appears to have built the great piers for the support of the dome, with their connecting arches and pendentives, but not to have begun the dome itself.

Fig. 23.—Bramante’s dome for St. Peter’s, from Serlio.

The scheme was to be a colossal one, and the dome was to be its main feature. We may presume that Bramante naturally shared the universal feeling of admiration for Brunelleschi’s dome, and that he wished to rival its imposing character. But his ardent and intelligent study of the monuments of Roman antiquity had given him a better appreciation of their superior structural merits, and in his project for the great dome he had sought to adhere more closely than Brunelleschi had done to the ancient principles and ancient forms.

In seeking guidance from the antique two monuments in particular appear to have appealed to him as offering appropriate suggestions, the Pantheon and the Basilica of Maxentius, then called the Temple of Peace. The first of these monuments gave the model for a mighty hemispherical vault securely suspended over a vast area, while the second offered an example of a stupendous system of piers and arches. In maturing his great scheme with these models before him, he conceived the idea of uniting their respective sublimities, and is said to have boasted that he would set the Pantheon upon the arches of the Temple of Peace. While it is probable that the majestic elevation of the dome of Florence haunted his imagination, and led him to feel that he must lift his dome high, he wished, at the same time, to give the design a more classic character, and a sounder structural form. In striving to accomplish this double purpose Bramante produced a scheme for an elevated dome of almost thoroughly Roman character, and at the same time of imposing external effect. The architect Serlio gives an illustration[40] (Fig. [23]) of this project which is highly instructive.[41] A comparison of it with the scheme of the Pantheon shows a close likeness in essential forms and adjustments. The points of difference are mainly such as Bramante’s desire to make his dome externally conspicuous would require. In the Pantheon (Fig. [24]) the dome springs from within the massive drum at a level far below the external cornice, so that the wall above the springing forms a solid and powerful abutment, reaching almost to the haunch of the vault. Above this a stepped mass of masonry, diminishing in thickness as it rises, is carried well over the haunch, effectively overcoming any tendency to yield to the force of thrust. A Corinthian order, surmounted by an attic, is carried around the wall of the interior,[42] while the wall on the outside is plain.

Fig. 24.—The Pantheon.

In Bramante’s project every essential feature of this ancient monument is reproduced, but with modifications which give a different aspect to the design as a whole, but do not constitute any such radical departure from the principles embodied in the Pantheon as those wrought by Brunelleschi in adapting the scheme of the Baptistery to that of the dome of the cathedral of Florence. In order to secure greater elevation for external effect, the architect has raised the springing level of the dome considerably, though he has still kept it below the top of the drum. The drum itself is of great thickness, and forms a strong continuous abutment at the springing, and the haunch of the vault is loaded with steps of masonry as in the Pantheon, though not quite so heavily. The lower half of the drum is a solid wall resting on the pendentives, while the upper part, which is less than half as thick (Fig. [25]), is pierced with eight wide openings, and its inner and outer faces are each adorned with an order of pilasters alternating with free-standing columns in the intervals. The upper wall stands on the inner circumference of the massive lower ring, while an encircling order of Corinthian columns is ranged on its outer circumference, and gives an effect of lightness and elegance to the exterior, which, together with the lantern at the crown of the dome, goes far to disguise the real likeness of the whole to the Pantheon scheme.

Fig. 25.—Plan of Bramante’s dome, from Serlio.

In these changes and additions Bramante was governed by a clear understanding of the exigencies of his project. He was obliged to raise the internal order from the place on the ground level which it occupies in the Pantheon, to the upper part of the drum, in order to provide a solid foundation resting on the pendentives; and this compelled him to eliminate the attic story of the Pantheon scheme. The most radical change was that of substituting the open colonnade for the solid wall on the outside of the drum. It is doubtful, I think, whether the drum thus lightened would have had enough strength to withstand the enormous thrusts of such a dome.

Like the dome of the Pantheon, Bramante’s dome was to be hemispherical and to have an opening at its crown. Over this he was to set the lantern which in outline recalls that of Brunelleschi, though it is of lower proportions, in keeping with the less elevated form of his dome, and has a small hemispherical dome instead of a conical roof. The shape of the lantern accords well with the composition as a whole, and contributes much to the aspiring expression which was now demanded, without wholly contradicting the classical spirit that the architect was striving to maintain.

The structural merit of this scheme lies in what it has derived from the forms and adjustments of the Pantheon. Its weakness consists in the increased elevation, lifting the dome away from its abutment to such an extent that it may be questioned whether it could have been made safe without chains. The thrusts of a hemispherical dome are vastly more powerful than those of a vault of pointed outline, like the dome of Florence, but if properly abutted, as in the Pantheon,[43] it is perfectly safe, and makes a better ceiling than a pointed vault. In reducing the efficiency of his abutment by raising the springing of the dome so high, the architect ought to have diminished the force of its thrust in a corresponding degree by giving it a pointed form. This would have made it more safe, but it would have been inconsistent with the classic Roman models to which he was striving to conform.

As for Bramante’s intended architectural treatment of the rest of the building we have, as before remarked, no precise information. It appears, however, most probable that he meant to cover the arms of the cross with barrel vaulting on massive square piers and arches, with a single order of pilasters such as Alberti had used in St. Andrea at Mantua, and such as Michael Angelo actually employed, though in a way peculiar to himself, and probably unlike that in which Bramante would have done it. For Bramante would, I think, have followed Alberti’s example in raising the order on pedestals—the great scale of St. Peter’s especially calling for such treatment. Bramante would have realized that a single order large enough to rest directly on the pavement and allow the entablature to pass over the crowns of the arches of the great arcades, would dwarf the apparent scale of the whole interior, as Michael Angelo’s order actually does. But whatever his intention was as to these details, Bramante died before they could be carried out, and we are left in the dark as to what the church as a whole would have been had he lived to complete it.

To the work of his numerous successors prior to the appointment of Michael Angelo, we need give no attention because their labours did not materially affect the final result. Their work was largely on paper only, and the building as it now exists is substantially Michael Angelo’s design, based on that of Bramante, but with extensive, and damaging, additions by subsequent architects.

Fig. 26.—The model.

Michael Angelo at the time of his appointment as architect, in the year 1546, was seventy-two years of age. He professed great respect for the original scheme of Bramante, yet he radically changed the form and adjustment of its main feature, the dome. In conformity with Bramante’s project, he made the drum massive at the base and thinner above, but in place of Bramante’s external colonnade surmounted by a solid ring of masonry, forming a continuous abutment at the springing of the vault, he substituted a series of sixteen isolated buttresses, and raised the dome so high above them that they do not meet its thrusts at all. The drum is carried up above the buttresses so as to form an attic over the order with which the buttresses are ornamented, and from the top of this attic the dome is sprung. The stepped circles of abutting masonry at the haunch are omitted, and instead of one solid vault shell, such as Bramante intended, Michael Angelo’s project provided for a variation of Brunelleschi’s double vault, and was to include (as the model, Fig. 26, shows) three separate shells.[44] The inner shell was to be hemispherical (Michael Angelo thus showing that he appreciated the superior character of the dome of the Pantheon and that of Bramante’s scheme to the dome of Brunelleschi as to internal effect), while the other two were to be pointed, with diverging surfaces. Following Brunelleschi, he introduced a system of enormous ribs rising over the buttresses of the drum, and converging on the opening at the crown of the vault. These ribs unite the outermost two shells, extending through the thickness of both, and support the lantern.

Of this hazardous scheme only the drum was completed when Michael Angelo died. But the existing dome, which was carried out by his immediate successors, is substantially his design, though the innermost shell of the model was omitted in execution, and the vault was thus made double instead of three-fold (Fig. [30]). This dome does not, however, divide into two shells from near the springing, but is carried up in one solid mass almost to the level of the haunch. Michael Angelo may have thought that this would strengthen it, but the solid part has not a form capable of much resistance to thrust, and the isolated buttresses are located so far below the springing that they contribute practically nothing to the strength of the system, as already remarked, and as we shall presently see.

Although this great dome has been almost universally lauded, it is entirely indefensible from the point of view of sound principles of construction. The work shows that Michael Angelo was not imbued, as Bramante had been, with a sense of the essential conditions of stability in dome building as exemplified in the works of Roman antiquity. He had conceived an ardent admiration for the dome of Florence, and in emulation of it he changed the external outline from the hemispherical to the pointed form, and, lifting it out of the buttressing drum, set it on the top.[45]

_A_  Fig. 27.  _B_

This vast dome is an imposing object, but it is nevertheless a monument of structural error. Not only does its form and construction render it much less secure than Brunelleschi’s dome, but its supporting drum is entirely unsuited to its function, save as to its strength to bear the mere crushing weight of the vault. In replacing the continuous colonnade, with its abutting load, of Bramante’s drum by the isolated buttresses, Michael Angelo ignored the true principle of resistance to the continuous thrusts of a dome. It has been thought that the rib system justifies this, that the ribs gather the thrusts upon the buttresses and give the dome a somewhat Gothic character. But this cannot be so. It is impossible for a dome to have any Gothic character. In addition to what has been already said (p. [20]) on this point, it may be further remarked that, so long as the surfaces between the ribs remain straight on plan, as in the dome of Florence, or are segments of a hemisphere, or of a dome of pointed form on a circular base, like the dome of St. Peter’s, no ribs can be made to act in a Gothic way. A circular vault on Gothic principles would necessarily be a celled vault, more or less like the small vault of the Pazzi already described (p. [27]). In such a vault there would have to be an arch (in a true Gothic vault a much stilted arch) in the circumference of the drum over the space between each pair of ribs. The crowns of these arches would reach to a considerable height, in a developed Gothic vault to nearly or quite the height of the crown of the vault itself. The triangular spaces enclosed by these arches and converging ribs would then be vaulted over by slightly arched courses of masonry running lengthwise of the triangle, or from the arches to the ribs, and approximately parallel with the crown of the cell (_A_, Fig. 27). Thus in place of an unbroken hemispherical or oval vault, we should have one consisting of deep cells. The drum would have to rise far above the springing, and the haunches would need to be loaded with a solid filling of masonry. The vault would thus be completely hidden from view on the outside. Nothing short of this would produce a circular vault on Gothic principles, or one in which the ribs could act in a Gothic way.[46] The nearest approach to such a form, in a vault that may with any propriety be called a dome, occurs over the crossing of nave and transept in the old cathedral of Salamanca in Spain (Fig. [28]).[47] But this vault has a very different character from the imaginary one just described. It rises from the top of a high drum resting on pendentives, and is built on a system of salient converging ribs. The spaces between these ribs are vaulted over with courses of masonry slightly arched from rib to rib, and thus running in a direction perpendicular to that of the courses in a Gothic vault cell, as in _B_, Figure 27. A series of hollowed gores are thus formed which give a scalloped instead of a plain circular plan to the vault as a whole. But such a vault differs fundamentally from a Gothic vault. For the line of the crown of each cell is the steep segmental curve _ab_ in _A_, Figure 29. In other words, the vault as a whole is a hemisphere with its surface broken into shallow hollows like the gores of a melon. It is obvious that in a vault with cells so shaped the thrusts are as great at all points in the circumference as they are in a simple hemispherical dome, and that such a vault can have no Gothic character. To develop this into any real likeness to a Gothic vault, it would be necessary to reduce it to an unbroken circular plan by cutting off the scallops at its base so that it would fit into the circular drum, upon the inner surface of which it would now intersect in series of small arches, one for each hollowed gore, with its springing at the point _d_ and its crown at the point _c_. Then these arches would have to be raised by stilting and pointing until their crowns were brought up to the level, or near the level, of the point _b_ as in _B_ of the same figure. Thus the line _dc_, which represents the height of the arches in the first stage of this development, becomes the line _ac_ in the second stage. So long as the chord of the arc _bc_ is a steeply inclined line, the vaulting cells cannot bear upon the ribs, nor can the thrusts of the vault be concentrated in a Gothic way.

Fig. 28.—Interior of dome of Salamanca.

_A_  Fig. 29.  _B_

The vault of Salamanca is not a Gothic vault in any sense, though its rib system and its hollowed cells conform with the earliest stage of apsidal vault development leading to Gothic.[48] It is a dome, and like the larger dome of St. Peter’s, it is sprung from the top of the drum; but unlike St. Peter’s dome, it is powerfully abutted by turrets and dormers built against its springing and its haunch, and it is loaded at the crown with a cone of masonry, so that from without it looks like a stumpy spire, and not like a dome.[49]

But Michael Angelo’s vault has not even such remote approach to Gothic character as the small dome of Salamanca has. Its surface is unbroken by any hollowing into cells. It is a perfect circle on plan, and its ribs, which are embedded and not salient on the inside, cannot, therefore, sustain the vault in any Gothic way. This dome has, moreover, so much of a spherical shape as to give it a stronger tendency to thrust than the dome of Florence has, and the thrusts are exerted equally on all points in the circumference of the drum. The isolated buttresses are therefore illogical, and being set against the drum only, and not even reaching to the top of the drum, they are ineffectual. Thus though the dome was bound with two iron chains, one placed near the springing, and the other at about half the vertical height of the vault,[50] it began to yield apparently soon after its completion. Fissures opened in various parts of both dome and drum which at length caused such apprehensions of danger, that Pope Innocent XI called a council of the most able engineers and architects of the time[51] to examine into the extent of the damage, and ascertain whether serious danger existed. This council concluded that the cupola was in no danger of disintegration, and the Pope, in order to restore confidence in its safety, charged Carlo Fontana, the architect, to write a book on the building and prove the groundlessness of any fears of its collapse. Thus the matter appears for the time to have been dropped. But subsequently the condition of the structure became so alarming that three eminent mathematicians, among whom was the celebrated Boscovich, were, in the year 1742, commissioned by Pope Benedict XIV to make a further examination and submit a report with recommendations for its consolidation.

The condition of the fabric at the time of this examination will be understood from Figure 30, a reproduction of the illustration subjoined to the mathematicians’ report.[52] They found the structure, as the illustration shows, rent into numerous fissures, some of which were large enough to allow a man’s arm to be thrust through them. In some places these cracks had been filled up with brick and cement, and new ones had opened in the filling.[53] At what time the ruptures had commenced could not be definitely ascertained, but the mathematicians express the opinion, for which they state their reasons, that they may have started very soon after the completion of the work.[54] That they were not due to any weakness in the substructure was shown by the fact that this remained apparently quite firm. Had the fractures been caused by any weakness in the piers or pendentives, the mathematicians say,[55] they would be wide at the base of the drum, whereas they were found (as shown in the illustration) to be small at the base and to increase in magnitude toward the top of the drum, and in the region of the haunch of the dome. This was thought by them to show that they were clearly due to weakness resulting from the form of the structure. The report states[56] that the weight of the lantern had caused the heads of the great ribs to sink, the dome to expand at the haunch and at the springing, and the wall of the drum to be pressed outward at the top. To consolidate the fabric they recommended that additional chains be placed at various levels, the old ones having, they thought,[57] burst asunder by the force of the thrusts; but this could not be verified because they are embedded in the masonry. They also recommended clamps of iron to hold in the buttresses.

CVPOLA DI S PIETRO

Fig. 30.—The dome with its ruptures.

The Marquis Poleni of Padua, a distinguished engineer of the time, was then called to examine the dome and to take such means as he might judge most effective to secure it against further disintegration. Milizia tells us[58] that Poleni, after careful examination, expressed the opinion that although the cupola had not the catenary outline which he considered the best for stability, it had nevertheless a good form, and was in no danger of ruin. He pronounced the fissures of no material consequence,[59] and attributed them to two classes of causes. The first of these he thought was haste in construction. The dome having been built in twenty-two months with materials of unequal quality, and not carefully laid, unequal consolidation resulted and caused numerous ruptures. The second class of causes he considered to consist in the action of heat and cold, humidity and dryness, lightning and earthquake. Thus far, says Milizia, the reasoning of Poleni is just, and worthy of his great mind, and he adds, “It would seem, then, that since the cupola was solid, it ought to have been left in peace, but he nevertheless advised five chains, which were placed under the direction of Vanvitelli.”[60] The first of these chains was located at the base of the drum, the second at the level of the attic, the third at the springing of the dome, the fourth above the haunch, and the fifth around the opening at the base of the lantern.

Plate III

DOME OF ST. PETER’S

Rome

Poleni, in his own book, is not wholly consistent in what he says of the dome. While affirming its form to be admirable and its strength sufficient, and attributing the ruptures to other causes than those inherent in the nature of the design, he yet, in another place, admits that they may be to some extent due to the shape of the vault which he here pronounces not high, or acute, enough.[61] His admiration for Michael Angelo makes it hard for him, as a rule, to find any radical defect in the composition; but this passage, and the fact that he caused the additional binding chains to be applied, would seem to show that he had misgivings, and did not consider the monument so entirely safe as the general tenor of his book would lead us to believe. And the same misgivings are betrayed in what is said by the numerous other writers whose opinions are cited by him, though like himself they write for the most part with a manifest bias in favour of Michael Angelo. Thus one of these writers proposes that the outer covering of lead should be stripped off on account of its weight, and be replaced with copper, to which Poleni objects,[62] affirming that the weight is an advantage, and tends to hold the dome together. Another writer suggests that the lantern be removed in order to relieve the fabric of its weight. Another thinks that the buttresses should be heavily weighted with statues. It was also proposed that additional buttresses should be set against the attic of the drum, and carried up against the dome itself; and again it was proposed that massive abutments be built up over each of the four great piers, but to this it was objected that the additional weight of such abutments would dangerously overcharge the substructure. The most radical suggestion was that both dome and drum be demolished and rebuilt in a more pointed form. All of these suggestions were rejected, and it was finally decided to employ the additional chains proposed by Poleni as already stated.

The dome of St. Peter’s (Plate [III]) was conceived in a grandiose spirit, which, while it drew inspiration in part from the ancient Roman source, recklessly disregarded the lessons which Roman art should teach as to principles of construction. I have said that Brunelleschi led the way in a wrong direction when he set his great dome on the top of its drum, and had resort to clamps and chains for the resistance to its thrusts that should have been given by abutment. In following his example, Michael Angelo wandered still farther from the path of true and monumental art. To make a dome on a large scale a conspicuous object, from the springing to the crown, is a thing that cannot be safely done in stone masonry. To make it stand at all, resort must be had to extraneous and hidden means of support, and even these are of uncertain efficiency for any length of time. The ancient Roman and the Byzantine builders settled, I think, for all time the proper mode of constructing domed edifices. Bramante had recognized this, and while striving to include in his design for the dome of St. Peter’s as much as he could of the new character embodied in Brunelleschi’s dome, he tried at the same time to keep safely within the limits of the principles that had governed the ancient practice. He gave as much elevation to his dome as he thought these principles would allow, but even this, as we have seen, was too much, and in greatly increasing this elevation, so as to leave the dome entirely without abutment, Michael Angelo took unwarrantable risks, and lent his genius to the support of false principles.

That this has not been generally recognized is due to the fact, already remarked, that the architects and leaders of taste of the Renaissance have made too little account of structural propriety, and structural expression, as a necessary basis for architectural design.

Recent writers have ignored the condition of this monument. They do not appear to be aware of it; and although it has been fully set forth, and discussed at great length by the earlier Italian writers, few of them have found the true cause in its flagrant violation of the fundamental laws of stability. They attributed the alarming progress of disintegration, as we have seen, to accidents and circumstances of various kinds; and have sought to shift the responsibility to the shoulders of Bramante. They have affirmed that he did not take enough care to make his foundations secure. There appears to be some justice in this, though since his work was strengthened by his immediate successors[63] the ruptures in the dome cannot, according to the mathematicians, be attributed to this. The remarks of the old writers on Bramante must, I think, be taken with some allowance. Their bias against him is very marked. Thus Poleni quotes Condivi, a disciple of Michael Angelo, as saying, “Bramante being, as every one knows, given to every kind of pleasure, and a great spendthrift, not even the provision given by the Pope, however much it was, sufficed him, and seeking to expedite his work, he made the walls of bad materials, and of insufficient size and strength.”[64]

A great deal has been said of the beauty of St. Peter’s dome. It has been held up as a model of architectural elegance by countless writers from Vasari down. But no abstract beauty, no impressiveness as a commanding feature in the general view of the ancient city that it may have, can make amends for such structural defects. Its beauty has, however, I think, been exaggerated. Its lack of visible organic connection with the substructure makes it inferior in effect to the dome of Florence, where the structural lines of the edifice, from the ground upward, give a degree of organic unity, and the buttressed half-domed apses, grouped in happy subordination about the base of the drum, prepare the eye to appreciate the majesty of the soaring cupola as it rises over them. The dome of St. Peter’s has not the beauty of logical composition. Beauty in architecture may, I think, be almost defined as the artistic coördination of structural parts. As in any natural organic form, a well-designed building has a consistent internal anatomy, and its external character is a consequence and expression of this. The dome of St. Peter’s violates the true principles of organic composition, and this I believe to be incompatible with the highest architectural beauty.

CHAPTER V
CHURCH ARCHITECTURE OF THE ROMAN RENAISSANCE

As for the rest of the church of St. Peter, we need give attention to that part only which was designed by Michael Angelo on the basis of the original scheme of Bramante, namely, all to the eastward of,[65] and including, the first bay west of the crossing. The western bays of the nave as it now stands were, as is well known, added at a later time by the architect Maderna. The plan (Fig. [31]) of the earlier part is thoroughly fine, and if the elevation had been made consistent with this plan, St. Peter’s might have been one of the noblest monuments in Christendom. But the architects of the Renaissance rarely sought consistency in design; they were prone, from first to last, to mix incongruous elements. The essentially Byzantine plan here adopted could not be carried out in elevation with classic Roman details with a noble result; and the attempt which Michael Angelo made to produce an architectural effect foreign to the real structural system led of necessity, not only to such inconsistencies as are common in Renaissance motives, but to some awkward makeshifts which have not, I believe, been hitherto noticed by writers on this edifice.

Following what appears to have been Bramante’s intention, Michael Angelo constructed barrel vaults over the arms of the cross,[66] supporting them on piers and arches which had been begun by Bramante. To this simple and reasonable scheme he applied a colossal order of Corinthian pilasters, a pair against each pier, as Alberti had done on a smaller scale at Mantua, and as Bramante appears to have intended in the great piers of the crossing, if not in all of the others. Apart from the superficial and purely ornamental character of the order, and its inappropriateness as ornament in such a system, its exaggerated scale dwarfs the effect of magnitude in the whole interior. The eye naturally estimates this magnitude by the customary proportions of a large classic order, and while these are by no means fixed, there is an approximate mean scale upon which we base our judgment. No beholder on entering St. Peter’s can, indeed, fail to be impressed with the unusual size of the order; but he is not apt to realize how far it exceeds the largest orders of antiquity. The order of the Parthenon is about forty-five feet high, and that of the portico of the Pantheon is about sixty feet. These are exceptionally large among the orders of Greek and Roman antiquity[67] but the order of St. Peter’s is one hundred feet high.

Fig. 31.—Plan of St. Peter’s, from Fontana.

The lack of due effect of scale in this interior has been often remarked, and it is generally attributed to the great magnitude of the structural parts. The size of these parts could not, however, well be different from what they are. Their magnitude is determined by the scale of the great dome and the width and altitude of the arms of the cross. The piers of the crossing are masses of masonry measuring on their longer sides more than fifty feet on the pavement, while the pendentive arches are one hundred and fifty feet high, and those of the arms of the cross are seventy-five feet high. But with appropriate treatment their scale might have been made more apparent. To adorn such piers and frame such arches with a classic order is to destroy the proper effect of scale, as well as to violate the true principles of architectural design by using structural members without any structural meaning.

Apart from the barbarism already remarked (p. [29]) of springing a vault from a classic entablature, the effect of the gigantic order is unhappy in other respects; the great salience of its cornice cuts off from view the lower part of the vaulting, and this pronounced overhanging ledge, extending around the whole interior, breaks the continuity of the upright lines into the vaulting, and diminishes the effect of altitude.

Fig. 32.—Section of aisle of St. Peter’s.

But not only did Michael Angelo employ this incongruous and ineffective ornamental scheme for the interior of St. Peter’s, he also adopted a corresponding design for the exterior which wholly contradicts the real character of the structure and led the architect into some curious makeshifts. For this exterior he used another gigantic order surmounted with an attic story. This obliged him to carry up the enclosing walls of the aisles to a height equal to that of the nave, and led to difficulties within. For the aisle vaulting was now far down below the top of these walls, and it therefore became necessary, unless the space above this vaulting was to be left open to the sky, with the enclosing wall standing as a mere screen answering to nothing behind it,[68] to construct a flat roof at the level of the attic cornice. Figure 32, a section through this part of the structure, will explain this and some other awkward expedients to which the architect was driven by the use of this colossal external order. Of the two compartments through which the line _AB_ (plan, Fig. 31) passes, one has a barrel vault and the other a dome, and, as each of the other corresponding parts of the plan are vaulted in the same way, there are four small domes in all. The effect of four smaller domes grouped around the great central one would be happy for both internal and external effect, if they were properly related in proportions, and the scheme were carried out in a structurally consistent and rational way; but such a scheme could not be developed here. For from the level of the aisle arches a dome, even on a proportionately high drum, could not be made to reach the level of the cornice of the enclosing wall unreasonably elevated for the sake of the gigantic external order. But Michael Angelo nevertheless constructed such a dome (_A_, Fig. 32), although it had to be sunk up to its crown beneath the aisle roof, and then, for external effect, he built another dome over it (_B_, Fig. 32). To light the lower dome it was necessary to sink oblique openings, _a_, through the massive masonry of the roof, and to light the useless vaulted chamber, _b_, which he was obliged to make over the barrel vault of the inner compartment (the crown of which is still farther down below the roof), the well, _c_, had to be sunk. Thus instead of making a reasonable design with ornamental details appropriate to its structural forms, Michael Angelo first conceived an ornamental scheme consisting of the inappropriate colossal order, and then fitted the building to it, filling up vacant spaces with extravagantly massive solids and useless voids, and resorting to other tortuous devices to piece out a fundamentally irrational system.

Such is St. Peter’s church, which, though it has been much criticised, has been more generally lauded as a model of architectural greatness. Its real character has rarely been analyzed or rationally considered. That it has qualities of majesty and grandeur need not be denied; but these qualities are mainly due to its vast magnitude, and to what it retains of the design of its first, and greatest, architect. The manner in which the scheme of Bramante was modified and distorted by his successors, and chiefly by Michael Angelo, notwithstanding his professions of admiration for Bramante’s intentions, is far from admirable, as I think the foregoing account of its structural and artistic aberrations must show. The building as a whole is characterized by incongruity and extravagance, and when we consider further that the ornamentation of the interior is for the most part a cheap deception, the rich coffering of the vaulting and the pilasters of the great order being wrought in stucco on a foundation of brickwork, we get the measure of the ideals and architectural standards of men who, like Vasari, could write of it that, “not in Christendom, nor in all the world, can a building of greater magnificence and grandeur be seen.”[69] And this short-sighted admiration did not abate as time went on, as we learn from the estimates quoted by Fontana in his well-known book,[70] among which are the following: “Temple more famous than that of Solomon,” “Unique miracle of the world,” “Chief among the most celebrated of Christendom,” “Compendium of the arts,” “Basis of the Catholic faith,” “Unique edifice of the orb of earth,” etc., etc.

Before leaving St. Peter’s a word may be said of a project for the building which was prepared by Antonio San Gallo the younger, Michael Angelo’s immediate predecessor as architect for the fabric. This design, no part of which was ever carried out, is embodied in a wooden model preserved with that of Michael Angelo in the existing edifice. The most meritorious feature of this model is the dome which, from a structural point of view, is better than the one that was built, since it is well abutted both at the springing and at the haunch. This important condition is secured, however, by an architectural treatment that cannot be commended, and consists of two superimposed concentric arcades, the lower one surrounding the drum and abutting the vault at the springing, while the upper one is set in retreat and fortifies the haunch. The architectural effect of these arcades, which are of course adorned with classic orders, is not happy because an arcade with a classic order is not an appropriate form of abutment, though it may be made mechanically effective, and also because the upper circle, rising from within the circumference of the lower one, gives the composition an unpleasantly telescopic effect.

Our consideration of St. Peter’s has led us to an advanced phase of the church architecture of the Roman Renaissance, and we must now go back and examine a few of the earlier structures in Rome and elsewhere that were produced under the distinctly Roman influence.

Fig. 33.

The church of Sant’Agostino is spoken of as a building of the early Roman Renaissance, and is said to have been built by the architect Giacomo da Pietra Santa between 1471 and 1484. But it is incredible that such a church could have been designed by any architect of the Renaissance, or by an Italian architect of any time. Letarouilly says of it that from the thirteenth century the Augustinians had a convent and small church in Rome, and that two centuries later they resolved to enlarge the church, and employed as architects Giacomo da Pietra Santa and a Florentine named Sebastiano.[71] The character of the building is such, however, as to warrant the belief that it is a mediæval structure with slight interior ornamental additions of the Renaissance, which may be by Pietra Santa, and a façade, dating from before the close of the fifteenth century, by Baccio Pintelli. In general character the church is in the style of the Rhenish Romanesque architecture of the twelfth century. It has a nave with groined vaulting in square compartments, each embracing two vault compartments of the aisles. It has also the Rhenish alternate system with plain square piers, and archivolts of square section, originally without mouldings, and the main piers have each a broad pilaster-strip carried up to the springing of the vaults. The triforium space has no openings, and the clerestory has plain round-arched windows. It is thus a thoroughly northern Romanesque scheme, entirely logical in its simple construction and fine in its proportions. The Renaissance interpolations consist of a few ornamental details only. A stilted composite column is set against the pilaster-strip of each main pier (Fig. [33]), this column is crowned with an entablature-block reaching to the level of the triforium, and upon it is set a short pilaster surmounted with a smaller entablature-block at the vaulting impost. This superfluous and irrational compound, breaking the reasonable and effective continuity of the mediæval pilaster-strip, greatly disfigures the originally noble design. The only other neo-classic details of the interior are mouldings at the arch imposts and on the archivolts, and coffering on the soffits of the arches. These are quiet and less injurious in effect, though equally superfluous and inappropriate. Thus did the sophistication of the Renaissance designers often blind them to real architectural excellence, and lead them to fancy that they could improve such an admirable and consistent interior by incongruous and meaningless features.

Fig. 34.—Façade of Sant’ Agostino.

The façade (Fig. [34]) is wholly of the Renaissance, and has no mediæval character except in its general outline, which conforms with that of the building itself. It is a simple design, and foreshadows those of Vignola and Della Porta for the church of the Gesù, to which it is superior in merit, being more reasonable and quiet. Shallow pilasters of considerable elegance mark the divisions of the interior, the portals are framed with simple classic mouldings without orders, and the aisle compartments are surmounted with reversed consoles after the manner of those introduced by Alberti in the façade of Santa Maria Novella in Florence. These consoles are, however, so different in character from the rest of the façade, having their details in higher relief and being set a little in retreat, that they would appear to be later interpolations. Answering to nothing in the building, they are superfluous ornaments, and do not improve the composition, which without them is as reasonable as a composition made up of superficial classic details can well be. A peculiar feature of this front is the truncated pediment that crowns the lower division, and forms the basis of the clerestory compartment. The small rectangular tablets that break the wall surfaces are also noticeable as foreshadowing a treatment that was subsequently much affected by Vignola. Contemporaneously with the façade, and by the same architect, a dome on a drum resting on pendentives was built over the crossing. The present dome rising directly from the pendentives is an alteration of a later time.

Fig. 35.—Exterior of Santa Maria della Consolazione, Todi.

In the earlier churches that were wholly built under the Roman Renaissance influence, the Byzantine scheme largely prevails in the plan and structural forms, probably because it lent itself to the most effective display of a high central dome. Among the first of these buildings is the church of Santa Maria della Consolazione outside the wall at Todi. The design is attributed to Bramante,[72] and it seems to bear enough resemblance to what we know of his work to justify the attribution. The arms of the cross here take the form of apses, the eastern one being semicircular on plan, and the others polygonal. The dome (Fig. [35]) is raised on a high drum, and is almost an exact reproduction of that of San Pietro in Montorio. Its thrusts are thus entirely unbuttressed, but it is probably bound with chains, as was the custom at this time in domes constructed in this manner.[73] The half-domes of the apses are better adjusted. They spring from within the supporting walls, which are carried up high enough to give effective abutment, and are loaded at the haunch by stepped rings of masonry, as in the Pantheon. The details of the interior (Fig. [36]) consist of two superimposed orders of small pilasters, with great pilasters on the angles of the crossing reaching from the pavement to the springing of the pendentive arches, and from ressauts of the upper entablature converging ribs rise against the surfaces of the vaults. Several further awkward results are here noticeable as a consequence of this application of the inappropriate classic details to the Byzantine structural scheme. The entablature which is carried around the whole interior at the springing of the vaults, has to do duty at once for the small order of the upper stage and for the great angle pilasters, and thus in so far as it is in good proportion for the one it cannot be so for the other. Then the true magnitudes of the piers and the pendentive arches are falsified by the pilasters and simulated archivolts which spring from them. These piers and arches really embrace in width both the pilasters and archivolts and the spaces of wall and vaulting between them and the pilasters of the smaller orders and ribs which spring from them. The proper and impressive massiveness of the essentially Byzantine system employed is thus contradicted by an apparent skeleton of classic orders simulating an organic structural scheme which has no real existence.

Fig. 36.—Interior of Todi.

Fig. 37.—Plan of San Biagio.

The exterior of this monument (Fig. [35]) has much merit in its general form and proportions. The great central square mass, visible from the ground upward, gives the sense of support for the dome which the eye demands, and the apses with their half-domes are effectively grouped in subordination to the crowning feature. But this merit, which Todi shares with many other buildings of the Renaissance, is primarily due to the Byzantine scheme adopted, and cannot, therefore, be wholly credited to the Renaissance architect.

Fig. 38.—Interior of San Biagio.

Fig. 39.

A variation of this scheme occurs in the church of San Biagio at Montepulciano by Antonio San Gallo the elder, and begun in the year 1518. Here the arms of the cross (Fig. [37]) are square, with an apse added to the eastern arm. The interior is ornamented with a single, and very heavy, Doric order (Fig. [38]), framing arched recesses in the imperial Roman manner. The use of pilasters on the angles makes the awkward combination of a pilaster coupled with a column necessary, and since the entablature is in the plane of the wall, it has to be broken into very salient ressauts in order to cover these members. Above the entablature is a low ledge in retreat, broken into ressauts in conformity with those of the entablature, and from these ressauts coffered archivolts are sprung under the ends of the barrel vaults which cover the arms of the cross. The Doric order is designed here, for the most part, in close conformity to ancient models, save for the pilaster on the angle, which does not generally occur in Roman monuments. The common Roman treatment of the angle is shown in the arch of Septimius Severus (Fig. [20], p. 41), where the end column of the order is placed at some distance from the end of the façade, which is left in retreat without any pilaster. But Serlio[74] describes the ruins of an ancient Roman building (Fig. [39]) that appears to have been a sort of open arcade or stoa, used as a meeting place for merchants, on the angles of which pilasters are set together with columns, somewhat as they were by Alberti in Santa Maria Novella, by San Gallo here in San Biagio, and by many other architects of the Renaissance. He speaks of the treatment of the angles of this building as follows: “The corner pilasters are larger than the others, and were truly made with excellent judgment, for they strengthen the angle with good effect; and from this architects may learn how to design angles with columns and pilasters bound together, in order that the corner may be brought into line with the column, which gives more solidity to the angle. If the said angle were withdrawn into line with the middle pilasters, the façade, when viewed obliquely, with the round column on the angle, would appear imperfect, and for this ... I strongly commend this form of angle because it may be fully seen from all sides.”

Fig. 40.—San Biagio, Montepulciano.

Externally the composition is remarkably good in its larger features (Fig. [40]). The dome, of slightly pointed outline, on a high drum, rises grandly from the substructure, and is well proportioned in relation to it. The wall surfaces are treated broadly, with no orders carried across them. They are divided into two stages, with a pediment over each façade. Super-imposed pilasters are set on the angles, and a Doric entablature, carried across the whole front, with ressauts over the lower pilasters, divides the two stages. The wall of the lower stage is entirely plain, with a severely simple rectangular portal surmounted by a pediment. The wall of the upper stage is divided into rectangular panels, as in the attic of the Pazzi chapel in Florence, the central panel being pierced with a square-headed window and framed with an order of which the capitals are Ionic and the entablature Doric. The cornice of the top story and the raking cornice of the pediment of each façade are broken into ressauts over the pilasters, and an order of Ionic pilasters, with a very high entablature broken into ressauts, surrounds the drum which supports the dome. Square detached towers are set in the reëntrant angles of the west side, only one of which was carried to completion. The completed one is in three stages, each adorned with a heavy order, Doric, Ionic, and Corinthian respectively. In these orders half-columns are coupled with angle pilasters, as in the interior, and the entablatures have ressauts on the angles over these members. An octagonal spire-like lantern, with a tall drum adorned with an order of Corinthian pilasters and surmounted by an attic, crowns the tower. Small obelisks set on the tower angles and reversed consoles against the angles of the attic give a simulation of Gothic form to the neo-classic scheme, and show the strong hold that mediæval ideas still retained upon the minds of the designers. The first of these spire-like towers of the Renaissance appears to be that of the church of Santo Spirito in Florence, which is spoken of by Milizia as the most beautiful of Italian bell towers.[75] It was designed by Baccio d’Agnolo, who, beginning as a wood carver, imbibed the new enthusiasm for the antique, and after studying the ancient monuments of Rome[76] began the practice of architecture. This campanile is thus noteworthy as the first of a large class of modern towers with spires of which Wren’s famous steeples were the ultimate outcome. The scheme is based on the mediæval campanile, the earliest form of which is the Lombard Romanesque tower. The Lombard tower is characterized by its simple rectangular outline, the walls rising sheer from the ground to the cornice, and strengthened and adorned with shallow pilaster-strips, corbelled string-courses marking the successive stories, and by small grouped openings. The tower of Santa Maria Novella in Florence is designed on this model, and the neighbouring tower of Prato and Giotto’s famous campanile are later and richer modifications of the same type. In the tower of Santo Spirito (Fig. [41]) Baccio d’Agnolo has taken the Lombard scheme and clothed it with a pseudo-classic dress. While his classic details have much of that elegance which belongs to the best Italian work, they are out of place in such a structure. The tall pilaster-strips of the mediæval tower gave an expression as of an organic skeleton running through the building. They had been developed out of the classic pilaster to meet the needs of the mediæval type of structure, and in substituting the superimposed classic orders for the appropriate continuous members, the artist did violence to the true principles of design.

Fig. 41.—Tower of Santo Spirito.

The lantern with which this tower is crowned is an adaptation of Brunelleschi’s lantern on the dome of the cathedral, but made more aspiring in form, so that the general outline is like that of a Gothic spire. But the form of a Gothic spire is far removed from anything that is proper to classic composition.

Returning to San Biagio, it may be said that the orders here have a closer conformity with those of classic antiquity than occurs in the earlier monuments already mentioned, except the Tempietto of San Pietro in Montorio by Bramante.

Fig. 42.—Santissima Annunziatta, Arezzo.

In the nave of the church of Santissima Annunziatta in Arezzo, the same architect produced a different design. The nave (Fig. [42]), of only three bays, is covered with a barrel vault, and the aisles have small domes on pendentives. The supporting piers are square with a shallow Corinthian pilaster on the face of each and an entablature passing over the crowns of the arches. The archivolts are deep, and each one is moulded on the face and plain on the soffit. These are carried on plain pilasters with simple impost mouldings. The wall above the entablature is plain and unbroken, except by a round-arched window over each bay of the ground story, and is crowned with a heavy cornice from which the vaulting springs. We have here a structural system of imperial Roman massiveness, necessitated by the use of the great barrel vault.

After the early part of the sixteenth century Italy produced few architects of a high order of genius. Most of the more advanced neo-classic art is the work of mediocre men who, while professing to be ardent advocates of grammatical correctness according to the ancient rules, were hardly less capricious in their misuse of classic elements than their predecessors had been. To enter upon the examination of any large number of buildings in this later Renaissance style would be tedious and unnecessary; but in addition to what we have already seen of it in the work of Michael Angelo in St. Peter’s, we may give some attention to a few characteristic works of the two leading architects of the later time: Vignola and Palladio.

Few men did more to make the neo-classic ideas authoritative than Giacomo Barrozzi, called Vignola. Beginning like so many others with painting, Vignola was led early to the study of architecture, in which he strove to gain an exact knowledge of classic Roman forms by drawing and measuring the remains of the ancient edifices. He thus became a devoted partisan of the antique, and he wrote a treatise on the Five Orders which has been widely accepted as an authoritative guide in modern architectural practice. To him, says Milizia, “Architecture is under lasting obligations because he established it upon system, and prescribed its rules.”[77] And the same author tells us further that Vignola “purified architecture from some abuses which neither his contemporaries nor the ancients had perceived”; yet nevertheless, he adds, “his book has produced more harm than good, for to make the rules more general, and more easy of application, he has altered the finest proportions of the antique.” No system of architecture, Milizia says further, “is more easy than that of Vignola, but the facility of it is obtained at the expense of architecture itself.”

In his book,[78] which is made up largely of drawings and diagrams, Vignola shows how the proportions of an order may be regulated by a module down to the smallest details. He explains how to construct Ionic volutes and other curves from centres, and how to describe the details of Corinthian and composite capitals by means of plan and elevation. He thus introduces a mechanical system modelled after the formulas of Vitruvius.

Fig. 43.—Vignola’s entablature.

But notwithstanding his ardent advocacy of the principles of ancient Roman art, Vignola, in his own practice, not only altered the proportions of the orders as Milizia says, but made many fanciful changes in them. He introduced details which have no counterparts in correct Roman design, and freely mixed those of different orders. An instance of this occurs in an entablature figured in his book,[79] which he calls his own invention. In this composition (Fig. [43]) we have a pseudo-Doric frieze between an architrave with multiplied faciæ, and a cornice on modillions. In the place of triglyphs this frieze has consoles with two channels, like those of a triglyph, on the curved face of each. To such travesties of classic design did the striving after novelty, which was curiously mingled with their ardour for the antique, lead the men of the later Renaissance. For an advocate of classic correctness such aberrations are the more surprising as they are expressly condemned by Vitruvius, who warns his readers against them as follows: “If dentiled cornices are used in the Doric order, triglyphs applied above the voluted Ionic, thus transferring parts to one order which properly belong to another, the eye will be offended, because custom otherwise applies these peculiarities.”[80] The Roman writer might, indeed, have given a better reason why the purity of the orders ought to be maintained, namely, because to each of them the fine artistic genius of the Greeks had given its appropriate details.

Fig. 44.—Half plan of Sant’Andrea.

In designing entire buildings Vignola shows no less freedom in unclassic and incongruous combinations. This is manifested in the earliest of his church edifices, that of Sant’Andrea di Ponte Molle outside of the Porta del Popolo at Rome (Figs. [44], [45], and [46]). It is a small, oblong, rectangular enclosure covered with a dome of oval plan on pendentives. The structural scheme is thus primarily Byzantine, but the architectural treatment is Roman. The dome is built in a praiseworthy form, and follows the construction of the dome of the Pantheon. An enclosing drum is carried up from the pendentives to a considerable height, and the haunch of the vault is well fortified by stepped rings of masonry. These rings are criticised by Milizia[81] as awkward and unnecessary because, he affirms, the vault might have been made secure without them. He probably means that it might have been bound with chains in the usual manner of the Renaissance. As in the Pantheon, the drum rises so high above the springing that but little of the dome is visible externally. The character of the rectangular substructure is puzzling to the eye of a beholder who looks for meaning and congruity in architectural forms. Wrought in shallow relief upon its façade is an order of Corinthian pilasters surmounted by a classic pediment, and the entablature of the order is returned on the sides of the building. The effect of the whole may be compared to that of a Greek temple with an attic supporting a dome built upon it. So awkward is the combination that it might be supposed to be a piece of patchwork in which a building of Greek temple form had been altered to gain more height within, were it not that we find in the architect’s own book the plan and section reproduced in Figs. 44 and 45, which show that the building as it now exists was originally designed in its present form.[82]

Fig. 45.—Longitudinal section of Sant’Andrea, from Vignola’s book.

On reflection we discover that the scheme suggests a derivation from the Pantheon. Not only is the dome shaped and adjusted as in that ancient monument, save for its oval plan, but the rest of the composition is pretty clearly from the same source. To realize this it is necessary only to eliminate, in idea, the portico of the Pantheon with the exception of its pediment, and to conceive this pediment as drawn back into the plane of the rectangular façade. The pediment would then surmount the order of Corinthian pilasters which adorn this façade, and the resulting composition would be substantially identical with that of the façade of St. Andrea. The minor differences are unimportant, as where Vignola has placed a pair of pilasters, instead of only one, at each end of the façade, has given the whole order more shallow relief, and has omitted the fluting on the pilasters. Even the niches on either side of the portal are reproduced from the Pantheon, though Vignola has pierced them with windows.

Fig. 46.—Sant’Andrea di Ponte Molle.

The likeness extends farther. The return of the entablature along the side walls and the cornice of the attic are the same in both instances; but the second pediment in the Pantheon façade Vignola has not reproduced. St. Andrea is thus a close, though a modified, copy of the rectangular part of the Pantheon, with the rectangle elongated and surmounted by a dome designed on the Pantheon model. It was not known in the sixteenth century that the ancient monument is not a homogeneous structure, but an awkward patchwork, the result of successive alterations and additions.[83] Vignola took it entire as an example of that ancient style which he regarded as authoritative, and based his design for St. Andrea upon it, just as many modern architects have taken motives from Vignola himself. If it were proposed to erect a dome upon the Parthenon, few people would fail to see that the result would be an architectural monstrosity, yet this would not be very different from what was done in St. Andrea by an architect who has been looked upon as a champion of classic correctness in design.

M. Palustre has called attention to the fact that, in the interior of St. Andrea (Fig. [45]), the two parts of the entablature which have no _raison d’être_ under a vault have been omitted.[84] But the impropriety of a complete entablature in connection with vaulting is no greater than that of any part of a classic order, which has no justification in such connection, as we have already remarked.

The pilgrimage church of Santa Maria degli Angeli, built over the oratory of St. Francis at Assisi, is a more extensive monument which was begun by Vignola in the year 1569. Though completed by other architects, and extensively restored in 1832, the building as it now stands is uniform in style throughout, and bears the marks of Vignola’s manner of design. It is cruciform in plan, with a long nave and aisles, and a square chapel opening out of each bay of each aisle. The nave and transept have barrel vaulting, a half-dome covers the apse, and a dome on a high drum resting on pendentives rises over the crossing. The aisles have domical groined vaulting with transverse ribs, and the side chapels have barrel vaults with their axes perpendicular to that of the nave. These chapels thus form abutments to the inner vaulting, so that no external buttresses are needed. The entire fabric is of brick, but the details, including the orders of the interior, of the west front, and of the drum, are wrought in stucco. For the interior the architect has employed a great order of Doric pilasters, a single pilaster on the face of each pier, and on the sides of the piers, under the aisle archivolts, he has placed pairs of smaller pilasters. The soffits of the archivolts are very wide, and have each a pair of salient sub-archivolts corresponding with the pilasters. It had been common for the architects of the Roman Renaissance to break the entablature into ressauts over the columns or pilasters of the orders when used in this way, as San Gallo had done in Montepulciano and Michael Angelo in St. Peter’s. But the effect of thus breaking the continuity of the cornice line is unpleasing, and Vignola has avoided it here by confining the ressaut to the architrave, frieze, and bed-mouldings, leaving the corona of the cornice unbroken as in Figure 47. The great piers of the crossing show the influence of St. Peter’s in being splayed, and the forms of the pendentives lose their spherical surfaces in being fitted to the straight line of the splay, as they do in St. Peter’s. The design of the façade expresses with unusual truthfulness the divisions of the interior, which are marked by pilasters like those of the great order within, and by an arch coinciding with the curve of the vaulting.

Fig. 47.—Order of Santa Maria degli Angeli.

Fig. 48.—Plan of the Gesù.

The Gesù in Rome, another large church by Vignola, and built at about the same time as Santa Maria degli Angeli, is a variation of the same scheme, and shows in a more marked degree the influence of St. Peter’s. A plan of this building, the intended façade which Vignola did not live to construct, and the existing façade by Jacomo della Porta are given in the addendum to the edition of the architect’s book on the Five Orders published in 1617 already referred to (p. [84]), and are reproduced in Figures 48, 49, and 50. The aisles are omitted here so that the side chapels, which communicate with each other by narrow openings in the dividing walls, open directly out of the nave. The transept is short, and extends on either side beyond the nave only by the thickness of its walls. An elevated dome on pendentives, circular on plan within and octagonal on the outside, rises over the crossing, and barrel vaults cover the nave and transept arms. The side chapels are vaulted, with small domes on pendentives, except those in the angles of the crossing, which do not require pendentives because their supports are shaped to the circular form as shown in the plan, Figure 48. These supports are made heavier than the others in order to strengthen the crossing piers, which, in consequence of this reënforcement, do not need to advance so far into the space under the great dome as they otherwise would. In Santa Maria degli Angeli the aisles prevent this treatment, and the crossing piers extend far into the nave and narrow the spans of the crossing arches.

The scheme of the interior of the Gesù is a close reproduction of that of St. Peter’s, though the great pilasters are of the composite, instead of the Corinthian, order, and other minor differences are noticeable. It is worthy of remark that the entablature has no ressauts except at the crossing, and the vaulting is raised upon an attic, so that no part of it is hidden from view by the cornice of the entablature, as it is in St. Peter’s. It is also noticeable that, while capricious in the use of elements derived from the antique, Vignola in his church architecture eliminates mediæval forms more completely than most architects of his time. Where in St. Peter’s, for instance, the apses have celled vaults on converging ribs, he employs the plain half-dome of Roman antiquity.

_Questa facciata no fu messa in opera per la morte del architecto_

Fig. 49.—Façade of the Gesù, Vignola.

_Facciata del Giesu come al presente si troua fatta da Iacomo della Portta._

Fig. 50.—Façade of the Gesù, Della Porta.

Vignola’s design for the façade (Fig. [49]) presents the familiar features of his style as already embodied in the earlier façade of St. Andrea, but with additional infractions of propriety, as well as of classic form in its more elaborate details. This façade corresponds in outline with the form of the building, except for the podium of the upper story (which contradicts the roof lines of the side chapels), and the abutting walls of curved outline over the side compartments. The chief aberrations of detail are the broken pediments of the doors and windows, and the barbaric scrollwork and hermæ, the use of which this architect did much to establish. How far the barbarism of breaking the pediment was an independent freak of the Renaissance I do not know. Instances of somewhat similar treatment occur in the Roman architecture of Syria, as in Baalbek (Fig. [51]), where the middle part of the pediment is in retreat of the rest, so that the ends form ressauts. Of the complete removal of a part of the cornice I know no instance in the Roman architecture of antiquity. To this, however, the architects of the later Renaissance were, in their desire for novelty of design, led. But the cornice of a pediment is, like the roof of an entire building, suggestive of shelter for the parts below. The actual necessity for such shelter may be slight, but any justification which the raking cornice has must be for expression, if nothing more, of a sheltering roof to what it surmounts (unless we are to assume that architectural design is a matter of purely fanciful composition of lines with no structural meaning or expression). To cut a piece out of the middle of it is an architectural solecism.

Fig. 51.—Pediment of Baalbek.

The actual façade by Della Porta (Fig. [50]) follows the main lines of Vignola’s design, but the details are much altered. The podium of the upper story is raised in height, reversed consoles are substituted for the plain curved abutments of Vignola, and the raking cornices of the small pediments are made whole. But other aberrations take the place of those which are eliminated, as that of placing one pediment within another over the central portal, and the ugly shapes and framings of the tablets and niches that break the wall surfaces. Della Porta had acquired these habits of design from his master, Vignola, and how far Vignola himself could go in such monstrosities is shown in some of the figures of his book already spoken of. Figure 52 from this book affords an instance.

Fig. 52.—Tablet from Vignola.

If Vignola did much to make authoritative the later ideas of the sixteenth century as to the principles of ancient art and their application to modern uses, Palladio did even more. By the example of his numerous architectural works, as well as by his writings, the influence on modern art of this famous neo-classicist has been greater than that of any other architect of the Renaissance, so that we have, in the principal countries of Europe, a style of architecture which is known as Palladian. Palladio was the first architect of the Renaissance who was not at any time either a painter or a sculptor. He begins his well-known book[85] as follows: “Guided by natural inclination, I began in my earliest years to devote myself to the study of architecture, and having been always of the opinion that the ancient Romans were in building, as in many other things, far in advance of all those who came after them, I took for my master and guide Vitruvius, who is the only ancient writer on this art, and I set myself to the investigation of the remains of the ancient edifices which, injured by time and the violence of barbarians, are still extant. And finding them much more worthy of attention than I at first thought, I began with great diligence to measure most minutely every part of them. I became so ardent an investigator, not having known with what judgment and fine proportion they had been wrought, that not once only, but many times, I visited different parts of Italy and elsewhere, in order to understand and delineate them completely. And seeing how far the common manner of building differs from what I have observed in the ancient edifices, and read in Vitruvius, and in Leon Batista Alberti, and in other excellent writers since Vitruvius, and from that new manner which I have practised with much satisfaction, and which has been praised by those who profited by my work, it has seemed to me right, since man is not born for himself alone, but also to be useful to others, to publish the drawings of these edifices, which at the cost of much time and peril I have gathered; and to state briefly that which has seemed to me most worthy of consideration in them, together with those rules which I have observed, to the end that those who shall read my book may profit by such good as may be in it, and supply that which may be wanting (for much, perhaps, may be) so that, little by little, we may correct the strange abuses, the barbarous inventions, avoid the superfluous cost, and (what is more important) the various and continued deterioration which we see in so many buildings.”

The implicit confidence of the neo-classicists in the art of Roman antiquity as the embodiment of all true principles of architectural design, and their unquestioning belief that mediæval art was wholly false in principle and barbaric in character, have seldom been more naïvely expressed.

Fig. 53.—Orders of San Giorgio.

Of church architecture by Palladio we have two important buildings, San Giorgio Maggiore and the Redentore, both in Venice. The first of these stands on the island of San Giorgio, opposite the Piazzetta, and is a characteristic Palladian design, though some parts of the west front may have been added after the architect’s death. This church is cruciform, and has barrel vaulting with interpenetrations for light, and a dome on pendentives over the crossing. The piers are heavy, with a single engaged column of the composite order, raised on a high pedestal, against each one, except at the crossing, where the columns are coupled with pilasters, while the wide archivolts rest on pairs of smaller pilasters of the Corinthian order, without pedestals (Fig. [53]). Both columns and pilasters have strong entasis, and the frieze of the entablature is rounded in profile. In raising the great order on pedestals Palladio conformed more closely to ancient Roman practice than Michael Angelo and Vignola had done; but the pedestals have a clumsy effect thus ranged along the nave, and their sharp angles are in the way of moving crowds of people. It is noticeable, too, that Palladio has introduced complete orders under the archivolts, giving an entablature to each pair of small pilasters. The entablature had before been omitted in this situation. The whole scheme shows in a marked degree how inappropriate is the use of classic orders in a church interior. The application of such orders to a building with aisles and a high nave obliges the designer to make awkward combinations, and to violate true classic usage in manifold ways, as we have already abundantly seen. He must associate large and small orders, and give them relationships and adjustments that belong to mediæval, rather than to classic, composition. The façade of this building (Fig. [54]) has the merit of conforming in outline to the shape of the nave and aisles. It is the outline of the primitive Christian Roman basilica without any disguises in the way of reversed consoles over the aisle compartments, or divisions contradicting those of the interior. Instead of the superimposed orders of Vignola’s west fronts, Palladio has here, in the nave compartment, one great order of engaged columns, on high pedestals, rising through the entablature of a small order of pilasters, which is carried across the whole front, reaching to the height of the aisles. The total scheme gives a suggestion of mediæval organic composition, but has no real organic character pertaining to the building.

Fig. 54.—Façade of San Giorgio.

In the façade of San Francesco della Vigna, also in Venice, and by the same architect, the design of San Giorgio is repeated, with some notable changes in detail. In this case the small order, as well as the larger one, consists of columns, except that on each angle a pilaster takes the place of a column, and both orders rise from the same level, the smaller one resting on a continuous podium, and the larger one on pedestals which are ressauts of the podium. The entablature of the small order is here not continuous, but is broken by the nave compartment, though a fragment of it is inserted in the central bay of this compartment over the small columns that flank the portal.

Fig. 55.—The Redentore, Venice.

The scheme of the Redentore differs from that of San Giorgio. It has no transept and no aisles, but in the place of aisles a series of side chapels. A square area in front of the sanctuary is covered with a dome on pendentives, while the nave has a barrel vault, and the side chapels have barrel vaults with their axes perpendicular to the axis of the nave. From the dividing walls of these chapels solid abutments in pairs are carried up through the lean-to roofs over the chapels to meet the thrusts of the nave vaulting, as shown in the general view of the exterior (Fig. [55]). The plan of the east end is peculiar. A round apse opens out of the north and south sides of the square covered by the dome, and a colonnade on a curved plan forming the sanctuary bounds this square area on the east side. Beyond this is an oblong enclosure the eastern wall of which is on a curved plan, and the sanctuary is flanked by small towers. The interior has a great order of Corinthian columns, one against each pier, resting directly on the pavement, and the small pilasters under the archivolts carry entablatures which extend to the outer wall and from them the barrel vaults of the chapels spring. The entablature of the great order is not set in the wall and broken by ressauts to cover the columns, as in San Giorgio; but is carried by the columns, and thus overhangs the wall with a supporting corbel in the middle of each intercolumniation which forms a keystone to the arch beneath. The façade of the Redentore is a variation of that of San Giorgio with the pedestals omitted from the great order, as in the interior, and it has an attic behind the pediment like that of Vignola’s small church of St. Andrea at Rome. Such is the nature of Palladian church architecture. We shall see more of Palladio’s art when we come to the consideration of the later civil and domestic architecture of the Renaissance.

It is unnecessary to multiply examples of the church architecture of the Roman Renaissance, _i.e._ that architecture which derived its character primarily from the influences that were active in Rome from the beginning of the sixteenth century. For while the churches of this style differ considerably one from another in details, they agree essentially in architectural treatment growing out of a closer contact with ancient monuments, though with no strict conformity to them. Descriptions of minor differences in the forms of such buildings, and in the composition of their ornamental details, are tedious, and enough of them have now been given. We may, therefore, in the next chapter, pass on to the consideration of the palace architecture of the Renaissance.

CHAPTER VI
PALACE ARCHITECTURE OF THE FLORENTINE RENAISSANCE

While it was in church edifices that the neo-classic ideas in architecture were first embodied, it was in vast palatial houses that they were most extensively carried out. Early in the fifteenth century luxurious living began to prevail among the upper classes of society, and sumptuous private dwellings on an unprecedented scale were now erected in Florence. Magnificent palaces had, indeed, been built in the later Middle Ages which were among the chief ornaments of the mediæval towns; but those were civic monuments expressive of the communal spirit and artistic culture of their time. Such buildings as the Palazzo Vecchio in Florence, the Palazzo Pubblico of Siena, the Ducal Palace of Venice, and many others were the material manifestation of a state of municipal pride, and popular love of beauty and propriety in public monuments. Upon these buildings the best craftsmanship was lavished, while the dwellings of the most wealthy citizens were modest in scale, though often beautiful in design.

Plate IV

PALAZZO RICCARDI

Florence

A fine example of an unostentatious, though dignified, house of a Florentine Patrician of the thirteenth century still extant is the Palazzo Mozzi. Its broad walled front of two stories over a high basement, with narrow string courses of simple profile and moderate projection, its well-faced and finely jointed masonry, and its plain window openings of the characteristic mediæval Florentine form in which the extrados is pointed while the intrados is round (Fig. [56]), is a model of architectural simplicity, while it expresses the superior social station of its inmates. A few smaller houses of similar character as to quietness and simplicity of design, many of them suited to the needs of humbler citizens, have been preserved in some of the Italian towns. A few interesting examples of these may be found in Perugia. They have plain stone fronts, with simple string courses marking off the stories, and windows in some cases wholly round arched, in others having the extrados pointed with an ogee curve (Fig. [57]).

Fig. 56.

Fig. 57.

But early in the fifteenth century vast structures for private use began to arise which rivalled in scale, and in costly splendour, the great civic monuments of the former time. The first of these larger palaces in Florence is the one now known as the Riccardi, designed by the architect Michelozzi for Cosimo de’ Medici in 1430. It is a princely edifice, and though comparatively plain in general aspect, it is in many ways superior in architectural character to all of those which followed it. Like other buildings of its class it is in plan a survival of the ancient Roman house, having the form of a rectangle enclosing an open court. In elevation (Plate [IV]) it has two stories over a high basement, and is grandly simple in design, and fine in its proportions. In buildings of this class there is no peculiar internal system which requires attention before the outside can be understood. The apartments have generally flat wooden ceilings, and where vaulting occurs, as usually in the basement and sometimes in the upper stories, it is of a kind that calls for no buttresses against the wall, the thrusts being met by the thickness of the walls, and by the weight of the upper stories. The façades of the Riccardi have no engaged orders, but the great cornice has classic profiling, and its bed mouldings have dentils and other classic details, while modillions of semi-classic form support the corona. The window openings are of thoroughly mediæval character in their larger features, and are each composed of a round arch embracing two smaller arches with a central shaft and jamb shafts, but the shafts have the tapering form with entasis, and the _congé_, of classic design. The capitals are of nondescript form, with a channelled bell, an ovolo with egg and dart ornament, reversed Corinthianesque leafage depending from its angles, and a Corinthian abacus in each. The openings are uniform in each story, and their archivolts are treated in the mediæval Italian manner, the extrados being struck from a higher centre than the intrados. The graduated heights of the stories, and the varied treatment of the wall surfaces by rough-faced rustication in the basement, smooth-faced rustication in the principal story, and smooth close-jointed masonry in the top story, add much to the beauty of this finest of early Renaissance palaces. It is worthy of notice that here, as in the Italian domestic architecture of the Renaissance generally, the roof is not visible in a near view of the building, and no dormers or chimney-stacks appear. The conditions of climate did not call for a high-pitched roof, nor for any of those features that are naturally developed in the architecture of more northern countries. The general outline of the edifice is thus severely simple, and its agreeable effect is due to its fine proportions and arrangement of parts. It is noticeable, too, that the reveals are shallow on the outside, in marked contrast to the deep reveals of the later Renaissance architecture. This is not only conducive to quietness of effect, but it has the advantage of giving to the interior the maximum of light—since the farther out the glass is placed the less will be the shadow thrown upon it, while the internal reveal, especially when it is splayed, reflects light into the interior.

The interior court of the Renaissance palace has a vaulted arcade on each of its four sides beneath the overhanging upper stories. These arcades are, in the Riccardi (Fig. [58]), supported on columns of classic form with capitals of a composite type, but of no great beauty. The arches spring directly from these capitals, and have classic profiles, while two string courses, with an interval forming a semblance of a frieze, give the effect of an entablature passing over the crowns of the arches.

The spacious apartments of these early Florentine palaces are generally fine in their proportions and simple in their architectural treatment. They are, however, rarely well lighted. The ceilings are at a great height above the comparatively low windows, and the windows are disposed for external effect, rather than for convenience within. Thus while these apartments are stately, they are rarely adapted to cheerful indoor life, and in a northern climate they would be intolerably gloomy. When used, as they now often are, as galleries for the display of works of art, they do not serve well, very small portions of their vast wall spaces being well lighted, and the disposition of the openings often such as to produce embarrassing cross lights and reflections.

Fig. 58.—Court of the Riccardi.

Vasari tells us that “after Brunelleschi, Michelozzi was held to be the most consistent architect of his time, and the one who with best judgment planned either palaces, monasteries or houses.” And concerning the Riccardi he adds, “All the more praise is due him since this was the first Palace in Florence built in the modern manner, and which has a disposition of apartments both useful and beautiful.”[86] He does not explain in what the superior planning of the Riccardi consists, and it is doubtful whether these remarks were based on any definite idea. But however this may be, the building is indeed a stately and magnificent one, of quiet aspect, and for the most part free from meaningless features.

Hardly any other one of the Florentine palaces of the Renaissance equals the Riccardi in beauty and dignity. That part of the Pitti which was begun by Brunelleschi in 1435, though equally free from meaningless features, is almost too bald to be called an architectural design. Each story of its long façade is as monotonous as the Claudian aqueduct which it closely resembles.

The front of the small palace called the Strozzino is in the style of the Riccardi, and is attributed to the same architect. It has but one story above the high basement, and the treatment is even more mediæval in character, the window arches having the pointed form.

The Palazzo Strozzi, begun in 1489 by Benedetto da Majano, follows the same general scheme as the Riccardi, but is less admirable in its proportions. Vasari tells us that Majano carried the exterior almost to completion, but that the court and the great cornice were the work of Simione Pollaiuolo, called Il Cronaca. This cornice, he says, was copied from an ancient model in Rome which the architect had drawn and measured with great exactness, but he had here enlarged the scale to suit the proportions of the building.[87] I think it may be said that he enlarged it too much, and that, in common with the cornices of most of these Renaissance palaces, it is too heavy. The Strozzi, more than any other of the palatial houses of its time, has the fortress-like character which indicates the turbulent condition of Florence in the fifteenth century. The vast basement of ponderous masonry, with no window openings near the ground, gives a gloomy and forbidding aspect to the front, and marks a survival of the savage habits of feudal life in this epoch of advanced Italian civilization and culture.

The Palazzo Pazzi, now known as the Quaratesi, is attributed to Brunelleschi, and has the marks of his style in the details of the windows. It has the same general scheme of design as the foregoing houses, and its stories are proportioned with the same pleasing gradation in their heights that we have noticed in the Riccardi; but the wall surfaces are different, being uniformly overlaid with stucco. A series of small circular openings, with mouldings over the windows of the topmost full story, resembling those of the drum of the Pazzi chapel, seem to give further evidence of Brunelleschi’s hand. Still another building in this style, though of even plainer external character, having small undivided openings, is the Palazzo Gondi, designed by Guiliano da San Gallo toward the close of the fifteenth century. The arcades of the court of the Gondi have Corinthian columns of great elegance, and the arches have ornamental keystones.

Another type of Florentine palace of the early Renaissance is exemplified in the Palazzo Guardagni, attributed to Simone Pallaiuolo. It has an open loggia at the top, and the portals and windows have the round arched form with the extrados pointed. This is a thoroughly reasonable and appropriate Italian style of domestic building, and if it had been consistently adhered to, without any admixture of the classic elements that were soon introduced, the domestic architecture of Florence in the fifteenth and sixteenth centuries might have merited our unqualified admiration. On the simple and appropriate scheme of the Palazzo Guardagni there was opportunity for such variations of disposition, proportion, and details as utility and taste might call for, without any resort to neo-classic elements.

The foregoing buildings, though larger and more elegant than the private houses of the Middle Ages, are still in their main features largely mediæval in character. But before the later buildings of this class were erected, another phase of design in palatial architecture arose in which the spirit of the Renaissance is more manifest in the application of the classic orders to the walls of the façades. This application, as is well known, occurs first in the Palazzo Rucellai, designed by the architect Alberti and built just after the middle of the fifteenth century. We have already (pp. [35–42]) seen something of Alberti’s use of classic orders in church architecture, and we have now to consider further the influences which were guiding the public taste as they are reflected in the works of the man who on the whole did most at this early epoch to establish the new architectural ideas. Alberti was a scholar and a man of high social station. Like most men of culture in Florence he had a taste for the fine arts, but, as Vasari tells us, he “applied himself not only to discover the principles and the proportions of antiquity, but also, being naturally so inclined, much more to writing than to practice.”[88] The moving purpose with him was thus primarily archæological and literary, rather than artistic.

Fig. 59.—Façade of the Rucellai.