CHAPTER III

.

CIRCULAR CHURCHES.

CONTENTS.

Circular Churches—Tomb of Sta. Costanza—Churches at Perugia, Nocera, Ravenna, Milan—Secular Buildings.

In addition to the Pagan basilicas and temples, from which the arrangements of so many of the Christian edifices were obtained, the tombs of the Romans formed a third type, from which the forms of a very important class of churches were derived.

The form which these buildings retained, so long as they remained mere sepulchres appropriated to Pagan uses, has been already described (pp. 342 to 346). That of Cæcilia Metella and those of Augustus and Hadrian were what would now be called “chambered tumuli;” originally the sepulchral chamber was infinitesimally small as compared with the mass, but we find these being gradually enlarged till we approach the age of Constantine, when, as in the tombs of the Tossia Family, that called the Tomb of Helena (Woodcut No. 227) and many others of the same age, they became miniature Pantheons. The central apartment was all in all; the exterior was not thought of. Still they were appropriated to sepulchral rites, and these only, so long as they belonged to Pagan Rome. The case was different when they were erected by the Christians. No association could be more appropriate than that of these sepulchral edifices, to a religion nursed in persecution, and the apostles of which had sealed their faith with their blood as martyrs; and when the Sacrament for the dying and the burial service were employed, it was in these circular churches that it was performed. But besides the viaticum for the departing Christian, the Church provided the admission sacrament of baptism for those who were entering into communion, and this was, in early days at least, always performed in a building separate from the basilica. It would depend on whether marriage was then considered as a sacrament or a civil contract, whether it was celebrated in the basilica or the church; but it seems certain that the one was used almost exclusively as the business place of the community, the other as the sacramental temple of the sect. This appears always to have been the case, at least when the two forms existed together, as they almost always did in the great ecclesiastical establishments of Italy. When the church was copied from a temple, as in the African examples above described, it is probable it may have served both purposes. But too little is known of the architecture of this early age, and its liturgies, to speak positively on the subject.

The uses and derivation of these three forms of churches are so distinct that it would be extremely convenient if we could appropriate names to distinguish them. The first retains most appropriately the name of basilica, and with sufficient limitation to make it generally applicable. The word _ecclesia_, or _église_, would equally suffice for the second but that it is not English, and has been so indiscriminately applied that it could not now be used in a restricted sense. The word kirk, or as we soften it into church, would be appropriate to the third,[286] but again it has been so employed as to be inapplicable. We therefore content ourselves with employing the words Basilica, Church, and Round Church, to designate the three, employing some expletive when any confusion is likely to arise between the first two of the series.

The most interesting feature of the early Romanesque circular buildings is that they show the same transitional progress from an external to an internal columnar style of architecture which marked the change from the Pagan to the Christian form of sacred edifice. It is perhaps not too much to assert that no ancient classic building of circular form has any pillars used constructively in its interior.[287] Even the Pantheon, though 143 ft. 6 in. in diameter, derives no assistance from the pillars that surround it internally—they are mere decorative features. The same is true of the last Pagan example we are acquainted with—the temple or tomb which Diocletian erected in his palace at Spalato (Woodcut No. 194). The pillars do fill up the angles there, but the building would be stable without them. The Byzantine architects also generally declined to avail themselves of pillars to support their domes, but the Romanesque architects used them almost as universally as in their basilicas.

Another very striking peculiarity is the entire abandonment of all external decoration. Roman circular temples had peristyles, like those at Tivoli (Woodcut No. 193) and that of Vesta in Rome. Even the Pantheon is as remarkable for its portico as its dome, so is that known as the Torre dei Schiavi,[288] but it is only in the very earliest of the Christian edifices that we find a trace of the portico, and even in them hardly any attempt at external decoration. The temples of the Christians were no longer shrines to contain statues and to which worship might be addressed by people outside, but had become halls to contain the worshippers themselves while engaged in acts of devotion.

The tomb of the Empress Helena (Woodcut No. 227) is one of the earliest examples of its class. It has no pillars internally, it is true, but it likewise has none on the exterior—the transition was not then complete. The same is the case with the two tombs on the Spina of the Circus of Nero (Woodcut No. 396). They too were astylar, and their external appearance was utterly neglected.

[Illustration: 422. Baptistery of Constantine. (From Isabelle.) Scale 100 ft. to 1 in.]

When from these we turn to the Tomb or Baptistery of Constantine, built some time afterwards (Woodcut No. 422), we find the roof supported by a screen of eight columns, two storeys in height, and through all its alterations can detect the effort to make the interior ornamental. It has, however, a portico, but this again is practically an interior, both ends being closed with apsidal terminations, so that it really forms a second apartment, rather than a portico. In both these respects it is in advance of the building next to it in age that we know of—the Octagon at Spalato—which it otherwise very much resembles. The eight internal pillars instead of being mere ornaments have become essential parts of the construction, and the external peristyle has disappeared, leaving only the fragment of a porch.

[Illustration: 423. Plan of the Tomb of Sta. Costanza, Rome. (From Isabelle, ‘Édifices Circulaires.’) Scale 100 ft. to 1 in.]

The tomb which the same Emperor erected to contain the remains of his daughter Constantia, is another example of the same transitional style. The interior in this instance is vaulted, but so timidly that twenty-four pillars are employed to sustain a weight for which half that number would have been amply sufficient. In the square niche opposite the entrance stood the sarcophagus of the princess, now in the Vatican. The roof of the aisle is adorned with paintings of the vintage and scenes of rural life, which, like all those on the tombs of Pagan Rome, have no reference to the sepulchral uses to which the building was dedicated. The whole internal diameter of the tomb is 73 ft., that of the dome 35.

In front of the building is a small crypto-porticus similar in arrangement to that of her father’s tomb, and beyond this is an oblong space with circular ends, and surrounded on all sides by arcades; its dimensions were 535 ft. by 130, and, though so ruined as hardly to allow of its arrangements being restored, it is interesting as being perhaps the only instance of the “_forum_,” which it is probable was left before all tombs in those times, and traces of which may perhaps be found elsewhere, though as yet they have not been looked for.

[Illustration: 424. Plan of San Stefano Rotondo. (From Gutensohn and Knapp.) Scale 100 ft. to 1 in.]

The only other important circular building within the walls of Rome of this early age is that known as S. Stefano Rotondo. Though there is nothing to fix its date with any precision, it is almost certain that it belongs to the fifth century of the Christian era.[289] It is 210 ft. in diameter, and its roof was supported by two ranges of columns, circularly disposed in its interior; and on the first or inner range rested a horizontal architrave like that of St. Peter’s. In the outer one the pillars support arches like those of St. Paul’s.[290] All the pillars are taken from older buildings. The outer aisle was divided into eight compartments; but in what manner, and for what purpose, it is not now easy to ascertain, owing to the ruined state of the building, and to its having been so much and so frequently altered since it was first erected. Nor can it be determined exactly how it was roofed; though it is probable that its arrangements were identical with those of the great five-aisled basilicas, which it closely resembles, except in its circular shape.

[Illustration: 425. Plan of Sti. Angeli, Perugia. (From Isabelle.) Scale 100 ft. to 1 in.]

This is more clear in another church of the same age, that of Sti. Angeli, at Perugia, which is very similar in its disposition. Of this building a section is here shown, as given by M. Isabelle—perhaps not quite to be depended upon in every respect, but still affording a very fair representation of what the arrangements of the circular wooden roofed churches were. Its dimensions are much less than those of San Stefano, being only 115 ft. in diameter; but it is more regular, the greater part of its materials being apparently original, and made for the place they occupy. In the church of San Stefano, the tomb-shaped circular form was probably used as symbolical of his martyrdom. That at Perugia was most likely originally a baptistery, or it may also have been dedicated to some martyr; but in the heart of Etruria this form may have been adopted for other reasons, the force of which we are hardly able at the present day to appreciate, though in all cases locality is one of the strongest influencing powers in so far as architectural forms are concerned.

[Illustration: 426. Section of Sti. Angeli, Perugia. (From Isabelle, ‘Édifices Circulaires.’) No scale.]

[Illustration: 427. Plan of Baptistery at Nocera dei Pagani. Double the usual scale, or 50 ft. to 1 in.]

At Nocera dei Pagani, on the road between Naples and Salerno, there is an extremely beautiful circular church, built undoubtedly for the purpose of a baptistery, and very similar in plan and general arrangement to the tomb of Constantia, now known as the Baptistery of Sta. Agnese, though somewhat larger being 80 ft. in diameter. Its principal merit is the form of its dome, which is not only correct in a scientific point of view, but singularly graceful internally. Externally this building for the first time introduces us to a peculiarity which had as much influence on the western styles as any of those pointed out above. As before observed (p. 540), the early Romanesque architects never attempted to vault their rectangular buildings, but they did frequently construct domes over their circular edifices. But here again they did not make the outside of the dome the outline of their buildings, as the Romans had always done before the time of Constantine, and as the Byzantines and Saracens invariably did afterwards; but they employed their vault only as a ceiling internally, and covered it, as in this instance, with a false wooden roof externally. It may be difficult to determine how far this was a judicious innovation; but this at least is certain, that it had as much influence on the development of the Gothic style as the vaulting mania itself. In the 10th and 11th centuries many attempts were made to construct true roofs of stone, but unsuccessfully; and from various causes, which will be pointed out hereafter, the idea was abandoned, and the architects were forced to content themselves with a stone ceiling, covered by a wooden roof, though this became one of the radical defects of the style, and one of the principal causes of the decay and destruction of so many beautiful buildings.

[Illustration: 428. Section of Baptistery at Nocera dei Pagani. (From Isabelle, ‘Édifices Circulaires.’) No scale.]

RAVENNA.

Ravenna possesses several circular buildings, almost as interesting as those of the capital; the first being the baptistery of St. John belonging to the original basilica, and consequently one of the oldest Christian buildings of the place. Externally it is a plain octagonal building, 40 ft. in diameter. Internally it still retains its mosaic and other internal features added in the 5th century, which are singularly elegant and pleasing. Its design is somewhat like that of the temple at Spalato, but with arcades substituted everywhere for horizontal architraves; the century that elapsed between these two epochs having sufficed to complete the transition between the two styles.

[Illustration: 429. Plan of St. Vitale, Ravenna. (From Isabelle.) Scale 100 ft. to 1 in.]

Far more interesting than this is the great church of St. Vitale, the most complicated, and at the same time, perhaps, the most beautiful, of the circular churches of that age. In design it is nearly identical with the church of St. Sergius at Constantinople (see Woodcut No. 311), from which it was undoubtedly copied, and probably by Greek artists from that town. It was built in the reign of Justinian by St. Ecclesius, archbishop of the see, and was consecrated in 547, eight years after the taking of Ravenna by Justinian’s generals. The principal difference of the plan lies in its being enclosed within an octagon instead of a square, as in St. Sergius, probably to mask the irregularity of the main entrance from a street which did not run in the direction of any of the cardinal points. The recesses are loftier in proportion than those of St. Sergius, and in the lower storey arcades take the place of beams. The aisles being covered with timber roofs, it was necessary to raise the walls of the octagon higher than those of St. Sergius, and small arches take the place of the usual pendentives: the springing of the dome, which is 50 ft. in diameter, is on the level of the sill of the windows the arches of which therefore form penetrations into the dome.

[Illustration: 430. Section of St. Vitale, Ravenna. (From Isabelle.) Scale 50 ft. to 1 in.]

The church is built in bricks with thick mortar joints, the dome being constructed in an ingenious manner with hollow pots fitted the end of one into the mouth of the other; the lightness of this vault has enabled the builders to dispense with the immense arches and buttresses found in St. Sergius and in Sta. Sophia. Similar construction with pots had been employed in the East for domes and roofs,[291] and they form as permanent a method as stone itself, in addition to the stability, facility of construction, and lightness which such an expedient affords.

[Illustration: 431. Capital in St. Vitale, Ravenna.]

Internally a good deal has been done in modern times to destroy the simplicity of the original effect of the building; but still there is a pleasing result produced by alternating the piers with circular columns, and a lightness and elegance about the whole design that render it unrivalled in the western world among churches of its class. This seems to have been admitted by its contemporaries as much as it is in modern times. Charlemagne at all events copied it for his own tomb at Aix-la-Chapelle, and the architects of many other circular buildings of that age appear to have derived their inspiration from this one.

[Illustration: 431_a_. Capital in St. Vitale, Ravenna.]

The church of San Lorenzo at Milan, had it not been so much altered in modern times, would take precedence of San Vitale in almost every respect. The date of its erection is not known, though it certainly must be as early as, if not earlier than, the time of Justinian. Down to the 8th century it was the cathedral of the city. It was burnt to the ground in 1071, and restored in 1119; the dome then erected fell in 1571, on which it underwent its last transformation from the hands of Martino Bassi and Pellegrini, who so disfigured its ancient details as to lead many modern inquirers to doubt whether it was really so old as it was said to be.

Its plan, however, seems to have remained unchanged, and shows a further progress towards what afterwards became the Byzantine style than is to be found either in St. Sergius or in San Vitale. It is in fact the earliest attempt to amalgamate the circular church with one of a square shape; and except that the four lateral colonnades are flat segments of circles, and that there is a little clumsiness in the angles (due possibly to the additions made in 1119 and 1571, when the plan of the dome was changed to an octagon, the original dome being probably circular, and carried on four spherical pendentives), it is one of the most successful designs handed down from that early age.

The dome as it now stands is octagonal, which the first dome certainly could not have been. Its diameter is 70 ft., nearly equal to that of the Minerva Medica, and the whole diameter of the building is internally 142 ft.

[Illustration: 432. Plan of S. Lorenzo at Milan. (From Quast, ‘Altchristlichen,’ &c.) Scale 100 ft. to 1 in.]

In front of the church, in the street, is a handsome colonnade of pillars, borrowed from some ancient temple—it is said from one dedicated to Hercules; this leads to a square atrium, now wholly deprived of its lateral arcades; and this again to a façade, which has been strangely altered in modern times. Opposite this, to the eastward of the church, is an octagonal building, apparently intended as a tomb-house; and on the north side a similar one, though smaller. On the south is the baptistery, about 45 ft. in diameter, approached by a vestibule in the same manner as that of Constantine at Rome, and as in the tomb of his daughter Constantia: all these, however, have been so painfully altered, that little remains besides the bare plan of the building; still there is enough to show that this is one of the oldest and most interesting of the Christian churches of Italy.

The building now known as the baptistery at Florence is an octagon, 108 ft. in diameter externally. Like the last-mentioned church, it was originally the cathedral of the city, and was erected to serve as such apparently in the time of Theodelinda, queen of the Lombards. If this was so, it certainly had not originally its present form, and most probably those columns which now stand ranged round the walls at that time stood in the centre, as in the Roman examples. If the original roof was of wood, it was probably in two storeys, like that of the baptistery of Constantine, or it may have been a dome of more solid materials, like that of the Sta. Costanza.

At the same time when the new cathedral was built, the older edifice appears to have been remodelled both internally and externally by Arnolpho da Lapo, and both its form and decoration so completely changed, that it must now be considered rather as a building of the 13th century than of the 6th, in which it seems originally to have been erected.[292]

[Illustration: 433. Half Section, half Elevation, of the Baptistery at Novara. (From Osten.) No scale.]

The baptistery of Novara, which may date from the time of Charlemagne, is interesting in that it contains the germ of those external galleries under the roof which form not only one of the most common but also one of the most beautiful features of the later Lombard and Rhenish churches. From the elevation (Woodcut No. 433) it will easily be seen what was the motive and use of this arrangement, the first trace of which dates perhaps as far back as the baptistery of Nocera (Woodcut No. 428); for wherever a wooden roof was placed over a circular vault, it is evident that the external walls must be carried up higher than the springing of the arch. But it was by no means necessary that this additional wall should be so solid as that below it, and it was necessary to introduce light and air into the space between the stone and the wooden roofs. Add to this the incongruity of effect in placing a light tiled wooden roof on a massive solid wall, and it will be evident that not only did the exigencies of the building, but the true principles of taste, demand that this part should be made as light as possible. Such openings as those found in the baptistery at Novara suggested an expedient which provided for these objects. This was afterwards carried to a much greater extent. At first, however, it seems only to have been used under the roofs of the domes with which the Italians almost universally crowned the intervention of naves and transepts, and round the semidomes of the apses; but so enamoured did they afterwards become of this feature, that it is frequently carried along the sides of the churches under the roof of the nave and of the aisles, and also—where it is of more questionable taste—under the sloping naves of the roof of the principal façade.

There is nothing in the Lombardian and Rhenish styles so common or so beautiful as these galleries, the arcades of which have all the shadow given by a cornice without its inconvenient projection, while the little shafts with their elegant capitals and light archivolts have a sparkle and brilliancy which no cornice ever possessed. Indeed so beautiful are they, that we are not surprised to find them universally adopted; and their discontinuance on the introduction of the pointed style was one of the greatest losses sustained by architectural art in those days. It is true they would have been quite incompatible with the thin walls and light piers of pointed architecture, but it may be safely asserted that no feature which these new styles introduced was equally beautiful with those galleries which they superseded.

[Illustration: 434. Tomb of Galla Placidia, Ravenna. (From Quast.) No scale.]

There can be little doubt that many other similar buildings belonging to this age still exist in various parts of Italy; for it is more than probable that, at a time when the city was not of sufficient importance, or the congregation so numerous as to require the more extended accommodation of the basilica, almost all the earlier churches were circular. They either, however, have perished from lapse of time, or have been so altered as to be nearly unrecognisable. We here, in consequence, come again to a break in the chain of our sequence; and when we again meet with any circular buildings in Italy, their features are so distinctly Gothic or Byzantine, that they must be classed with one or other of these modifications. The true Romano-Byzantine style had nearly come to an end when Alboin the Lombard had made himself master of the greater part of Italy about the year 575.

Before leaving this branch of the subject there are two small buildings at Ravenna which it is impossible to pass over, though their direct bearing on the history of this subject is not so apparent as it is in the case of other buildings just described. The first and earliest is the tomb of Galla Placidia (Woodcut No. 302), now known as the church of SS. Nazario and Celso, and must have been erected before the year 450. It is singular among all the tombs of that age from the abandonment in it of the circular for a cruciform plan. Such forms, it is true, are common in the chambers of tumuli and also among the catacombs, while the church which Constantine built in Constantinople and dedicated to the Apostles, meaning it however as a sepulchral church, was something also on this plan. Notwithstanding, however, these examples, this must be considered as an exceptional form, though its diminutiveness (it being only 35 ft. by 30 internally) might perhaps account for any caprice. Its great interest to us consists in its retaining not only its primitive architectural form (which is that of a dome carried on pendentives, and one of the few instances in which both dome and pendentives form part of one sphere), but its polychromatic decorations nearly in their original state of completeness (Woodcut 302). The three arms of the cross forming the receptacles for the three sarcophagi is certainly a pleasing arrangement, but is only practicable on a small scale.

[Illustration: 435. Capital of Pillars forming peristyle round Theodoric’s Tomb. (From Hubsch.)]

[Illustration: 436. Plan of Tomb of Theodoric. Scale 50 ft. to 1 in.]

[Illustration: 437. Elevation of Tomb of Theodoric, Ravenna. (From Isabelle, ‘Édifices Circulaires.’)]

Far more interesting than this—architecturally at least—is the tomb of Theodoric, the Gothic king, now known as Santa Maria Rotunda. The lower storey is a decagon externally, enclosing a cruciform crypt. It is 45 ft. in diameter, each face being ornamented by a deep niche. These support a flat terrace, on which originally stood a range of small pillars supporting arches which surrounded the upper storey. These have all been removed, though their form can be restored from fragments found, and as shown in Woodcut No. 435. On the face of the tomb itself are the sinkings for the architraves and vaults which they supported. The most singular part of the building is the roof, which is formed of one great slab hollowed out into the form of a flat dome—internally 30 ft. and externally 35 ft. in diameter—and which certainly forms one of the most unique and appropriate coverings for a tomb perhaps anywhere to be found. Near the edge are a series of projecting bosses, which evidently were originally used as handles, by means of which the immense mass was raised to its present position. In the centre of the dome is a small square pedestal, on which, it is said, once stood the urn which contained the ashes of its founder.

The model of this building seems probably to have been the Mole of Hadrian, which Theodoric saw, and must have admired, during his celebrated visit to Rome. The polygonal arrangements of the exterior, and the substitution of arcades for horizontal architraves, were only such changes as the lapse of time had rendered indispensable. But the building of the ancient world which it most resembles is the Tour Magne at Nîmes. In both cases we have the polygonal basement containing a great chamber, and above this externally the narrow ledge, approached by flying flights of steps. We cannot now tell what crowned the French example, though the fact of an urn crowning the tomb at Ravenna points to an identical origin, but we must obtain a greater number of examples before we can draw any positive conclusions as to the origin of such forms. Meanwhile, however, whether we consider the appropriateness of the forms, the solidity of its construction, or the simplicity of its ornaments and details, this tomb at Ravenna is not surpassed by any building of its class and age.

Though the investigation of the early history of these circular forms of churches is not so important as that of the rectangular basilicas, it is extremely interesting from the influence they had on the subsequent development of the style. In Italy it is probable that one-half of the early churches were circular in plan; and one such is still generally retained attached to each cathedral as a baptistery. Except for this purpose, however, the form has generally been superseded: the rectangular being much easier to construct, more capable of extension, and altogether more appropriate to the ritual of the Christian community. In France the circular form was early absorbed into the basilica, forming the Chevet or apse. In Germany its fate was much the same as in Italy, but its supersession was earlier and more complete. In England some half-dozen examples are known to exist, and in Spain they have yet to be discovered.

Had the Gothic architects applied themselves to the extension and elaboration of the circular form with the same zeal and skill as was displayed in that task by their Byzantine brethren, they might probably have produced something far more beautiful than even the best of our mediæval cathedrals; but when the Barbarians began to build, they found the square form with its straight lines simpler and easier to construct. It thus happened that, long before they became as civilised and expert as the Easterns were when they commenced the task, the Westerns had worked the rectangular form into one of considerable beauty, and had adapted it to their ritual, and their ritual to it. It thus became the sacred and appropriate form, and the circular or domical forms were consequently never allowed a fair trial in Western Europe.

SECULAR BUILDINGS.

Very few remains of secular buildings in the early Christian style are now to be found in Italy. The palace of Theodoric at Ravenna, though sadly mutilated, is perhaps the best and most perfect. In all its details it shows a close resemblance to that of Diocletian at Spalato, but more especially so to the Porta Aurea and the most richly and least classically decorated parts of that edifice, but much intermixed with mouldings and details which would seem to belong to a later style.

Another building, though perhaps of earlier date, is that which is now called the Palazzo delle Torre at Turin, and which still retains the architectural ordinance of the exterior of a Roman amphitheatre, but so modified by common sense that the pilasters are frankly accepted as purely decorative features, having only a slight projection. A similar style of work is found at Bordeaux in what is known as the “Palais Gallien,” but which in reality is a fragment of an amphitheatre built by the Emperor Gallienus (260-268 A.D.). The example at Turin is built with brick of large dimensions 15 in. by 11 in., which, coupled with its character and style, has led M. Cattaneo to ascribe it to the 3rd or 4th century of our era; the paucity of contemporary examples, however, renders it extremely difficult to trace the exact history of the style at this age.

[Illustration: 438. Palazzo delle Torre, Turin. (From Osten’s ‘Bauwerke in der Lombardei.’)]

In so progressive an art as architecture it is always very difficult, sometimes impossible, to fix the exact date when one style ends and another begins. In an art so pre-eminently ecclesiastical as architecture was in those days, it will probably be safer to look in the annals of the Church rather than in those of the State for a date when the debased-Roman expired, giving birth, phœnix-like, to the Romanesque. Viewed from this point there can be little doubt but that the reign of Gregory the Great (A.D. 590 to 603) must be regarded as that in which the Latin language and the Roman style of architecture both ceased to be generally or even commonly employed.

After this date we wander on through five centuries of tentative efforts to form a new style, and in the age of another Gregory—the VII.—we find at last the Romanesque style emancipated from former traditions, and marching steadily forward with a well-defined aim. What had been commenced under the gentle influence of a Theodelinda at Florence in the year 600, was completed in the year 1077 under the firmer guidance of a Matilda at Canossa.

##