Nimes (Nemausus). Situated at no great distance from Arles, and at the base of the hills which bound the plain of the Rhone, Nimes formed the capital of the Volces Arecomiques (or inhabitants of the flat country). In B.C. 121 it submitted voluntarily to Rome, and a few years B.C. Augustus planted a colony there. Being enriched with baths, &c., by Agrippa, Nimes soon became an important town surrounded with walls and towers, and provided with all the usual public buildings. It had reached the height of prosperity when it was ravaged by the Vandals in 407. In 472 it fell under the power of the Visigoths, who established themselves in the town, and made the amphitheatre their fortress. After suffering the usual course of sieges and destruction by the Saracens and Franks, Nimes early declared itself a Republic. In 1185 it came under the suzerainty of the Count of Toulouse, in which condition it continued to flourish till it finally passed to France under Louis VIII., along with the other domains of the Count of Toulouse after the Albigensian wars.

Although Nimes was a comparatively obscure town in the days of the Empire, the remains of its Roman monuments are the finest in Southern Gaul. The Amphitheatre (Fig. 19) is not quite so large as that at Arles, nor is the interior (Fig. 20) so well preserved, but the exterior is more complete. It measures 437 feet by 332 feet, with thirty-two rows of seats which contained about 20,000 spectators. The amphitheatre is now well seen, owing to the removal of the paltry buildings which had invaded it both within and without. Like all such Roman works it is constructed with the most massive materials, built without cement, and all bound together with solid stone lintels and arches. Fig. 21, a view in the corridor on the first floor, gives some idea of the colossal strength of the masonry. But these great stone lintels, massive as they are, indicate a vicious form of construction, many of them being cracked and shattered by the weight of the arches resting upon them. The exterior is of the usual design of such edifices having two arcades superimposed one on the other, with upright pilasters, or engaged columns, between the arcades supporting horizontal entablatures. Each arcade has sixty arches. The pilasters of the ground tier are square, and have no base, while the engaged columns of the upper tier are round and of the Doric order; above the latter is the attic, partly demolished, but still containing 120 bold consoles with holes to receive the masts which supported the velarium or awning.

There are four principal entrances at the four cardinal points; that of the North ornamented with a cornice resting on two bulls’ fore quarters. Similar ornamental bulls were introduced in the Temple which stood where the Cathedral is now built, and on the fine gate of Augustus of this city. Some therefore think it a kind of coat of arms given by the Emperors to the town. Others imagine that these features were adopted in order to flatter the Emperor Augustus, some bulls’ heads having been sculptured on the house in which he was born. A few sculptures are still visible on the amphitheatre, including two gladiators, and the Roman wolf.

A very large part of the ornament is left in block, only the western division being finished, the carving of the remainder never having been completed. The podium surrounding the arena is low, as at Arles, thus confirming Mérimée’s views as to the provisions which required to be adopted for the safety of the audience.

The interior has been greatly restored, so as to make it available for a large modern audience, and the amphitheatre is now used, amongst other exhibitions, for the annual branding of the young bulls of the Camargue, which, from the lively description of it given by Alexandre Dumas, seems to be a stirring spectacle, not unworthy of this classic arena.

In ancient times the lowest or first series of seats was set apart for the senators; the second series for the knights; the third for plebeians; and the top rows for the slaves. The last being the most quarrelsome it was considered desirable to endeavour to prevent squabbling by marking off each person’s seat. This was effected by means of lines cut in the stone, some of which are still visible in situ.

Some years ago there existed in the first precincts divisions similar to those of boxes in modern theatres.

The celebrated Maison Carrée at Nimes (Fig. 22) is probably the purest piece of Roman work to be found north of the Alps, and cannot fail strongly to impress the beholder, especially if he here sees for the first time a genuine Roman temple. The design doubtless owes much of its beauty and purity to the Grecian spirit of the locality. The building is small, being only about 80 feet by 40 feet. The portico, with its ten Corinthian columns, and enriched pediment, is very fine; but the effect of the flank view, in which the columns attached to the cella are visible, is not so satisfactory. The temple is surrounded by thirty columns in all, including those of the portico, which stand free, and the engaged columns of the flanks and rear. This is what is called the pseudo-peripteral plan—the true peripteral temple having the columns detached so as to form an ambulatory all round the cella. The former is the arrangement usual in Roman temples, which, according to Fergusson, never follow the genuine peripteral type. It is,

however, worthy of note, that the Plan of this building with its deep porch is rather Italo-Etruscan than Greek, and thus adheres to the traditional type observed by the Romans.

The cornice is perhaps rather over-enriched and is indicative of a late date, when classic art was in decadence; but the frieze is beautifully designed, and the style as a whole is remarkably pure and elegant.

Various ingenious attempts have been made to decipher the letters of the bronze inscription (which were originally fixed on the frieze of the portico), by means of the holes formed by the bolts which attached them to the stone work. The reading which seems most probable from its agreeing with the style of the building, indicates that it was dedicated to two nobles distinguished with the title of “princes of youth.” It is as follows:—

M. CAESARI AUGUSTI F. COS. L. CAESARI
AUGUSTI F. COS. DESIGNATO PRINCIPIBUS JUVENTUTIS.

This inscription necessarily places the Temple in the age of the Antonines, since the only princes known to whom the above names and title of Principes Juventutis will apply, after the sons of Agrippa, were Marcus Aurelius, and Lucius Verus, adopted sons of Antoninus Pius.

From excavations around the edifice it has been ascertained that the Temple formed a centre from which colonnades extended on either hand. It thus probably stood at the end of a Forum, the colonnades around which enclosed shops and places of business or pleasure. This edifice has passed through many vicissitudes; and it is marvellous how it has survived all the various uses or abuses to which it has been subject. It was naturally in the course of events first changed from a Pagan Temple into a Christian Church; in the eleventh century it formed the council chamber of the municipal body; and at a later time it was degraded into a stable, when the flutings of the columns were grated off to allow carts to pass between them. It then became attached to an Augustinian Convent, and was used as a mausoleum and place of burial. More recently it was occupied as the Hall of meeting of the revolutionary tribunal, and still later as a corn market. Now it has been put in good order, and contains the local museum of antiquities. This Museum comprises some good sculpture, especially a fine statue of Venus (Fig. 23), and numerous antiquarian fragments,—many for want of room being ranged round an enclosure in the open air. Portions of Roman mosaics and foundations of an earlier Roman building have been discovered under the soil of the Maison Carrée, thus shewing that it has been erected at a later period than the first occupation of the site by the Romans.

According to Mérimée the style accords with the time of the Antonines, when the decadence had begun, and when richness and multiplicity of details replaced the simple majesty of the first century. He also points out various irregularities in the structure which would never have been tolerated in the earlier period,—such as, that the columns are not equally spaced, that there is an unequal number of modillions on the opposite sides, that the caps are too low, and the shafts of the columns too long (being 10¼ diameters in height). But notwithstanding these defects the Maison Carrée is a building of which Nimes and France may well be proud.

The Nymphæum or Temple of the Nymphs at Nimes (Fig. 24), with its accompanying fountain, is another charming and quite unique structure. The fountain bursts forth in great abundance at the base of a hill called Mont Cavalier. It is enclosed in a space which was formerly a Roman Bath, and is then led away through wide open conduits or canals, all lined with stone and faced with pilasters. The whole is situated in a pretty public garden to which the fountain gives a special character. In this garden too, are found the ruins of the above temple, formerly called of Diana, which, however, is now supposed to have been a Nymphæum, or Temple dedicated to the Nymphs, and forming part of the Baths. The interior contains twelve niches of good design, and the roof was constructed with large stone arches or transverse ribs, between which the space was filled in with a plain waggon vault or flags of stone. This kind of vaulting was also adopted, as will be further explained afterwards, in the construction of the early Christian churches of Syria, and had undoubtedly great influence on the design of the first vaulted churches of Provence. The Nymphæum now contains a museum of busts and statues. This temple is shewn, by an inscription, to have been built along with the Baths in the time of Augustus. The variety and elegance of its details are further evidence of the Grecian taste of the people of the district. The aqueduct from the Pont du Gard terminated in a reservoir near this point.

The Tour Magne (Fig. 25), on the top of the hill above the Fountain of the Nymphs, is a Roman building, the object of which has given rise to much discussion, without any definite conclusion being arrived at. It seems, however, most likely to have been a mausoleum. The plan is octagonal, and the walls are built with rough ashlar. The structure is hollow, and from 90 to 100 feet high. It was attached to the walls of Augustus, and in later times was converted into a fortress by the Count of Toulouse. The general resemblance of the design of this monument to that of Augustus at La Turbie, which we shall meet with further on, is very striking.

Two of the Roman gates of Nimes remain. The Porte d’Auguste, founded B.C. 16, has a double arch for vehicles, and two side openings for foot passengers flanked by two towers. Like the Roman gates of Autun these two towers contained stairs leading to the walls, and formed posts of observation. The other gate, the Porte de France, lies to the west of the amphitheatre, and has one wide archway.

Pont du Gard (Fig. 26). This magnificent specimen of Roman engineering is situated at a distance of about 13 miles N.E. from Nimes, on the way to Avignon, and can now be reached by rail. It formed part of an aqueduct (partly in tunnel and partly in open canal) of about 25 miles in length, which brought an ample supply of water to Nimes. This work is said to have been built by M. Agrippa, son-in-law of Augustus, 19 years B.C. It has thus for 1900 years defied all the attacks of man, both barbarian and civilised, as well as the elements, to which so many other Roman monuments have succumbed, and still stands almost as perfect as at first. The arcades abut at either end on the slope of the hills at the base of which flows the river Gardon. The aqueduct measures 160 feet in height, and 882 feet in length on the top. It is composed of three stages, all built with enormous blocks of stone placed together without cement, and presents probably the most stupendous example of the solidity of Roman workmanship in Gaul. But it is roughly and irregularly constructed, as if utility alone had been considered, and no regard paid to beauty; the arches are unequal in span, and the structure itself is bent in its length. The arches are constructed, those of the two upper arcades with three, and those of the lower arcade with four distinct courses in the breadth of the structure. These courses are composed of stones of immense size, placed side by side, but not otherwise joined together. Above the upper tier lies the conduit for the water, 5 feet high, and 2 feet wide, covered over with immense flags, which even the Goths seem to have despaired of being able to destroy. The conduit is lined with strong Roman cement, which still remains sound and good. The projecting blocks observed on the flanks and under the arches were intended to receive scaffolding for the execution of repairs, should these ever be required in a work so simple and substantial. “What a grand faith,” exclaims Mérimée, “must the constructors of this aqueduct have had in the eternal duration of the Roman Empire, when they made provision for repairing this gigantic and enduring work!” The bridge placed alongside the lower arches is of modern construction, having been erected in 1743.

Leaving Arles for Marseilles we traverse a country as bare and uninteresting as an African desert. To the right, on the western side of the Rhone, lies the great plain of the Camargue, the delta of the river, composed of mingled salt mud and stagnant pools, the result of the contest between the waters of the Rhone and the sea; the former constantly pouring down immense volumes of débris, and the latter, obeying the impulse of the wind, as constantly driving it back upon the land. But the railway, keeping on the eastern side of the river, runs through a different but not less remarkable plain called the “Crau.” This consists of an immense accumulation of shingle, composed of water-worn and rounded stones of all sizes—the fabled scene of the fight of Hercules with the Ligurians, when Jupiter rained down these stones to provide the hero with ammunition. This extensive plain was a barren wilderness until a system of irrigation was introduced by the construction of the Canal de Craponne, whereby the water of the Durance is brought down for its fertilisation. Having at last crossed the Crau we arrive at St Chamas, where the eye is relieved by the bright and peaceful prospect over the Etang de Berre, an extensive branch of the Mediterranean almost entirely surrounded with land. St Chamas is a quaint old town, with some of its houses hollowed out of the rock and traces of ancient ramparts. About half-a-mile distant may be seen an interesting Roman Bridge called the Pont Flavia. It is constructed with the usual solid masonry, and spans the river Touloubre with one arch, which is abutted by the rocky banks. The entrance at either end to the roadway over the bridge is through an arch, decorated with Corinthian columns and entablature. These archways are well preserved and are illustrated in Fergusson’s “Handbook of Architecture.” The columns are surmounted with lions, and the frieze bears an inscription shewing that the structure was erected by one of the Flavii.

Some distance north from this, on the slope of the chain of the Vernégues, which divides the valley of the Durance from that of the Rhone, are to be found the relics of a small Corinthian temple, originally preceded with a peristyle of four columns in front, and pilasters of return on each side, of which, however, only one single pillar now survives. This was doubtless the site of the ancient Ernaginum.

In early Christian times this temple seems to have been converted into a church, and a circular-headed window opened in the wall of the cella. A chapel dedicated to St Césaire was in the tenth century erected against the north wall, with a door into the main church, now built up. The temple is well illustrated in Texier and Pullan’s “Byzantine Architecture,” and is said to be “full of the sentiment of pure Greek art.” The carving of the capital, as shewn in Texier’s drawing, is in the best style. “The proportions of the entire column, which are excellent,” says Texier “and the foliage of the capital, which seems to have been inspired by that of the monument to Lysicrates, prove that this little building, concealed amongst the mountains of Provence, was the work of a Greek artist of the colony of Massilia.”

It has already been pointed out how capriciously the Roman remains have been preserved in Southern Gaul. While a small provincial town like Nimes possesses so many splendid examples, the great and ancient cities of Marseilles and Narbonne have scarcely a single relic of their Greek or Roman civilisation left. At Marseilles some fragments of walls with an archway and some subterranean vaults under the Church of St Sauveur are the only remains of the splendid edifices which no doubt once adorned this ancient and important city.

All along the coast between the Rhone and the Pyrenees, many towns existed and flourished under the Empire, but there is now scarcely a fragment of Roman work to be found in the whole province.

Leaving therefore for the present this south-western district, we shall now follow the great Aurelian way which conducted from Spain and Gaul eastward into Italy. This road passes through the celebrated Riviera, the favourite winter resort of the delicate from every country in Europe, and even from America. It consists of a narrow strip of land between the lower spurs of the Alps and the sea; but this level strip is frequently interrupted by branches or roots sent down from the mountains which run out as Capes into the Mediterranean, enclosing in their arms beautifully sheltered sunny bays, each having a town or village of its own. The Roman road clung to the mountains, the engineers finding it easier to span with bridges the higher rugged ravines of the torrents than the broad channels of the rivers near their mouths, where the shingly and shifting foundation was found insecure. Of the towns and stations which existed along this route in Roman times, some vestiges may still be traced.

Toulon, now the great naval arsenal of France in the Mediterranean (formerly Telo Martius), contains no Roman buildings; but some miles to the eastward, on the road by the coast leading to Hyères, the ruins of an ancient Roman town called Pomponiana have been discovered and partly excavated—exposing to view portions of the walls of houses, vaults, walls of enceinte, frescoes, fragments of sculpture, aqueducts, baths, &c. The wall of a quay presents the peculiarity of being built above a basement formed of large cubes of stone, superimposed, but not united with cement, which seems to be of Cyclopean work.

Moving eastward we pass Le Luc (Forum Voconii) in the middle of the fertile “garden of Provence,” where one Roman sculpture of a boar hunt has been preserved; and following the course of the river Argens, with the rocky mountains of Les Maures on the right we arrive at Fréjus, an important sea-port in Roman times, and then known as Julii Forum.

This town is supposed to have been first occupied by the Phœnicians, and afterwards by the Greek colonists. It was enlarged and improved by Julius Cæsar and Augustus. It then possessed a valuable harbour at the mouth of the river Argens, to which Augustus sent the fleet of galleys which he took from Anthony at the battle of Actium; but the sediment of the river has now silted up the harbour, and formed a flat plain of about a mile in breadth between the ancient port and the sea. The protecting walls of the harbour, with a solid obelisk at the end, which no doubt marked the entrance, still remain, but are now high and dry on the plain. Adjoining these are the walls of a strong fort or castellum for the protection of the port, built with Roman masonry of small sized cubic stones. The “Porte dorée,” is an archway close to the railway, built with similar masonry, divided with courses of brick work, now greatly restored and renewed. It is

supposed to have been the gate between the port and the town. Some ruins of the baths have been discovered adjoining this. Considerable remains of the ancient Roman city walls, enclosing five times the extent of the present town, still remain. Close to the railway station relics of the “gate of Gaul,” and other Roman works are observable. Following these from the railway station towards the left, the ruins of the Roman Amphitheatre (Fig. 27), through which the public road passes by a picturesque archway, are soon reached. The interior is fairly preserved, together with the arches which sustained the seats, staircases, &c., but the exterior walls and arcades (if the building ever had an ornamental exterior, which is doubtful) are now completely awanting. The Amphitheatre is 375 ft. long by 273 ft. wide. The east side rests on the slope of a hill, so that little building was required in that position, but the west side of the structure is raised from the level plain.

Continuing round the old walls of the town to the eastward, we find in a garden the ruins of a Roman Theatre. The dimensions of this building, which was of small size compared to those we have met with at Orange and Arles, are quite traceable, but the scena is gone all but the foundations, and only some walls and ruined arches of the auditorium remain above ground.

A little further round the walls, traces are observed of the great aqueduct which brought the water of the river Siagnolles to Fréjus from a distance of above 20 miles. On turning the north-east angle of the walls, the ruined piers of the aqueduct are seen stretching across the plain. At the above point the conduit is in a canal owing to the height of the ground. On reaching the main road leading from Fréjus to the eastwards, the aqueduct takes a sudden bend to the east, and follows the road for a considerable distance. At this bend was an entrance gate of the town, called the gate of Rome, a portion of which still exists. From here a branch canal took the water to the port. In its long course the aqueduct is sometimes in cutting, and sometimes carried on lofty piers and arches 87 feet wide. Those near the town (Fig. 28) are amongst the finest specimens, but some portions in the more remote valleys also still retain their arches, and at one place the aqueduct is carried in two parallel canals on separate arches.

Between Fréjus and Cannes, the Roman Via Aurelia passes inland through the chain of the Esterelle mountains, whence the Romans obtained much of the granite and porphyry found in their monuments. At Cannes and neighbourhood there are a few Roman relics. A bridge over one of the small streams which descend from the hills through the town is said (but this is doubtful) to be of Roman origin. A delightful walk of an hour from Cannes over the hills leads by Vallauris to Clausonne, where the well preserved remains of the Roman aqueduct (Fig. 29) which conveyed the water supply to Antibes are still to be seen.

At Antibes, the ancient Roman Antipolis, there are no Roman remains; but according to M. Lenthéric, a stone has been found here with a Greek inscription, giving proof of the ancient worship of the Hellenes in this region in the fifth century B.C.

At Vence, the ancient Ventium, a town some seven miles inland, a number of Roman inscriptions are built into the wall of the Cathedral, and two granite columns are preserved, which are supposed to have been anciently

presented to the town by the city of Marseilles (see Part VI.)

Crossing the wide and dangerous channel of the Var (formerly the boundary between France and Savoy) we arrive at Nice.

Nice (or Nizza), although now the most important town on the Riviera, possesses no ancient buildings. In Roman times Cemenelum (now Cimièz), the chief city of the Maritime Alps, stood on a lofty site about three miles up the river Paglione from the modern town. This ancient city has almost entirely disappeared, its only relics being the ruins of a small amphitheatre (Figs. 30 and 31), through the centre of which the public road now passes, and some excavated hypocausts in the garden of a villa adjoining. The amphitheatre measures 214 feet long by 178 feet wide, and it has been calculated that it was capable of containing about 8000 spectators. The form of the arena and the slope of the first series of seats can be distinctly seen, but otherwise the building is a complete ruin. A few of the perforated corbels for the support of the poles which carried the velarium may, however, be still observed on the exterior. But the want of architectural features is to some extent compensated by the grandeur of the views obtained from the walls, comprising the whole of the coast from Bordighera on the east, to the Cap d’Antibes on the west. Proceeding in that direction, a drive along the magnificent Cornice-road soon brings us to the ancient boundary between Gaul and Italy at

La Turbie (Turbia or Trophæa), a small town standing on an inland pass formed by a notch in the mountains, which here rise in great precipices directly from the sea. On this neck a trophy was built in commemoration of the victories of Augustus over the Alpine tribes. The monument (Fig. 32) has been of great size, and is built with large blocks of stone. It probably stood on a square base, on which was erected the great circular mass above. It was adorned with statues, and a colossal figure of the emperor crowned the top. The design would thus resemble a great many of the splendid mausoleums erected about that time in Italy. As above noticed this edifice bears a strong likeness to the Tour Magne at Nimes. The massive Roman work is still traceable in the lower parts filled in with rubble between. Fragments of an inscription have been found in the ruins commemorating the triumphs of the divine Emperor and High Priest Augustus. In mediæval times this monument was, as usual, converted into a fortress, as the work of the upper part still shews. It is executed in inferior masonry, and the cornice is Italian in character. The fortress was blown up by Marshal de Villars in the seventeenth century. The gateways of the town (see Part VI.) and other structures have been built with massive stones from the ruins of the trophy, which, as so often happens, has been used as a convenient quarry.

A splendid view of the coast is obtained from the summit, including Monaco, Monte Carlo, Mentone, and point after point to the eastward leading into Italy. But though we now stand on the borders of Italy, we should still have far to travel through the land ere we encountered such a fine series of Roman structures as those we have just been contemplating. Not till we reach Verona, or Rome itself, are monuments to be found comparable with the amphitheatres of Arles and Nimes, or the theatre of Orange; and there is probably no temple even in Rome so complete and striking in its unity and spirit as the Maison Carrée at Nimes. But our way lies not across this border. We must now turn back and follow in the later edifices the course of Roman Art after the Fall of the Empire, and the growth and development of the new styles which sprung from it.

V.

THE transition from the architecture of Rome to that of mediæval times forms one of the most interesting and instructive epochs in our art. The whole history of Roman art is that of a transition from the external trabeated style, with its horizontal entablature, which was common to the early races of Greece and Italy, to the complete development of the internal arched architecture, which was the final outcome of Roman constructional forms.

The leading features of that Italo-Greek architecture contain a reminiscence or survival of the primitive elements of wooden construction, from which they were doubtless traditionally derived, although in the course of time their origin had been lost sight of. Thus the upright pillars with their flutings are idealised descendants of the Egyptian column, which again represents a bundle of reeds tied together. The horizontal entablature is derived from the beams laid across the heads of the pillars, in accordance with the earliest and most natural mode of wooden construction. The pediment is the evident continuation (both in place and time) of the couples and ties of a wooden roof of the simplest and most primitive design; while the side cornice represents the projection of the eaves, and the triglyphs and modillions are the imitative survivals of the ends of the cross beams or ties and the sloping rafters of the wooden roof. For centuries this trabeated principle prevailed in Rome; but together with it there existed a disturbing element, which at first appeared to be small and insignificant, but which nevertheless contained the elements of the greatest revolutions in architecture which the world has yet seen. That little feature was the arch, the distinguishing principle of true stone construction—the seed containing the germ from which, through Roman cultivation, have sprung all the great families of mediæval architecture, whether Byzantine, Gothic, or Saracenic.

The earlier architecture of the Romans was doubtless chiefly derived from that of the Etruscans, who, like the Greeks, followed the trabeated principle. This origin is distinctly traceable in the plans of the Roman temples, which are never truly peripteral, or surrounded with a detached colonnade, like those of the Greeks, but have a deep portico at one end only, in front of the cella. Of this arrangement we have seen a beautiful example in the Maison Carrée at Nimes. But the Central Italians must have early received some impressions from the Hellenic art of Magna Grecia, and the way would thus be opened for the introduction at a later period of the finer developments of Greek architecture which were so universally followed during the Empire. Meanwhile the arch, the antagonistic element to the trabeated principle, was gradually progressing; and from its primitive obscure use in substructures, conduits, and similar engineering situations, it had forced itself into notice above ground, and had gained recognition in the elevations as a proper architectural element. Hence arose the combination, so conspicuous in the architecture of the Romans, of trabeated features, such as pilasters and entablatures, with the arched method of construction which they had adopted from an early period, and of which they ultimately shewed themselves such masters. The amphitheatres and the triumphal arches of the empire well illustrate this mixture of arched construction, as shewn in the round-headed wall openings, combined with trabeated decoration, in the form of horizontal entablatures supported on engaged columns or pilasters. This mixed style long prevailed, and examples of it are to be found in every part of the Roman world. But in later times, when purity of taste had begun to decay, the Romans gradually gave fuller scope to their noble constructive powers, and allowed them to find a worthier expression in their designs. This took place chiefly in their engineering works, such as the Pont du Gard, and in their interior architecture, as, for instance, in the great halls of the Baths of Diocletian and Caracalla, the Basilica of Constantine (or Maxentius) and other similar works, in some of which immense intersecting vaults were successfully executed. The simple barrel or tunnel vault is of very ancient origin, and was adopted by the Romans from the earliest times. They also freely employed round intersecting vaults for covering spaces of all sizes up to the great examples above referred to. But the most astonishing feat of the Romans in connection with vault construction is their adoption and application of the dome. In the Pantheon at Rome we have an example of that species of vault introduced at once in its perfect form in the largest example in the world. The portico of this temple belongs to the age of Augustus, and it is therefore thought by many that the rotunda and dome are of the same date. It is very remarkable that no smaller Roman domes of earlier date are to be found, and that this style should, as it were, be born in perfect manhood without having passed through the stages of infancy and growth. These no doubt existed, although we have as yet been unable to trace them. Possibly, as Professor Baldwin Brown suggests, the dome is of eastern origin, and its enlarged construction may have been worked out in some of the Hellenistic cities, such as Alexandria, where the earlier examples have now perished.

Along with the introduction of the above new and splendid development of vaulting in their interiors, the Romans still adhered in the decoration of their exteriors to the Italo-Grecian portico, with its entablature and pediment. It was not till the time of the Lower Empire that these elements came to be modified and slowly abandoned. The stages by which the trabeated forms were by degrees stripped off can, however, be distinctly traced. The arches and vaults employed in the baths, tombs, &c., no doubt conduced to that result. In these the arch became the important feature internally, and naturally in course of time it assumed a more prominent position externally also. Archivolts, or curved architraves running round the arches, such as were in common use in buildings like the Colosseum, had gradually intruded themselves amongst the Greek pilasters and entablatures of the exterior elevations; while in later edifices, such as the palace of Diocletian at Spalatro, the straight architrave was omitted, and only the arched one retained. The early Christian sarcophagi shew the same important step. In these a common design consists of an arcade containing the figure of an apostle in each arch, and these arches or archivolts spring directly from the caps of the columns, without any straight architrave being employed. Of this a good example has been given above, page 63.

In all transitional styles it is difficult, and indeed scarcely possible, to draw the line where one style terminates and another begins. This is especially difficult in connection with the passage from Roman to mediæval architecture. The latter was in fact for centuries not a different style but simply a continuation of that of the Empire.

After the adoption of Christianity the purposes to which the Christian buildings were applied was certainly very different from that of their prototypes, but the architecture was the same. The circular domed edifices raised by the Romans as mausoleums were imitated by the Christians in their circular baptisteries; while the style of construction employed in the great basilica or pillared hall lighted by a clerestory, was exactly copied in the nave or large vessel of the Christian Church. The continuity of style is complete; there is no break. The same Corinthian or Ionic pillars, the same entablatures, the same roofs and vault are used in both. So close is the resemblance between the Christian circular baptisteries (several of which we shall meet with in Provence) and Roman circular monuments, that the former are generally regarded as Roman temples converted to Christian uses. The early churches are usually called basilicas, and have hitherto been supposed to be derived from the Roman basilica. But Professor Baldwin Brown, in his recent interesting and learned work “From Schola to Cathedral,” endeavours to prove that this is not the case. The basilica had no doubt the form of a pillared hall with central and side aisles, the former lit by a clerestory, but it had no apse, or if there was one it did not occupy the prominent position of that feature in the early churches. The origin of the apse, which was an essential feature in all churches, containing as it did the seat of the Bishops in the centre and those of the presbyters on either side, is attributed by Professor Baldwin Brown to the memorial cellae erected by Pagans and Christians alike in the cemeteries. These often assumed a domed or apsidal form, and were much resorted to on saints’ natal days, for commemorative festivals and religious ceremonies, held in the cemeteries above the spot where the martyr’s bones reposed in the catacomb below. At a later time, when these relics had been transferred to crypts below the altars of the churches, the apse was a feature naturally introduced to complete the resemblance to the original tomb. As regards the nave, the scholae or halls of meeting of private societies are regarded by Professor Baldwin Brown as the principal model of the early church. Under the emperors the Christians were allowed to form burial guilds, and these, like other guilds, had their scholae. The schola often had an apse containing the seat of the president; and the above author is of opinion that the large churches built after the conversion of Constantine are rather enlarged scholae than copies of basilicas.

However this may be, the type of the early Christian church or basilica presented to view an elongated hall with two or four rows of pillars, dividing it into three or five aisles, with a lofty triumphal arch at the end of the central nave, leading into a wide open space raised some steps higher than the nave, and in which stood the altar. Beyond this was the invariable apse with its semi-domed ceiling adorned with mosaics, and containing, elevated by a few steps above the floor, the throne of the Bishop, and the seats of the Presbyters. The whole building was covered with an open wooden roof.

Some of these early churches have been preserved or restored in Rome—such as San Paolo fuori le Mura, Sta Maria Maggiore and San Clemente.

There is every reason to believe that the above was the usual form of early churches in the West. At Ravenna, which was the principal city in Italy during the Lower Empire, being the seat of the Exarch, the representative of the Emperor in the West, there are fine examples of the various kinds of early Christian religious edifices, dating from the fifth to the seventh century. The great Church or Basilica, used for the assembly of the whole congregation, is represented in St Apollinare Nuovo. It has the usual row of columns on either side of the nave, separating it from the side aisles, and supporting a flat upper wall splendidly decorated with mosaics, the whole being ornamented with Roman details. The upper portion of the wall is pierced with clerestory windows, and at the east end is the great apse.

The Baptistery or Ceremonial Church is as usual octagonal and is domed. Here also the walls are covered with fine mosaics.

Another extremely interesting building at Ravenna is the church of San Vitale. This edifice (whether designed as a monument or as a church is uncertain) is octagonal and domed, very much after the style of the temple of Minerva Medica and similar Roman structures.

San Vitale has a special interest from its having formed the model adopted by Charlemagne for the church which he erected at Aix-la-Chapelle, to serve also as his own mausoleum. It thus constitutes an example of a Roman design reproduced in Ravenna, under the late Empire, as a Christian structure, and again serving as a model for a mediæval mausoleum as late as the eighth century. This shows distinctly the continuity of Roman design and its direct influence on the art of later times.

The above three edifices at Ravenna present fully developed examples of the three chief buildings required in connection with the church services up to the ninth century, viz., the church, the baptistery, and the mausoleum. As we proceed we shall meet with proofs that the same classes of edifices were in use and were carried out in a similar manner in other parts of the Western Empire. The circular or octagonal baptistery is of frequent occurrence in Southern Gaul. Examples of circular churches are also not awanting, but there is every ground for believing that the basilican form of church, like that of St Apollinare, was the plan most generally adopted in Western Europe.

At Ravenna, an early circular tower or campanile, generally similar to the square ones at Rome and elsewhere, still exists. This is a feature the origin of which has not yet been accurately determined. The prevailing opinion, however, now is that these towers were at first erected as places of observation and defence, being in that respect somewhat similar in their conception to the round towers of Ireland. As in San Vitale, one form of a Roman octagonal-domed building is followed, so at San Lorenzo in Milan another design of a somewhat similar character is carried out, showing that the basilican form, although general, was not universal.

In consequence of the destruction caused by the invasions of the Barbarians, by fire or otherwise, very few edifices now exist in Western Europe of the time between Justinian and Charlemagne. During all that time of disaster in the West, the Eastern Empire still maintained itself in splendour, and gave encouragement to architecture and the fine arts. From an early time the Byzantine architects showed a preference for the dome over the intersecting vault, and it is possible to follow in the still existing edifices, the mode in which the domical form of roof was gradually worked out, until in the great church of Sta. Sophia, erected under Justinian, in the sixth century, the largest and noblest building of the style was successfully completed.

In the details of the style of the Lower Empire, as practised in the East, there is considerable evidence of Greek taste. The sharp thistle-like sculpture of the foliage is designed in a manner not unlike that of the Corinthian capitals of the Choragic Monument of Lysicrates at Athens. The Byzantines also excelled in flat and delicate carving, such as that generally executed in ivory or fine wood, and in ornamental metal work and jewellery. When the West began to revive, this Byzantine art naturally produced some influence on it. A very remarkable example of this occurs in the church of St Mark’s at Venice, erected about A.D. 950, which in every feature—in plan, in distribution of parts, in the use of the dome, and in its mosaic decorations,—is a distinct importation from Constantinople.

But the art of the East was destined to produce, at a later period, a much stronger effect, as we shall afterwards see, in Provence and Aquitaine. Besides the domical structures of Constantinople, another series of Christian buildings which had a great influence on Western architecture exists in the East. A large number of churches have been brought to light in Syria by the work of Count Melchior de Vogüé. These correspond in general features with the early churches of the West. They comprise a central nave and side aisles, separated by rows of piers, with nave arches thrown longitudinally between them. The nave is also crossed transversely with arches cast between the piers, and these are abutted by arches thrown over the side aisles. The latter, in order to resist the thrust of the central arch, require to be placed at a considerable height. The side aisles are thus rendered unnecessarily lofty, and are therefore divided into two storys with a floor which forms a gallery. The nave piers and their transverse arches are placed pretty close together in order to carry the great flag stones of which the roof is frequently composed, and which are supported upon them. Although the roof is in some cases flat, the general system of construction of these Syrian churches is very similar to what is found in the oldest churches of Southern Gaul; and which, as already mentioned, was also used in the Nymphæum at Nimes. There can be little doubt but that the Syrian structures were carefully studied by the numerous monks who visited the East in the eleventh century, while Palestine was in the hands of the Crusaders, and that they were thereby helped forward in the enterprise which was then absorbing the attention of the Western architects, viz., how to roof their churches with stone vaults.

Hitherto the Western basilicas had been roofed with timber. A few examples of these early basilicas have escaped the universal destruction, and serve to indicate what the other churches which existed before the eleventh century were like.

The Basse Œuvre at Beauvais is a well known specimen. It has a row of square piers on each side of the nave, separating it from the side aisles and carrying, on round arches, the upper walls containing the windows of the clerestory—the whole being covered in with a wooden roof. It was probably terminated to the east with a semicircular apse, and at the west with a narthex or porch.

These early churches were no doubt all of very simple construction, the only ornaments being the marble columns and carved work which in some localities were available from Roman buildings. Where these existed the style adopted naturally followed the Roman forms, but in districts where they were absent the style gradually passed into the Romanesque, under the influence of the new elements imported by the Northern invaders. We have seen how Charlemagne attempted to follow a Roman structure in his great church at Aix, and that is a distinct indication of the general tendency. The chief object at this period of transition was to produce an effective internal design, the exterior being invariably very simple. In this also the system by which Roman architecture had been developed continued to be carried out.

When the new political conditions of the different divisions of Europe had become somewhat settled, these principles were worked out separately and independently in each country and province, and produced a great variety of styles, all comprehended under the general title of Romanesque. They were in reality all derived from ancient Roman architecture, but by their very variety they indicate the new spirit which was now beginning to express itself.

As above mentioned the great desideratum in the eleventh century was a simple form of stone roof. The earlier wooden roof had been found so liable to destruction by fire, that great efforts were now made to provide a fire-proof covering.

At San Miniato, near Florence, there still stands a very fine basilica of the beginning of the eleventh century, which shews one method in which this was attempted to be done, and which recalls the mode of construction of the Syrian Churches above referred to.

San Miniato is divided into three long bays in its length by circular stone arches, springing from clustered piers, thrown across the nave, each bay being again subdivided by three longitudinal archivolts resting on simple pillars.

The above great transverse arches do not, as in the Syrian examples, carry the roof, which is in this instance of wood, and is thus not quite fireproof; but even if the timbers were destroyed by fire, the three transverse arches would tend to bind the structure all together, and prevent further ruin.