pavements or in wall decorations, are extremely rare, although fragmentary pieces of either are of very frequent occurrence; the non-geometrical patterns were no doubt more costly, and therefore, perhaps, less used in pavements than those of the geometrical class, the comparatively simple forms in which could be produced with less labour.

Such patterns were, however, very largely used in the decoration of walls, and were generally executed in glass; detached pieces are found in thousands in the ruins of villas near Rome, but instances in which the connection and disposition of the pieces can be traced are of extreme rarity; only three, as far as my knowledge extends, in which glass is the material can be cited.

While, therefore, tessellated work, whether in pavements or as decorations for walls, is familiarly known to us, and has been largely illustrated, sectile work, and especially the non-geometrical variety, has been but little noticed, but, as I hope to be able to show, the remains of it which exist, although but very scanty, well deserve the attention of the student of art or antiquity. Those in which figures of men or animals appear have especial claims to notice, but those consisting merely of patterns are also interesting from the fact that they preserve their original colours unaltered, and that we have comparatively very small remains of the polychromatic decorations of the ancients-at least, of those in which variety of colour and form, and not the representation of natural designs, was the primary object.

Before entering into a description of the existing remains it may, however, be well to say a few words on the history of this sectile work.

Pavements made of variously coloured marbles were in use among the Persians; in the Book of Esther (chap. i. 6) mention is made of "a pavement of red and blue and white and black marbles" in the palace of King Ahasuerus at Susa. The Greeks probably learnt from the Persians the art of constructing such work, but no existing monument can be referred to. One instance of the use of coloured glass in architectural decoration in the best period of Greek art, though trifling in itself, is very suggestive; it is in the temple of Minerva Polias at Athens; here pieces of blue glass are inlaid in the plaited torus between the volutes of the capitals of the portico."

Pliny (Hist. Nat. book xxxvi. c. 15) mentions a temple at Cyzicus, the walls of which were of polished stone, with threads of gold between the joints; it seems These authors say "coloured

a Stuart and Revett, Antiquities of Athens, vol. ii. p. 73, note a.

stones or glass," but Mr. H. March Phillips informed me that he had noticed these decorations, and that they were pieces of coloured glass.

probable that we should by this passage understand not that the building was constructed with plain rectilinear courses of white stone, but that coloured marbles formed patterns on the walls. In the former case the threads of gold would have produced little or no effect, in the second the beauty of the work would be greatly increased by the lines of gold parting off and harmonising the colours of the stones. That the intervention of a line of gold or of white between positive colours has that effect may be seen in many ornamental works of all ages and countries, and, in the case of the incrustations of walls, white lines of mortar often occur between the various pieces of marble, the work having evidently been purposely so constructed; instances of this may be seen in San Giovanni in Fonte at Ravenna and in Santa Sabina in Rome, both of the fourth or fifth centuries of

our era.

Of incrustations on walls in glass at Rome perhaps the earliest instance on record is that mentioned by Pliny (Hist. Nat. book xxxvi. c. 15), when he tells us that Scaurus, the stepson of Sylla, constructed a theatre of three storeys, the lowest of marble, the second of glass, and the third of gilded wood. It cannot be supposed that the glass was used in solid masses; doubtless it was attached to the walls in the form of "crustæ."

In the time of Cæsar, sectile pavements were in use, as Suetonius (book i. c. 46) states that that general was accustomed to carry with him on his campaigns both tessellated and sectile pavements. Casaubon, in a note on this passage, suggests that what he really carried with him were the materials for making such pavements. It was the practice to place the chair of a Roman official of high rank, e.g., a prætor or a consul, on an ornamented pavement, and it was no doubt with a view to the formation of such pavements that Cæsar carried these materials with him. Very beautiful examples of sectile decoration for walls have been discovered in the recent excavations made by the Emperor of the French in the portion of the palace of the Cæsars at Rome which was built by Nero. These, presenting very ingenious and elaborate patterns made up of forms chiefly non-geometrical, are wholly composed of marbles and granites or porphyries. Glass, however, was about the same time used for the same purpose, as is proved by examples discovered at Herculaneum and preserved in the museum at Naples. Excavations in the ruins of a villa on the Via Cassia (about four Roman miles from the Porta del Popolo), which is said to have belonged to Lucius Aurelius Verus, the son-in-law of Marcus Aurelius, brought to light a large quantity of these decorations executed in glass. Whether they can be assigned to the period of Lucius Verus or not, is a question not perhaps very easy to decide; my learned

friend Padre Garrucci was of opinion, that, judging from the character of the ornament, a period not very remote from that of Constantine was probably that of their execution. It may however be noticed, that pieces of glass identical in pattern and colour to these are to be found in the mosaics of San Andrea, from which we may infer that the two were constructed at periods not very remote from each other.

In the third century we have distinct reference to such employment of glass, for Vopiscus (in Vita Firmi, cap. 3) tells us of Firmus, “De hujus divitiis multa dicuntur; nam et vitreis quadraturis, bitumine aliisque medicamentis insertis, domum induxisse perhibetur." From this it would seem that such decorations were deemed uncommon and expensive luxuries when executed in glass, but they would appear to have been more common in marble. Richerius Rhodiginus, a writer of the fifteenth century, says, "In turribus et decursoriis mœnium Romæ videre est, opere tessellato expicta, distinctaque pavimenta, nec non obductos crustis parietes." These "crusta" were doubtless of marbles or porphyries; I have myself noticed the traces of such decoration in towers of that part of the wall which is near the Porta San Giovanni, and which probably dates from the time of Aurelian.

a

The system of decorating walls by means of incrustations of marbles was largely used in the churches constructed in and after the time of Constantine. San Paolo fuori le Mure at Rome, and probably almost every large church built in the fourth and fifth and sixth centuries, was so decorated; the most remarkable examples remaining are at Rome, Santa Sabina (A.D. 425-450); at Ravenna, San Giovanni in Fonte, the baptistery of the cathedral (fifth century), and San Vitale (sixth century); at Parenzo (sixth century) the Duomo ; at Constantinople, Santa Sophia; at Thessalonica, St. Demetrius (early sixth century). In none of these, however, as far as I am aware, is any glass used in the sectile work, although the tessellated work in the same churches is almost or entirely composed of that substance. The materials are chiefly marbles of various colours and porphyries, but brick is employed at San Giovanni in Fonte, Ravenna, and Parenzo, and mother-o'-pearl at Parenzo, and at San Vitale in Ravenna.

From the period of the above-mentioned churches to the present time this system of decoration has continued in use, so far as marble is concerned. Though it may be difficult to cite examples dating from the centuries between the sixth and the eleventh, it is probable that in the territories ruled by the Emperor of the East it was used from time to time. In the eleventh century we find it in

a Lect. Ant. lib. 26, c. 32, p. 1,480. Geneva, 1620.