heat; the mass thus obtained was then heated and worked into vessels, tablets, or the like. Many of the pieces employed in this mosaic are apparently fragments of vessels made of this millefiore glass, which was used in Rome for dishes, plates, and other table utensils, much as we now use porcelain. It was used in immense quantities, as we may learn from the fact that every year thousands of pieces are found about the sites of ancient villas. Doubtless the Transtiberines, who, as Martial tells us (Ep. lib. i. 42), made a living by exchanging sulphur matches for broken glass, sold the fragments of these vessels to the artist in mosaic, while the plain glass was returned to the melting-pot.

It

may be worth mention, as having some bearing on the question of date, that one small piece, red, with five white spots, which forms part of the chair of one of the seated figures, is identical with pieces forming part of the decorations of the villa near Rome mentioned above.

The other large picture represents a consul (or other official) clad in the “toga (or læna) picta" or "triumphalis " of purple and gold, proceeding in his chariot "of to preside at the games. The white horses are of "palombino ;" the chestnut of giallo antico;" the stockings worn by the men on horseback of " palombino ;" the garments as well of these as of the consul of glass; as also are the trappings

of the horses, with the exception of the discs in the breast and headbands of the horses attached to the biga, which are of mother-o'-pearl. The spokes of the chariot-wheels are, I believe, also of this last material, but my memoranda are deficient on this point.

ROMAN

WALL

DECORATION

IN OPUS

SECTILE.

These two mosaics are preserved in the palace of the Prince del Drago at the Quattro Fontane in Rome, which when they were originally placed there belonged to Cardinal Nerli, and has subsequently been known as the Palazzo Albani. The only other portions which remain of the decorations of this basilica are two compartments preserved in the church of San Antonio Abate, one of which is shown in the annexed cut, reproduced from a photograph. In this the background and the stripes of the tigress are of green porphyry; the remainder of the tigress's skin of giallo antico; the bullock of a pale fawn-coloured marble; the eyes, I believe, of mother-o'-pearl. No glass is to be found in these compartments. The ground appeared to be some variety of alabaster.

In these examples the work has been done in a somewhat rough and irregular manner, for they were placed originally at a considerable distance from the eye; the pieces are not made to fit accurately, and were evidently put together much as they came to hand, the junctions of pieces of the same colour occurring at irregular intervals without regard to the design and wherever convenience dictated. The glass employed is almost all opaque, scarcely any other than the deep blue being transparent.

It is interesting to consider how near an approach was made in these pictures to what we call "painted glass windows;" but, as the system of joining glass by narrow strips of lead was not invented until many centuries later, the idea of constructing transparent pictures, if it ever occurred to the mind of the artists of these mosaics, necessarily remained barren.

At the time that these mosaics were constructed, decorations of like character were probably by no means uncommon at Rome. Herr von Minutoli (Über die Anfertigung und die Nutzanwendung der farbigen Gläser bei den Alten, Berlin, 1836, p. 13) tells us that he was informed by Signor Luigi Vescovali that the walls of a chamber in a palace between the gate of San Sebastian and that of St. Paul at Rome" were found to be covered up to a man's height from the ground with choice marbles, and above that height with coloured glass plates and arabesques (mit farbigen Glastafeln und Arabesken). Signor Vescovali, he says, showed him some figures of "giallo antico" engraved and executed in a very good style; these were originally let into the wall between the glass plates, and many of the appendages, as shields, swords, the tunics or chlamydes, consisted of coloured incrusted (inkrustirten, i.e. mosaic or millefiori) glasses. Canon Gorio, a Herr v. Minutoli says that the chamber was that the floor of which consisted of the beautiful mosaic now in the museum of the capitol, in which doves perched on the lip of a vase are represented. But this is said to have been found in Hadrian's villa near Tivoli.

he says, showed him at Naples fragments of decorations of various marbles which had been affixed to the walls of two chambers with plates of coloured glass between them.

But, although so few examples of sectile wall decoration have been preserved, the practice, as has been already said, of forming such with pieces of glass must have prevailed most extensively in Rome and its vicinity, for even now, after a lapse of some fourteen or fifteen hundred years, thousands of pieces of glass which have formed parts of such decorations are annually found, chiefly in the vineyards and market gardens which surround Rome and occupy the sites of villas. These are of all sorts of colours and forms; red, white, and green are the most common colours; the forms are geometrical, octagons, triangles, squares, portions of circles, or often non-geometrical. Some of these nongeometrical forms are obviously adapted to form parts of the running patterns used as borders in antique work, but others are of shapes the adaptability of which to the formation of patterns is less obvious. I have, however, found that the most puzzling pieces had usually one straight side, and that if two such pieces were joined at these straight sides a figure was produced which might pass as a petal of a three, four, or six-petalled flower. The diagrams in the accompanying plate (Plate XX.) will show how these pieces were in all probability arranged. The coloured parts represent the pieces which exist.

The rosettes thus formed were probably inclosed in octagons, and these combined so as to form patterns of the character which we usually call diapers, repetitions of small parts in a symmetrical manner. Such patterns are not very often met with at Pompeii or in the sepulchres or other remains at Rome, the chief sources from whence we derive our knowledge of Roman wall decoration for domestic purposes. But one instance at Pompeii may be cited; it was found in a house in the Strada di Stabia discovered in 1872; an engraving will be found in Ritcher's Secrets de l'Art Decoratif, plate 1.

Having thus described in some detail the examples of sectile mosaic in wall decoration which have come under my notice, I proceed to the history and description of the very remarkable building which has afforded the only examples of this method of depicting subjects into which human or animal forms entered which have been preserved to our time.

The history of this building, from the time when it was converted into a church and dedicated by Pope Simplicius, to that of its destruction, has long been sufficiently well known, but that of the earlier period has been misunderstood.