2

So far the paintings which we have considered can only be regarded as an extremely ingenious and, in the main, tasteful. form of wall-decoration; they tell us little of that which we most wish to know-the style and treatment of substantive works of painting. The gap is in some measure filled by the central panels of Pompeian walls, which are usually adorned with subject-paintings, often mythological in subject, clearly marked off from the rest of the wall and intended to take the place of pictures. In the Architectural style these are usually framed in a species of pavilion or aedicula, painted in perspective; but this motive gradually loses its importance. In the Third style ("ornate") distinguished by Mau the architectural design ceases to be intelligible as the counterfeit of real construction, and becomes a purely conventional scheme of decoration; and in the Fourth or Intricate style, which again reverts to true architectural forms, however fantastic and bewildering in their complexity, the figure-subjects are plainly conceived as pictures and framed with a simple band of colour. The subjects of these frescoes are for the most part taken from Greek mythology, and it has been argued that in the main we have to deal with reproductions of Hellenistic paintings rather than of contemporary works of art. It is not to be denied that the motives of famous compositions of earlier date may have found their way into the repertory of the Pompeian artists; it is not unnatural, for example, to conjecture that the figure of Medea here reproduced (Plate VI. fig. 30) may have been inspired by the celebrated painting of Timomachus above-mentioned. But there are reasons for thinking that the debt owed by the Pompeian artists to the Greek schools of the Hellenistic age is not so direct as was believed by Helbig, whose Untersuchungen über die kampanische Wandmalerei won a general acceptance for the theory. It seems clear that in the central subjects of walls decorated in the Architectural style we are intended to see, not a picture in the strict sense, but a view of the outside landscape, generally with a small shrine or cult-statue as the centre of the piece; and the importance of the figure-subject was therefore at first subordinate. These subjects are, it is true, taken from Greek mythology, but this only proves that that source of inspiration was as freely drawn upon in the art as in the literature of imperial Rome. In the later styles figure-subjects without landscape are extremely common, but it has been shown that, e.g. in the triclinium of the Casa dei Vettii, which is decorated with a cycle of mythological paintings, the lighting is carefully calculated with a view to illusionistic effect under the local conditions, so that the conception of an outlook into external space is not given up. We sometimes, as in one of the rooms in the Farnesina" house, find framed pictures directly imitated, and here the models were clearly of a relatively early period; but this is exceptional. The Pompeian paintings, therefore, may fairly be used as evidence for the methods and aims of art in imperial Rome; and when allowance is made for their decorative character and hasty execution, we must admit that they give token of considerable technical skill-the modelling of figures is often excellent, the colourscale rich, the "values" nicely calculated. The composition The composition of subject-pictures is somewhat theatrical. Amongst the wallpaintings which have been preserved are some which from their classicistic style have been thought to represent Greek originals; the most famous is the Aldobrandini Marriage (Plate V. fig. 28), now in the Vatican library. As a matter 1 The Elder Pliny's Chapters on the History of Art, p. 238.

66

2 The most striking example is that from the "House of Livia " on the Palatine.

of fact, the composition is formed by the juxtaposition of sculpturesque types, after a fashion familiar to Roman wallpainters. Mention may here be made of the combination of ornamental work in plaster with painting which is found at Pompeii, in the work of the Flavian period at Rome, and in tombs of the 2nd century A.D. In the Augustan period we find exquisitely modelled relief-work in plaster, used to ornament vaulted surfaces in the "Farnesina house; it might seem natural to treat of these under the heading of Sculpture, but in point of fact they are translations from painting into stucco. At a later time both painter and modeller worked in conjunction, with admirable effect; the results are best seen in the tombs on the Latin Way.

Little can be said as to Roman portrait-painting. We know that in this branch of art the technique generally used was that called "encaustic." The colours were mixed with liquefied wax and fixed by heat; whether they were applied in a molten state or not has been disputed, but it seems more likely that the pigments were laid on cold, and a hot instrument used afterwards. Several examples of such wax-paintings have been found in Egypt, where it was the custom during the 2nd and 3rd centuries A.D. to substitute panel portraits for the plastic masks with which mummy-cases were adorned; but these cannot be described as works of high art, though they sometimes have realistic merit. A good example in the Berlin Museum (Antike Denkmäler, ii. pl. 13) is executed in tempera on primed canvas. The medium used in ancient as in medieval tempera painting appears from the statements of ancient writers to have been yolk of egg mixed with fig-sap or natural gums.

To the little we know of purely Roman painting something is added by that which we learn from the remains of the sister art of mosaic, which, being less easily destroyed, have survived. in large numbers to the present day. It has been estimated by Gauckler that considerably more than 2000 mosaics with figure-subjects have been discovered; and the number is steadily increasing. For the origin of the art reference may be made to the article MOSAIC, where the reader will also find an explanation of the essential differences of principle between the arts of painting and mosaic. It is to the credit of the Roman artists that they were, generally speaking, alive to this distinction of method, and did not seek to produce the impression of painting executed with a liquid medium by the use of solid materials. Indeed, it seems not improbable that in this respect they had a truer conception of the function of mosaic decoration than their Greek forerunners. Amongst the mosaics of Roman date which employ a large number of exceedingly minute cubes in order to produce an illusion akin to that of painting, the most conspicuous examples are the pavement in the Lateran Museum signed by the Greek Heraclitus, which appears to reproduce the unswept hall" of Sosos of Pergamum (see MOSAIC), and the Mosaic of the Doves from Hadrian's Villa, preserved in the Capitoline Museum, which may be supposed to have been inspired by the "drinking dove" of the same artist. The former of these contains about 120, the latter as many as 160 cubes to the square inch.

As shown in the article MOSAIC, a distinction must be drawn between opus tessellatum, consisting of cubes regularly disposed in geometrical patterns, and opus vermiculatum, in which a picture is produced by means of cubes irregularly placed. The two methods were commonly used in conjunction by the Romans, who recognized that a pavement should emphasize the form of the room to which it belonged by means of a geometrical border, while figure-subjects should be reserved for the central space. A good example is furnished by a mosaic pavement discovered on the Aventine in 1858, and preserved in the Museo delle Terme (Plate VI. fig. 29). Enclosed within a geometrical framework of guilloches and scroll-work, diversified with stilllife subjects and scenic masks which break its monotony, we find a landscape evidently taken from the banks of the Nile, as the hippopotamus and crocodile, as well as the papyrus and lotus, clearly show. These Egyptian scenes are likewise found

at Pompeii, and the celebrated pavement at Palestrina, with a bird's-eye view of the Nile and its surroundings, is the finest, as well as the latest, example of the class. The conclusion to be drawn is that the Roman mosaic-workers of the early Empire owed much to Alexandrian models. Their finer works, however, were restricted in size, and formed small pictures isolated in geometrical pavements. Such mosaic-pictures were called emblemata, and were often transported from the great centres of production to distant provinces, where pavements were prepared for their reception. The subjects of these emblemata, like those of the wall-paintings of Pompeii, were, for the most part, taken from Greek mythology, and it is not easy to determine what degree of originality is to be assigned to Roman artists. We note a certain interest in the great figures of literature and philosophy. A subject of which two somewhat different versions have been preserved, commonly known as "The Academy of Plato," shows us a group of Greek philosophers engaged in discussion. In provincial pavements it is not uncommon to find portraits of poets or philosophers used to fill ornamental schemes of decoration, as in the famous mosaic at Trier signed by Monnus. And it is possible to trace the growth of interest in Roman literature at the expense of that of Greece. Fig. 31 (Plate VI.) shows a mosaic discovered in the tablinum of a villa at Sousse (Susa) in Tunis (the ancient Hadrumetum). It represents the poet Virgil seated, with a scroll on his knee, upon which is written Aen. i. 8; beside him stand the muses of tragedy and history. In one of the side-wings (alae) of the atrium was a mosaic representing the parting of Aeneas from Dido, and this was no doubt balanced by another scene from the Aeneid. It has also been shown that the mythological scenes depicted by the mosaic-workers of the later imperial period are frequently inspired, not by Greek poetry or even Greek artistic tradition, but by the works of Ovid; and the popularity of the legend of Cupid and Psyche is doubtless to be traced to its literary treatment by Apuleius.

The mosaic shown in fig. 31 is notable for the simplicity of its composition; and it may be laid down as a general rule that the later workers in this field preferred such subjects, consisting of few figures on a neutral background, which lend themselves to broad treatment, and are best suited to the genius of mosaic. The finer pavements discovered in the villas of the landed proprietors of the African provinces, Gaul, and even Britain, are distinguished by the excellent taste with which ornament and subject are adapted to the space at the disposal of the artist. Beside a well-chosen repertory of geometrical patterns, the mosaic-workers make use of vegetable motives taken from the vine, the olive, the acanthus or the ivy, as well as conventional figures, such as the seasons,1 the winds, the months and allegorical figures of all kinds, forming elements in a scheme of decoration which, though often of great richness, is never lacking in symmetry and sobriety.

It is much to be regretted that the destruction, partial or complete, of the great thermae and palaces of the early Empire has deprived us of the means of passing judgment on the opus musivum proper (see MOSAIC), i.e. the decoration of vaults and wall-surfaces with mosaics in glass, enamel or precious materials. Effective as are the pavements constructed with tesserae of marble or coloured stone, they must have been eclipsed by the brilliant hues of the wall-mosaics. We can form but little idea of these from the decoration of fountains at Pompeii and elsewhere, and must depend chiefly on the compositions which adorn the walls and apses of early Christian basilicas. An attempt has, indeed, been made to prove that one of these the church of S. Maria Maggiore-is nothing else than a private basilica once belonging to a Roman palace, and that its mosaics date from the period of Septimius Severus ; 2 but it is impossible to accept this theory. The earliest monument of the class which we are now considering is the baptistery of S. Costanza at Rome, built by Constantine in the early years 1 At least fifty examples of these have been found.

of the 4th century A.D. Unfortunately the mosaics of the cupola were destroyed in the 16th century, and we derive our knowledge of them from drawings made by Francesco d'Olanda. The tambour was decorated with a maritime landscape diversified with islands and filled with a crowd of putti fishing; and the cupola itself was divided into twelve compartments, containing figure-subjects, by acanthus motives and caryatids. The mosaics of the annular vault which surrounds the baptistery are extant, though much restored, and purely pagan in design, showing that the decorative schemes (Eros and Psyche, vinepatterns, medallions, &c.), commonly found in pavements were also used by the musivarii. The mosaic-panels of the nave of S. Maria Maggiore already mentioned are (in the absence of earlier examples) very instructive as to the artistic quality of Roman opus musivum. Richter and Taylor's publication of some of the unrestored portions, which unfortunately form but a small fraction of the whole, serve to show that the musivarii had an accurate conception of the true function of mosaic destined to be seen at a distance. Their effects are produced by a bold use of simple means; a few large cubes of irregular shape serve to give just the broad impression of a human face or figure which suits the monumental surroundings and subdued light. Very remarkable is the success with which the atmospheric backgrounds are treated. To seek delicate gradations of tint by elaborate means would be waste of labour for the mosaic-worker, but the artists of S. Maria Maggiore are able to produce sky and cloud effects (cf. Plate V. fig. 25) of great beauty, when seen from the floor of the church, with the aid of broad masses of colour. Their gamut of tones is of the richest; and it is to be remarked that no gold is used except in the restored parts. Doubtless gold was employed in decorative wall-mosaics before the Constantinian period; but the Roman musivarius knew the secret of making a true mosaic picture with natural tints alone.

(4) Work in Precious Metals.-In the article PLATE the history of this branch of art in ancient times is treated, and it is there shown that it continued to be a living art, capable of producing works of the highest merit, in Roman times. The sections of Pliny's Natural History (xxxiii. 154 sqq.) which treat of caelatura deal only with the works of Greek artists, and Pliny ends with the statement that, as silver-chasing was in his time a lost art, specimens of embossed plate were valued according to their antiquity; but the extant remains of Roman plate suffice to disprove his statement, and in a previous passage (xxxiii. 139) he names the principal ateliers where such works were produced. The famous treasure of Bosco Reale (see PLATE) comprises specimens of silver-work belonging to various dates, many of which bear the inscription "Maximae"; this doubtless gives the name of the owner of the objects, whose skeleton was found near the treasure. But some of them had passed through other hands; for example, four "salt-cellars," probably of pre-Roman date, are also inscribed with the name of "Pamphilus, the freedman of Caesar." Certain pieces, too, seem older and more worn than others; two ewers, decorated with Victories sacrificing to Athena, are probably of Alexandrian origin-the lotus-flower on their handles most probably points to their Egyptian provenance. On the other hand, the various decorative styles characteristic of Augustan art are well represented, not merely the elaborate and conventional plantsystems of the Ara Pacis Augustae, teeming with animal life, which adorn two splendid canthari, but also the naturalistic treatment of vegetable forms, of which a cup decorated with sprays of olive furnishes a good example (Plate VII. fig. 32). But the most important pieces in the collection are those which show the silversmith at work on specifically Roman subjects. Amongst the cups with emblemata (for the meaning of the term see PLATE) were two which originally contained small portraitbusts of the master and mistress of the house to which the collection belonged. One of these became detached, and is now in the British Museum; the other is in the Louvre in its

empresses of the later Julio-Claudian period; but this is not

conclusive as to date, and the style of the male portrait (which recalls the realistic bronze busts found at Pompeii) points rather to an early Flavian date. Amongst the finest pieces of this collection is a large bowl with an emblema in high relief (Plate VII. fig. 35), which was at first taken to represent the city of Alexandria, on account of the sistrum which appears amongst the attributes of the figure. It seems, however, to be a personification of the province of Africa, which was conventionally represented with a headdress formed by an elephant's scalp with trunk and tusks. We have in this emblema the earliest example of the ideal types which the Roman artists of the Empire called into being to symbolize the subject-countries; the inexhaustible fertility of the African soil is indicated by the cornucopiae and the fruits carried in the bosom of the figure. But there is some trace of that overcharging of symbolism to which we drew attention in discussing the Prima Porta statue of Augustus; and, though the bowl was in a very fine state of preservation, there is little doubt that this was due to the care with which it had been kept-it was of course an ornament reserved for the table or sideboard-and that we should date it to the Augustan period. The same is clearly true of the most important pieces comprised in the treasurethe pair of cups reserved by Baron Edmond de Rothschild and forming part of his collection (Plate VII. figs. 33 and 34). In these we have examples of the crustae, or plaques decorated in repoussé, which were mounted on smooth silver cups. The manufacture of these or at least the designing thereof-was a special branch of caelatura, and Pliny mentions an artist named Teucer who achieved distinction therein; we may possibly identify him with the gem-engraver whose signature is read on an amethyst at Florence. Upon one of these (Plate VII. fig. 34), we see a seated figure of Augustus, approached by a processional group on both sides. To the left are three divinities, the foremost of whom presents a statuette of Victory to the emperor; to the right is Mars in full panoply, in whose train follow the conquered provinces, symbolized by female figures, amongst whom we recognize Africa with her elephant headgear (see above). On the other face of the cup we see Augustus again seated, receiving the homage of a group of barbarians ushered into his presence by a Roman commander. The schemes which are here found for the first time, became typical in Roman historical art, and thence passed into the service of Christianity to portray the homage of the Magi. The second cup celebrates the glories of Tiberius, whose triumphal procession appears on the one face, and a finely conceived scene of sacrifice on the other. For the occasion various dates have been suggested (13-12 or 8-7 B.C.); but it seems most likely that the return of Tiberius from Dalmatia in A.D. 9 is here commemorated.

The fortunate preservation of the Bosco Reale treasure has enabled us to appraise Roman silverwork at its true value. It also affords some confirmation of the rapid decadence of the art, which Pliny laments. Amongst the cups are two decorated with still-life subjects and signed by an artist who writes a Roman name (Sabinus) in Greek characters, which clearly belong to the last years of Pompeii, and are coarser in execution than the earlier pieces. And the simple emblemata of the classical period, which stand out against the background of the bowl in which they are framed, give place to such a crowded group as we find on a gold patera1 found at Rennes and preserved in the Cabinet des Médailles, where the artist has surrounded the central emblema with a frieze which detracts from its effect. This and still later specimens of Roman silversmiths' work are described in the article PLATE.

(5) Gem-Engraving and Minor Arts.-The art of the gemengraver, like that of the silversmith, was naturally held in high esteem by the wealthy Romans both of the Republic and

1 Works of pure gold have but rarely survived to modern times; but traces of gilding remain upon many of the specimens of plate described above. In the law books we have mention of cups adorned with golden crustae.

Empire; 2 and the period of its highest excellence coincides almost precisely with that which gave birth to the masterpieces of Roman silver-chasing. By far the greater part of the ancient gems which exist in modern collections belong to the Roman period; and the great popularity of gem-engraving amongst the Romans is shown by the enormous number of imitative works cast in coloured glass paste, which reproduce the subjects represented in more precious materials. Not only were intagli thus produced to suit the popular demand, but fine cameos were at times cut (not cast) in coloured glass; the most notable example of these is a portrait of Tiberius in turquoise-coloured glass bearing the signature of Herophilus (see below).

In the style of Roman intagli we can trace each of the phases through which Roman plastic art has been shown to pass.3 A black agate in the Hague Museum (Furtwängler, pl. xlvii. 13) supplies a characteristic portrait of the Ciceronian age; the splendid cornelian of the Tyszkiewicz collection (Furtwängler, pl. 1. 19) with the signature IопА· ААВАN. which portrays Augustus in the guise of Poseidon in a chariot drawn by four hippocamps, is doubtless (as Furtwängler showed) to be referred to the victory of Actium; the classicism of the early Empire is exemplified by a sardonyx in Florence (Furtwängler, pl. lix. 11), which probably displays an empress of the Julio-Claudian line with the attributes of Hera; a sardonyx in the hermitage at St Petersburg (Furtwängler, pl. lviii. 1) is noteworthy because the subject is borrowed from painting and occurs on a Pompeian fresco discovered in 1897; the portraiture of the Flavian epoch is seen at its best in the aquamarine of the Cabinet des Médailles signed by Euhodos, which represents Julia, the daughter of Titus (Furtwängler, pl. xlviii. 8). Amongst later gems one of the finest is the "Hunt of Commodus" in the Cabinet des Médailles (Furtwängler, pl. 1. 41), which is engraved in one of the stones most popular with the Roman artists-the " Nicolo," a sardonyx with a bluish-grey upper layer used as background and a dark brown under layer in which the design is cut.

But the masterpieces of Roman gem-cutting are to be found in the great cameos, the finest of which no doubt belonged to the treasures of the imperial house. These were engraved in various materials, including single coloured stones such as amethyst or chalcedony; but the stone most fitted by nature for this branch of art was the sardonyx in its two chief varieties -the Indian, distinguished by the warmth and lustre of its tones, and the Arabian, with a more subdued scale of colour. As examples of these we shall take the two master-works of the art-the Grand camée de France (Plate VII. fig. 37), and the "Gemma Augustea" (Plate VII. fig. 36), preserved in the imperial collection at Vienna. The latter is attributed by Furtwängler to Dioscorides, the artist who, as Pliny tells us, enjoyed the exclusive privilege of portraying the features of Augustus. We possess several gems inscribed with his name, as well as with those of his sons and pupils-Eutyches, Herophilus (see above) and Hyllos; and, though several of these are Renaissance forgeries, enough genuine material exists for an appreciation of his style. The Arabian sardonyx was amongst his favourite stones, and the Vienna cameo at least represents the work of his school. Blending the real with the ideal, the artist has represented in the upper zone Augustus and Rome enthroned. Behind them is a group of divine figures-the inhabited Earth, Time and Tellus, according to the most probable interpretation; to the left we see Tiberius. descending from a chariot driven by Victory, before which stands a youth, probably Germanicus. We seem to have here, as in the Bosco Reale cup, a scene from the triumphal

2 We first hear of collections of gems in the last century of the Republic. Pompey dedicated that which had belonged to Mithridates the Great on the Capitol; Julius Caesar placed six collec tions in the temple of Venus Genitrix; and Marcellus dedicated another in the temple of Apollo on the Palatine.

3 The references given in the text are to Furtwängler's great work, Die antiken Gemmen, in which all ancient gems of any considerable importance are reproduced.