In spite of the absence of “banquet” or “greeting” scenes, the parallelism with the Attic reliefs is very marked, and the sepulchral character of these vases is indubitable. It is, further, natural to suppose that there is some reference to the worship of a ἥρως or deceased ancestor, such as is known to have been a universal custom among the Greeks.^[1467] Reliefs have been found at Tarentum with subjects which obviously have this reference. Apart from these two classes, however, the majority of the vases of Southern Italy seem to have been made originally for ornamental purposes, such as the decoration of a house, as is implied by the distinction in the artistic merit of the two sides.

Artists’ signatures in this period are exceedingly rare; only three, in fact, are known. Of these one may be briefly dismissed—Lasimos,^[1468] who signed a fine Apulian vase in the Louvre, with sepulchral and other scenes; his style is hardly distinctive enough to admit of identifying any others as his work. But in the other two names, those of Assteas and Python, we find more interest. Five vases exist with the signature of Assteas, and one with that of Python, and it is interesting to note that they both use the form ἔγραψε (see Chapter XVII.). The list is as follows:

Assteas. (1) Krater from Paestum in Madrid. Reinach, i. 168 = Baumeister, i. p. 665, fig. 732 = Fig. 107. Herakles destroying his children.

(2) Krater from Paestum in Naples (3412). Wiener Vorl. B. 2. Phrixos and Helle.

(3) Krater from S. Agata dei Goti in Naples (3226). Millingen, Anc. Uned. Mon. i. 27. Kadmos slaying the dragon.

(4) Krater from S. Agata dei Goti in Berlin (3044). Wiener Vorl. B. 3, 1. Scene from farce (parody of Prokrustes?).

(5) Lekythos from Paestum in Naples (2873). Millin-Reinach, i. pl. 3. The garden of the Hesperides.

Python. Krater from S. Agata dei Goti in the British Museum (F 149). J.H.S. xi. pl. 6. Alkmena on the funeral pyre.

The characteristics of Assteas’ work are very marked, and, curiously enough, Python’s differs little from it. Both are essentially pictorial artists, trained in Greek traditions, and inheriting from Attic painters like Meidias the love of elaborate and minutely rendered draperies and picturesque grouping of figures at different levels. In the latter detail we also seem to see signs of the influence of Polygnotos.

There are many other vases in our museums which present the same features of style and treatment as these.^[1469] Besides those already mentioned, the fondness for half-figures in the background, the large heads, pronounced features, and heavy masses of hair in the figures on these vases connect them unmistakably with the school represented by the two artists. It is not the style of Lucania or of Campania, still less that of Apulia; and yet it is clearly an Italian fabric. Some previous writers have maintained that Assteas came from (or was resident at) Tarentum, arguing thus partly on epigraphical grounds, partly on the ground of his employment of scenes from the farces,^[1470] which, as we have seen, were popular in that city. But having regard to the fact that three out of five of Assteas’ vases were found at Paestum, and that he combines certain characteristics of Lucanian and Campanian fabrics, we may fairly assume that he (and therefore also Python) resided in that city, which lay on the border of the two districts.

We are thus enabled to establish a style of Paestum distinct from the other Italian fabrics—a conclusion at which the present writer and Signor Patroni arrived independently some years back. The latter has pointed out that several small details also point to that city—such as the gaily plumed helmet worn by Herakles on the Madrid vase, which resembles those worn by local warriors on paintings found in that city.^[1471] And in the Naples Museum there are several other vases in the style of Assteas from Paestum.^[1472] Signor Patroni dates Assteas about 350–320 B.C., Python a little later.

The Madrid vase and the Python krater are in their way masterpieces, and form almost the finest examples we possess of South Italian vase-painting. Both are extraordinarily rich in colouring as well as in detail. The former (Fig. 107) represents, as has been said, Herakles destroying his children the subject being treated in a manner which to us appears almost grotesque, not to say comic. But it is probable that this is due partly to the element of exaggeration which has been ascribed to the revival of tragedy (see p. 472). The whole conception is obviously theatrical, with the setting of Herakles and his child, the principal figures, against a background formed, after theatrical models, by the front of the palace, through openings in which appear the horrified faces of Alkmena and Iolaos, and that of Mania, the goddess of madness. Herakles has already set fire to a confused pile of household furniture—tables, chairs, and wool-baskets—and a child clings to him in agony, while Megara tears her dishevelled hair; but their pleadings have no effect. In the Python krater the action is less violent and theatrical, but there is the same gaudiness of colouring and richness of embroidered costume. Alkmena is seated on the pyre, to which Amphitryon and Antenor are about to set light, and raises her hand in supplication to Zeus, whose bust is seen above. In answer to her prayer the Hyades or rain-nymphs pour down water from their pitchers to extinguish the flames. It should be noted that in this painting we have several successive stages of time combined in one (cf. Vol. II. p. 10); the pyre is not yet lighted, but the water is already descending to extinguish it.

We now proceed to describe in detail the characteristics of the three principal fabrics, beginning with that of Lucania, as the earliest in character, if not necessarily in point of time. Lucanian vases stand nearer to the latest Attic fabrics than do those of the other districts, and do not present the same local peculiarities; nor do they sink like the others into a state of decadence and barbarism, but are very conservative in their style.

We note in them a much greater unity of style than in the vases of Campania, and everything points to one centre of fabrication. This is most probably Anzi, where the largest number have been found. Information as to provenance is unfortunately often vague, but few other places are given as sources (see p. 83), almost the only other names being those of Pisticci and Pomarico. But the number of vases that it is possible to attribute to Lucania is not large in any case.

The designs are usually somewhat severe and restrained, and characterised by a certain stiffness of drawing and largeness of scale. The heads of figures are abnormally large, with great staring eyes and masses of hair rendered without detail. The draperies are comparatively free from ornamentation, only broad black borders and patterns of small dots being admitted. The clay is of a rich red colour, but accessory colours are exceedingly rare. Hence they present a great contrast to the Apulian and Campanian, with their masses of white and generally gaudy appearance. Another peculiarity is that fillets in the hair are rendered simply by leaving a narrow band across the head in the colour of the clay. The figures often stand in the air without the usual dotted ground-lines, but sometimes the ground is represented by a heap of loose stones. A favourite device is that of a half-shield seen in the upper part of the scene, as a sort of indication of locality or action.^[1473] Fig. 108 gives a typical example of Lucanian vase-painting.

Among the favourite shapes are the bell-shaped krater and the amphora, also the hydria and column-handled krater. The hydria is generally employed, as in Campania, for sepulchral subjects. The vases are mostly of large size, whence a corresponding largeness of the figures; whereas Campanian vases are generally small, and make up for the absence of imposing figures by their colouring. An entirely new shape, peculiar to this style, is the four-handled krater, to which the name of nestoris has been somewhat absurdly given^[1474]; it is undoubtedly a local form, being found in the indigenous pottery of the district.^[1475] There are two varieties, one with a high neck, the other with sloping shoulder and no neck. The handles are usually ornamented with discs painted with rosettes, and the designs are in panels surrounded by ornament, sometimes on the second variety with a lower frieze of figures. Generally speaking, secondary ornamentation is largely employed on these vases, especially on the last-named shape. The palmette patterns under the handles are usually very luxuriant.

The vases of Campania present in many ways a striking contrast to those of Lucania. Their chief characteristic is, as has been noted, love of picturesque effect and variety of colour, even to the extent of introducing attempts at shading (see above, p. 471). The vases are mostly small, and none of the large kraters or amphorae belong to this class. The favourite shapes are the hydria, lekythos with bulbous body, and amphora; the latter is clearly an imitation of the Attic “Nolan” amphorae, which were so largely imported into the district, but the body is usually more symmetrical. The clay is usually of a buff or dull yellow ochre tone, and red and yellow washes are frequently used, as well as large masses of white; these tints are laid on very carelessly, and the white is of a kind that is apt to flake off and disappear. Yellow, purple, and white are largely used as accessories, and the drawing has a tendency to become very careless. The lines of the ground are indicated by occasional strokes of white, or by rocks strewn with flowers. Ornamental patterns are not so popular as in Lucania; the favourite is the wave, and the palmettes under the handles are thick and ugly, with angular leaves. Some decorative motives seem to be derived directly from nature.

The subjects are often interesting and uncommon, introducing recondite or unusual myths; many of the vases with comic scenes appear to belong to this class, and one in the British Museum has an Oscan inscription. Local peculiarities of costume and armour, which Signor Patroni calls Osco-Samnite, are often found; for instance, warriors wear a very short chiton with broad girdle, a helmet with waving crest and tall side-plumes of Italian type,^[1476] and a remarkable breast-plate formed of three circular plates of metal arranged in a triangle.^[1477] These same peculiarities are found on the wall-paintings at Paestum, and there are indications that Virgil was familiar with them.^[1478]

Signor Patroni, by dint of an exhaustive study of the Naples collection, has made a tentative classification of Campanian vases according to fabric; he distinguishes those of Cumae, Saticula (Santa Agata dei Goti), and Abella; but those of Capua, Nola, and Neapolis appear to have no distinctive style. The Cumae fabric, for studying which the Raccolta Cumana in Naples gives exceptional facilities, is represented by the long, straight-bodied amphorae, the hydriae with female heads under the handles, and kraters on which the design is framed by stylised floral patterns or heavy palmettes. Among the characteristic patterns are the wave, large flowers in profile, and ground-ornaments generally, such as ivy-leaves, branches, and small windows. The strong tendency to polychromy seems to be the result of using the late Attic polychrome vases as models. In the colouring a new feature is the use of a carmine red, which, according to Patroni, is only found in the Cumae fabrics.^[1479] Mythological subjects are rare,^[1480] sepulchral common, and shrines are found on these alone; but the majority have scenes from daily life,^[1481] banquets, return of warriors, etc. It is on these that the local costumes are usually found.

The Saticula fabrics are very uniform,^[1482] practically all bell-shaped kraters with red clay; colours are sparingly used, and then only white; a maeander takes the place of the wave-pattern as a border; ground-lines are usually indicated. Of subjects Dionysiac have the preference. The vases of Abella are of late date, chiefly hydriae of very pale clay with accessory colours; among the typical patterns are arabesques ending in white daisies. They sometimes show reminiscences of the Paestum style.^[1483]

There are a few peculiar fabrics which we may also attribute to a Campanian origin, including rude imitations of the B.F. style, chiefly small amphorae with single figures; imitations of Nolan amphorae, reproducing both their form and their scheme of decoration^[1484]; and bell-shaped kraters imitating the Attic style, which Signor Patroni has associated with Saticula. The imitations of Nolan amphorae have a slim body, twisted handles, and a sharply set-off shoulder forming a right angle with the neck instead of a graceful curve. As in their prototypes, the subjects are confined to one or two figures each side. The lustrous black glaze of the Attic vases is admirably reproduced. There is also a class of vases with designs painted in opaque red on the black ground, reproducing the method of the transitional vases described on p. 393.^[1485] They are very rude in character, with roughly incised details and subjects of a simple kind; the red pigment appears to have been made from fragments of pounded pottery (testa trita). There is, however, one remarkable exception—a small phiale in the British Museum,^[1486] dating from the third century, with the subject of a shepherd-boy with his dog. The design is carefully painted in opaque red and white in the style of the Pompeian wall-paintings, and the effect of light and shade produced by hatched lines is both remarkable and unique. A krater found at Civita Castellana (Falerii),^[1487] the paintings on which are in Campanian style, is unique in having Latin inscriptions over the figures, a group consisting of Zeus (... SPATER, Die]spater), Ganymede, Eros (CVPIDO), and Athena (MENERVA). The subject is conceived rather in the style of the Etruscan mirrors than that of the painted vases, and is obviously under local influence. As Falerii was destroyed in 243 B.C., a terminus ante quem may be obtained for the date of the vase, as for others found on this site (see p. 75).

The vases of Apulia are not only more numerous, but of more merit and greater interest than those of the other two classes. In them may be observed two or three stages of development, beginning with a fifth- or early fourth-century group of Attic type, consisting of large amphorae with two friezes of figures.^[1488] Both in shape and method of decoration these form the prototype of the large kraters and amphorae which comprise the second class; they are distinguished from the latter by severity of treatment and absence of colour. The second class includes the large vases with mythological and tragic subjects, the Under-world vases, and those with sepulchral scenes; they are all richly decorated from head to foot, with two main rows of figures, smaller subjects on the neck, and ornamentation over every available space. The theatrical characteristics of which we have spoken above (p. 472) are best illustrated by some of this series.

The third class includes some large vases, such as the so-called pelikae and the large phialae, and the smaller forms, the oinochoë and its varieties, and kanthari, rhyta, and other kinds of drinking-cups. Some shapes are peculiar to this class. In spite of the great variety of shape, there is a remarkable poverty of conception in the subjects, which show a tendency to become purely decorative, and are mainly confined to the vague “courting” scenes or “toilet” scenes, or to single figures of Eros and Nike. On the smallest vases the commonest subject is often that of a female head covered with a cap, sometimes of a relatively colossal size, and this also occurs, surrounded by foliage, on the necks of the large vases. The shapes, as in the case of the epichysis (p. 179), often tend to ugliness and over-refinement.

The conception of Eros on the later Apulian vases is one of their chief characteristics (cf. Plate XLIV.). An almost invariable participant in every scene, his form assumes an androgynous character; his hair is arranged in feminine fashion, and his person adorned with necklaces, earrings, and other jewellery. Among other peculiarities we may note the double line of white or yellow dots for ground-lines; the characterising of Oriental figures by tiaras and cross-belts; the general treatment of the hair of women, at first long, thick, and wig-like, but later gathered up in a cap, from which the ends float out behind; the thick but effeminate proportions of the men; and the small heads of the horses.

There does not seem to be any possibility of distinguishing different centres of fabric in Apulia. Nor can Tarentum have been a centre of vase-fabrics, although Lenormant stoutly upheld its claims, as the chief centre of Greek civilisation in that region. But Tarentum has been the scene of much excavation, and results do not point to that conclusion; most of the vases found there are purely Greek. On the other hand, enormous numbers have been found at Ruvo, and this was undoubtedly the chief centre, though without a distinguishing style of its own. Ruvo was famous for its red clay, and remains of furnaces and potteries have been found there. Other sites where vases have been found are Bari, Canosa, and Ceglie. At Canosa there was a preference for the tall amphora with scroll-handles, the large phiale, and the prochoös,^[1489] and purple accessories were largely used here. It is also interesting to recall that Canosa seems to have been the centre for the large ornamental vases of terracotta painted in tempera (p. 119).

On some of the column-handled kraters^[1490] local costumes appear, probably representing the Peucetians, and having some affinities with those of Lucania; the principal features are the tall pointed cap and short striped chiton worn by both sexes. Another group peculiar to Apulia is formed by the fish-plates^[1491]—a peculiar form of plate, with low stem, a sinking in the centre, and edge turned over, all being painted with fish of various kinds (Plate XLIV.). They were no doubt used for eating fish, the sinking being for the sauce; but they may also have been hung up as votive offerings in the temple of some marine deity.

The last efforts of vase-painting on the soil of Magna Graecia date from the latter half of the third century B.C. By this time vase-painting had reached a stage of complete decadence, devoid of style or taste, and rapidly verging on barbarism, as shown in some specimens, which seem to be the efforts of local craftsmen to copy the better examples, but with the same want of success as the Etruscans.^[1492]

Another direction which vase-painting took before it finally disappeared is illustrated by a group of vases mostly found at Egnazia (Gnathia) in Apulia, which clearly form a final stage in the evolution of the local fabric just discussed. Originally known from the place where the majority was found as vasi di Egnazia or Gnathia vases, they were in the view of Lenormant more probably made at Tarentum.^[1493] But we have seen that there is slight evidence of local fabric there,^[1494] and their connection with the fabrics of Ruvo and Canosa makes it more likely that they came from that neighbourhood. It is therefore probable that the old name is the correct one.

The characteristics of this group are: (1) the black varnish with which the whole vase is covered; (2) the designs painted in opaque colours—white, purple, and yellow; (3) the tendency to imitate vases of metal, as seen in the vertically ribbed bodies and other details of form. The important rôle played by the black varnish is interesting, as showing the increasing tendency to reduce the painter’s labour to a minimum, combined with a striving after novelty and the rejuvenation of the art. The practice, no doubt, arose from the discovery of the painter that it was easier to paint the figures on the black in opaque colour than to trace them out in the clay and work round them with the varnish, especially in the case of the elaborate foliage patterns which played so important a part in Apulian vases.

The subjects are usually confined to the shoulder or neck, at least of the larger vases; but figures are comparatively rare. One krater in the British Museum (F 543) which belongs to the comic series is a notable exception; and there is a pleasing subject on a skyphos in the Louvre^[1495]—a cock and goose confronted, and greeting one another with the respective salutations, “Ah, the goose!” “Oh, the cock!” But in the majority of cases the only designs are female heads, Erotes (Fig. 118), birds, comic and tragic masks suspended from wreaths, and simple foliage patterns. The reverse of the two-sided vases is often undecorated.

It is interesting to note that specimens of this ware are sometimes found on Greek sites, such as Athens, Myrina in Asia Minor, Melos, and Cyprus. At Curium in the latter island a fine hydria in this style, with figures on the shoulder (Fig. 109), was found in 1895.^[1496] Whether these were imported from Italy or made elsewhere is quite uncertain.^[1497]

Another interesting but much smaller class which belongs to the latter half of the third century is formed by a group of vases, mostly small phialae, which are distinguished by bearing painted Latin inscriptions.^[1498] Some also have figures (Eros, a female head, etc.), which are treated in the same manner as the Gnatia vases. It is probable that Rome was the place of origin of this class, in spite of the fact that most of them were found in Etruria.^[1499] But the Latin language at that time was more at home in Campania than anywhere else outside Rome. The inscriptions take the form: AECETIAI POCOLOM, Aequitiae poculum (B.M. F 604 = Fig. 110); IVNONENES POCOLOM, Junonis poculum; and so on,—Saturn, Mercury, and other Roman deities being included in the list. Reasons have been given for dating this series in the First Punic War, 260–240 B.C.

Formerly it was universally supposed that the art of vase-painting was brought to an end in 186 B.C. by the action of the Roman Senate when they issued their edict against Bacchanalian ceremonies, which undoubtedly affected Southern Italy. But this was only a natural view to be taken by writers who associated the painted vases with the Eleusinian mysteries and similar ideas; on other grounds it is hardly tenable. Especially in regard to the general putting back of the chronology of the art, it is impossible to suppose that painted vases with mythological subjects were still made in the second century. The character of the mid-third-century vases just described is sufficient to indicate that they represent the last stage to which Greek painting could ever have reached.

§ 3. Figure-Vases and Vases with Reliefs

We propose to conclude this sketch of the history of Greek vase-painting with a few words on a principle which, while always present in Greek pottery, yet at all times lay in the background, until the latest stages of the art, when it entered on a phase of increased popularity. This is the principle of combining the ceramic with the plastic art—in other words, the manufacture of vases in the form of human or animal figures or heads.

It has already been noted, in discussing the primitive pottery of Troy (p. 257), that the idea of associating the vase form and the human form is a very old one. At Troy it is of course seen in its most rudimentary stage, when correct modelling was a thing quite beyond the potter’s scope, and he could only roughly indicate features and limbs on the surface of the vase, which thus always remained a vase, and the figure idea never gained, as in later times, the predominance. In the Mycenaean period the advance in modelling was great, but only reached a high level in Crete. It is only since the discoveries at Knossos that we have been able to account for the astounding group of porcelain rhyta from the Enkomi tombs in Cyprus (see Plate X., fig. 4),^[1500] which at first sight seem to have been made by a sixth-century artist, so admirable and lifelike are they. Although the rams’ heads bear the palm, the female heads are, for the period, a tour de force, so advanced in type that it would be pardonable to argue—apart from the circumstances of their discovery—that they must belong to a later stage of art.

Apart from these, however, the principle did not find its way into Greece before the seventh century B.C., and then its origin is indubitably Oriental. It is best exemplified by the discoveries in Rhodes, especially at Kameiros,^[1501] where vases of porcelain and terracotta are found modelled in the form of helmeted heads or heads of animals (see Plate XLVI., fig. 1, and p. 127). The type adopted is that of the aryballos (p. 197); it was no doubt a comparatively easy matter to model its spherical body into the form required, applying paint where necessary to bring out the details as on the vases. In the Western Mediterranean the alabastron form seems to have been more popular.^[1502] It is often adopted for the Canopic vases of Etruria (see Chapter XVIII.). Many of these are unpainted, or rather are covered with a white slip and then painted in tempera like the ordinary terracotta figures; they are, in fact, figurines in essence, vases by accident; whereas in the first-named group the vase idea retains the predominance. But it is almost impossible to draw the line. A fine early instance of imitation of metal in early Greek pottery is the British Museum jug from Aegina (A 457) terminating in the head of a Gryphon.

During the sixth century painted figurine vases are rare, though there are not wanting various examples of the class just described, which belong to this period; but at all events hardly any examples can be traced to Athenian manufacture during the age of B.F. vase-painting. Towards the end of the century, however, the fashion was reintroduced by the potter Charinos, who belongs to the transitional period (about 525–500 B.C.). A vase signed by him, which was found at Corneto, is in the form of a female head surmounted by a kalathos.^[1503] It was made in a mould like the terracotta figures, but the painted decoration, which is remarkably elaborate and minute, is entirely B.F. in character. The patterns on the head-dress include maeander, stars, ivy-leaves, lozenge and net patterns, and a minute frieze of animals, painted in black on the clay ground. A similar vase, but later in date, is in the Berlin Museum^[1504]; in this example we may note the introduction of R.F. ornamentation, in the palmettes and diapering round the top.

These two stand at the head of a series of similar vases extending throughout the succeeding periods down to the end of the age of painted vases. They compare for style with the heads of the female statues found on the Acropolis, which belong to the same period. Two other potters, Kaliades and Prokles, made similar vases.^[1505]

The fashion started by Charinos continued throughout the fifth century, but the plastic conception tended to become subordinate to the ceramic, and it became more and more customary to decorate the non-plastic portions in the manner of the vases. Of this development the most noteworthy example is the beautiful rhyton in the British Museum in the form of a Sphinx (E 788), the upper or vase part of which is ornamented with the subject of Kekrops and Erichthonios. The body of the Sphinx is covered with a fine white slip, and the details are picked out with red and gilding. This vase dates from about the middle of the century. There also exist many examples of rhyta or kanthari, formed of a head or two heads back to back, usually a Maenad and a Seilenos.^[1506] Another favourite type is that of a jug in the form of a negro’s or Aethiopian’s head^[1507]; and there are also rhyta which terminate in the head of a lion, mule, or other animal finely modelled (Plate XLVI., figs. 2, 5).

Towards the end of the fifth century there is a reversion to the purely plastic figure-vase, usually in the form of a lekythos with spherical body, to the front of which the figure is attached (Plate XLVI., fig. 4). The vase is usually covered with black glaze, and the figure with a white slip like the terracottas, with polychrome colouring. Examples of this class are the series of lekythi representing Aphrodite Anadyomene in a scallop-shell, of which there are examples at Athens and Petersburg,^[1508] and the fine vase in the British Museum (E 716) with the bust of Athena Parthenos. A series of smaller lekythi, of which the British Museum possesses examples (G 2–7), represents Eros on a dolphin, the young Dionysos in a sort of canopy, Europa on the bull, a boy with a dog, and other subjects; the technique is similar to that of the larger specimens, with pink and green colouring. They form charming little objects, and are often well executed.^[1509]

In Southern Italy many of these types are continued, the most popular being that of the rhyton ending in an animal’s head (p. 193), of which many examples have been found in Apulia. They usually have some simple design painted on the upper part, such as a figure of Eros. There are also numerous examples of vases in the form of animals or human figures (Plate XLVI., fig. 3), some of which are in black glazed ware with patterns in white like the vases of Egnazia, others being covered with white slip like the terracottas. With the decay of painted decoration the plastic element gradually predominates more and more, until the vase form becomes, so to speak, purely accidental. Thus in the third century the fabrics of Canosa, of which we have spoken in a previous chapter (p. 118), entirely hold the field, and the vases pass out of the sphere of the history of vase-painting.

In all or nearly all of the vases just described we observe the same principle at work—namely, the tendency to imitate metal in terracotta. It is one that is constantly recurring throughout the whole history of Greek ceramics, with more or less persistency and prominence. Sometimes, as in the Melian, Proto-Attic, and other fabrics, the imitation is limited to the form of the handles, which is, strictly speaking, inappropriate in terracotta, though frequently found in early bronze vessels.^[1510] It is seldom found in Ionia, but in Western Greece there are many examples during the seventh and sixth centuries, as in some of the Proto-Corinthian and early Corinthian wares,^[1511] This is doubtless in a large measure due to the influence of the great centres of metal-work at that time, Corinth and Chalkis. We are not, therefore, surprised to find the tendency exemplified in the pottery fabrics of those two centres, and at Chalkis, as has already been noted (p. 321), it is especially conspicuous in the form and minor details of the vases. At Athens examples are rare, with the exception of the vases of Nikosthenes, who not only copies complete vases in metal, as in his peculiarly-shaped amphorae and in a small phiale mesomphalos in the British Museum,^[1512] but is also addicted to adorning the handles of jugs with female heads in relief, as on specimens in the Louvre and elsewhere.^[1513] After the sixth century the tendency is far less conspicuous, owing to the high esteem in which vase-painting was then held, and little is seen of attempts at imitating metal until the revival of the plastic element in pottery in the fourth century. An almost unique exception is the Berlin krater from Corinth (2882), which must date from the fifth century. It is of black ware, with designs in relief round the body.^[1514]

The tendency also manifests itself in a marked degree in another direction in early Greek art—namely, in that of ornamenting vases with reliefs. So much evidence of this has been yielded by discoveries on Greek soil that it is now certain that this method of decoration had its origin in Greece, and not in Etruria, although the close resemblance between early relief-wares from Rhodes and the large πίθοι of Cervetri (see p. 153) had led archaeologists in the past to regard Etruria as its original home. The Etruscans always preferred modelled vases or relief decoration to painted ware, as their bucchero fabrics show; but we know that they had no inventive power, and even in this they have proved to be only imitators.^[1515]

Turning to details of the early Greek vases with reliefs, we may note that there are two varieties: firstly, those in which the reliefs are made by rolling a cylinder round the vase, the design being repeated over again; secondly, those in which the reliefs are made from separate moulds, and attached with some kind of cement.^[1516] In both classes the shape usually affected is that of a large πίθος (cf. p. 151), of a somewhat coarse red clay. It is the first variety which so closely resemble the πίθοι found at Cervetri, and which are now known to be the prototypes, not imitations, of the Etruscan examples.^[1517]