The process of moulding gave scope for reducing the “walls” of the figure to the smallest possible thickness, thereby avoiding the danger of shrinkage in the baking; it also rendered them extremely light, and allowed of great accuracy in detail. A model (πρότυπος) was made in terracotta with modelling-tools, from which the mould (τύπος) was taken, also in terracotta,

usually in two pieces, which were then baked to a considerable hardness. From this mould the figure was made by smearing it with layers of clay until a sufficient thickness was reached, leaving the figure hollow. The back was made separately, either from a mould or by hand, and then fitted carefully on to the front, the join being concealed by a layer of wet clay. The base was usually left open, and a vent-hole was left at the back which may have served a double purpose—first to allow the clay to contract without cracking, and subsequently in some cases for the suspension of the completed figure.

The heads and arms were usually moulded separately and attached afterwards, and altogether the average number of moulds employed—say for a Tanagra figure—was four or five. M. Pottier^[405] quotes an instance of an Eros from Myrina which is made up of no less than fourteen; yet it is not a specially complicated figure.

Greek moulds, either for statuettes or reliefs, are somewhat rare; but the British Museum contains a fair number from Tarentum of all kinds (see Plate V.).^[406] Those that we possess are mostly for small objects, such as figures of animals; but in the Museum collection there are several moulds for reliefs, as well as for vases of the later class with reliefs (see Chapters XI., XXII.), such as the Calenian phialae with embossed designs.^[407] Moulds employed for making stamps of various kinds are also in existence; at Naukratis Mr. Petrie found several circular “cake-stamps” with various designs. Of the moulds used by forgers or others for copying coins we have already spoken (p. 106).^[408]

The shrinkage of the clay as it dried afterwards permitted the figure to be withdrawn easily from the mould, and it was then ready for the necessary retouching. It is obvious from a glance at any collection of terracottas that there is a great similarity between the various representatives of any one type, and that actual or virtual repetitions are by no means uncommon. This was, of course, due to the fact that only a limited number of moulds were used, corresponding to the different types. At the same time there are in almost all cases minute differences which redeem them from a charge of monotony, and these were obtained in various ways: by varying the pose of the head or attaching the arms in different positions; by retouching before the baking; or by the addition of attributes and colouring. As it has been neatly put by M. Pottier,^[409] “All the Tanagra figures are sisters, but few of them are twins.” But retouching is not invariable, and is, in fact, confined to the finer specimens, such as those of Tanagra. In the statuettes from the Cyrenaica and Southern Italy it is the exception. The difference which it effected may be well observed by comparing two statuettes of Eros in the British Museum from Myrina (C 535–36), which are from the same mould. They are identical in style and type, yet one is far superior to the other in artistic merit, just because of the greater finish of detail.

The process of baking required great care and attention; for if no allowance was made for the evaporation of moisture, or if too great a degree of temperature was reached, the result was bound to be disastrous. It does not appear that a very high temperature was reached, especially as compared with the pottery. The clay was further insured against too rapid drying by preliminary exposure to the air. A story told by Plutarch^[410] of the fate which befell the chariot cast for the temple of Jupiter on the Capitol illustrates the possibility of disasters either from accident or carelessness. The clay swelled up to such a size and hardness that it could only be extracted by pulling the kiln to pieces.

The colouring of statuettes may be considered a fairly universal practice, although not always suggested by their present appearance. The earlier archaic specimens were not always, or only roughly, coloured, and those of the Roman period seem to have been often left plain; but otherwise it is the general rule. The surface on which the colours were applied was formed by a white slip or engobe of a creamy colour and consistency, with which the whole figure (except the back) was coated. This when dry becomes very flaky, and is liable to drop off, carrying the colours with it; most statuettes retain at least traces of this coating.

The method of painting is that known as in tempera, the pigments being opaque, mixed with some stiffening medium. The colouring was as a rule conventional, aiming at giving the figure a pleasing appearance, without any particular regard to nature. It was applied after the firing, as in that process the colours would have been liable to injury. The tints are what are known as body-colours, without any attempts at shading, and those usually employed are red, blue, yellow, and black, the white slip forming a ground throughout, and left untouched over the nude parts and often over the drapery; of these the favourites, especially for drapery, were blue and red, as also we learn from Lucian.^[411] Pollux says it was a speciality of the κοροπλάθοι to colour their figures yellow, or with a golden tint.^[412] The reds range in shade from scarlet to rose-colour and purple. At all times there was a tendency to treat the drapery in masses of colour, and this we see especially in the Tanagra figures, in which the chiton is almost invariably blue, the himation rose-pink. At a later date it became more customary to leave the drapery white, with borders and stripes only of colour. Black was only used for details of features, such as the eyes; green is very rare; and yellow was employed (in a deep brownish shade) for the hair, and also for jewellery, etc. The use of gilding is at all times rare in the statuettes; but some good examples are known—as, for instance, two archaic statuettes from the Polledrara tomb, and a head of Zeus, all in the British Museum.^[413] Imitation jewellery in terracotta gilt is not at all uncommon. On many of the earlier figures from Cyprus the drapery is indicated by stripes of red and yellow laid directly on the clay, while animals are usually decorated with stripes of red and black; the method employed is the same as on the contemporary vases (p. 253). Similarly, in the terracottas of the Mycenaean and Geometrical periods, such as those from Boeotia, the technique of the painted vases is closely followed, and the same decorative patterns are employed.

The use of an enamelled glaze first appears at Athens in the fourth century, and it is also occasionally found at Tanagra. The colour is uniformly a dull ashen-grey. A few examples are also known from the Cyrenaica, but it was in Sicily that the practice found most favour. There we find attempts to reproduce the colouring of the flesh by an enamel coating varying in hue from rose-pink to orange, and also grey and purple tints.

It is probable that the colours employed for painting terracottas were made from the same earths, though of a coarser kind, as the ware itself. Some information on the subject may be derived from Theophrastos, Vitruvius, and Dioskorides.^[414] For white the artist used a white earth, such as Melos produces, and white lead; it is also said to have been produced from the burnt lees of wine, and from ivory. The reds were composed of a red earth, probably ochre from Sinope, and vermilion or minium. Yellow was obtained from Skyros and Lydia; and a yellow ochre was obtained by burning a red earth.^[415] The Egyptian smalto or cobalt served for blue, and a copper solution prepared with alkali and silica was also employed. Copper green was obtained from many places, and mixed with white or black.

This may be a convenient point at which to speak of a class of vases which come rather under the heading of terracottas than that of painted pottery. They are found at Calvi, Canosa, Cumae, and other places in Southern Italy, and belong to the Hellenistic period, forming a parallel development to the glazed wares with reliefs of which we shall speak later (p. 497 ff.). They combine in a marked degree the characteristics of the vase and the statuette, some being vases with moulded reliefs or small figures in the round attached in different places, others again actual figures or colossal heads modelled in vase form by the addition of mouth, handle, and base (see Plate VI.). They are usually of considerable—sometimes gigantic—size, and do not appear to have served any practical purpose; some, indeed, are only imitation vases with false bottoms. It is reasonable to suppose that they were manufactured for sepulchral purposes only, like the large painted kraters and amphorae of Apulia (p. 476).

Like the statuettes, they are covered throughout with a white slip laid directly on the unglazed clay, and this is often richly coloured in tempera. Some of the heads have the hair covered with intersecting pink lines to imitate a net, and the figures attached to them are usually coloured in the manner of the statuettes, with blue and pink draperies. There are some, however, in which the encaustic or a similar process seems to have been employed^[416]; one example, in the British Museum (D 185, shown on Plate VI.), has a Hippocamp painted on either side in white and colours outlined with black, the wings being elaborately rendered in blue, brown, yellow, and pink. The same process is employed for a large cover of a vase in the British Museum from Sicily (D 1), but the figures are now nearly obliterated.

The prevailing shape of these vases is that conventionally known as the askos, with spherical body, over which passes a flat handle and three mouths on the top; the latter are often covered in and figures placed upon them. On the front and back of these vases appliqué masks of Medusa or figures in relief are usually placed, flanked by the fore-parts of galloping horses. Others take the form of a large jug or bowl with appliqué ornaments.

It now remains to consider the small but interesting class of terracotta reliefs, which are nearly all of the late archaic period, dating from the beginning of the fifth century. Later reliefs are nearly all architectural in character, and have already been described, as have those which were made for the decoration of tombs and sarcophagi. But the purpose for which the reliefs were made, of which we are about to speak, is not so certain. One group appears from the character of the subjects to be votive, and they may possibly have been let into the walls of temples or shrines; but the others are mostly known to have been found in tombs. The former group are found at Athens and at Locri in Southern Italy; the latter at Melos and other sites round the Aegean Sea, being usually known as “Melian” reliefs.

The character of the work of these Melian reliefs (see Plate VII.) is exceedingly delicate and refined; the subjects are mainly mythological, and include the slaying of Medusa by Perseus and of the Chimaera by Bellerophon, Helle on the ram, Peleus seizing Thetis, Eos carrying off Kephalos, and the death of Aktaeon. Three classes have been distinguished,^[417] of which the peculiarly Melian type has the figures cut out, without background; in the second only the outer contours are cut round, and the third consists of rectangular plaques.

Brunn^[418] considers that they served a definite architectural purpose, being intended to cover a field enclosed by borders, and that the holes with which they are pierced show that they were used either for suspension or attachment. But his reasons for regarding them as an archaistic survival have not been generally accepted.

The Locrian type of relief takes the form of a square plaque.^[419] They are easily recognised by the rough micaceous character of the clay, and by their subjects, which mostly relate to the myth and cult of Persephone. They were probably dedicated in one of her shrines, as were those found on the Acropolis at Athens to Athena. All these reliefs seem to have been impressed in moulds, not modelled by hand, as many of them exist in duplicate. Those from Greece are sometimes coloured.

Many little figures in the shape of animals and other objects, such as goats, pigs, pigeons, tortoises, chariots or boats, boys or apes riding on animals, women making bread, and similar subjects, together with jointed dolls or νευρόσπαστα, were evidently used as children’s toys. They have been found deposited with the bodies of children in the tombs of Melos, Rhodes, and Athens. In Mr. Biliotti’s excavations at Kameiros in Rhodes in 1863, one child’s tomb was found containing two of the “Melian” reliefs, small vases of glass and black-glazed ware, a terracotta basket of fruit, and a sea-shell; in another were a bird, two dolls, a child in a cradle, two grotesque figures, a woman playing a tambourine, and two other terracotta figures.

The terracotta dolls were cast in a mould like the ordinary figures, but the bodies, legs, and arms are formed of separate pieces pierced with holes, so that they might be joined and moved with strings, like the modern marionettes; hence their name of νευρόσπαστα, “drawn by wires.” They all represent girls, and sometimes dancers with castanets in their hands; they are coloured in the usual manner, and date from various periods between 500 and 200 B.C. Allusion is sometimes made to these figures in the Greek writers—as, for instance, by Xenophon, who in his Symposium^[420] introduces Socrates inquiring of an exhibitor of these puppets what he chiefly relies on in the world. “A great number of fools,” he replies, “for such are those who support me by the pleasure they take in my performances.” Aristotle^[421] mentions dolls that moved their limbs and winked their eyes like marionettes, but this can hardly refer to terracotta figures.^[422]

It would require too much space to enumerate all the subjects represented in the terracotta statuettes. But it may be found convenient to give an outline of the subjects and principal types adopted at different periods.^[423] Roughly speaking, the range of subjects may be divided into seven groups: (1) figures of deities; (2) mythological subjects; (3) scenes from daily life; (4) imitations of works of art; (5) caricatures; (6) masks; (7) animals. Among the figures of the Olympian deities we find most commonly Demeter, Aphrodite, and Artemis; Hephaistos, Ares, and Hestia are seldom if ever represented; Zeus, Poseidon, Hera, and even Athena are also very rare. Of the inferior deities, Dionysos, Persephone, Eros, and Nike (Victory) are most frequently found, as well as Satyrs and similar personages. Nor is it always easy to ascertain definitely whether a figure is or is not intended to be mythological in significance.

This question is, in fact, closely bound up with that of the Uses for which the statuettes were made, as on such a purpose their interpretation in a mythological or human sense may largely depend. The uncertainty of identification arises from the practice which obtained of adhering closely to certain recognised types, which occur repeatedly at all periods. There is a strong probability that a clear distinction was not recognised by the Greek κοροπλάσται, but that the same type of figure might be used either for a votive offering to a deity, or as a mere ornament or article of tomb-furniture. And we are further met with the fact that a type which was mythological at one period ceases to be so at another, or at any rate is transformed by some slight alteration of details or omission of an attribute. Thus the seated figure of an Earth-goddess or Nursing-mother of a Rhodian or Cypriote tomb becomes the nurse and child of the fourth century at Tanagra, while the archaic standing type of a Persephone holding a flower requires little but the omission of her special head-dress to transform her into the girl-type of the Hellenistic age.

The earliest beginnings of the statuette proper show, as might be expected in primitive Greek art, a very limited range of ideas. As in marble, bronze, and wood, so also in clay, the type of the female deity reigns supreme. The primitive Hellenic type of goddess adopts two forms, both derived from an original in wood, the board-form or σανίς, and the column-form (κίων or ξόανον), each of which finds parallels in sculpture. The limbs are either completely wanting or of the most rudimentary description, the figure terminating below in a spreading base. Both these types are found in Rhodes, but on the mainland of Greece the columnar form is confined to the Mycenaean period. In the succeeding “Geometrical” age the board-like types rose into popularity at Athens and Tegea, and above all in Boeotia. Two varieties are found, a standing and a sitting type, and they are usually painted in the manner of the local vases (see p. 290). The later examples show a great advance in modelling, especially in the heads. The columnar form exhibits its development best in the terracottas of the Graeco-Phoenician period from Cyprus.

The standing and sitting goddess (Plate VIII.) are the two principal types in archaic Greek art, and are remarkable for their wide distribution and universal popularity. The name of the goddess may vary with the locality, but the types remain almost identical, and the attributes show little variation.

Another interesting archaic type is the so-called funeral mask or bust (Plate VIII.), of which the best examples have come from Rhodes. Being almost exclusively feminine, we must suppose that they ceased to represent the image of the dead person, as in Egypt and primitive Greece, and became images of the Chthonian goddess, Demeter or Persephone, represented under the form of a bust rising out of the earth.^[424] Thus they played in the tombs the rôle of protection against evil influences, like the mask of Demeter Kidaria, worn by the priest at Pheneus in Arcadia on certain occasions.^[425] Male masks are occasionally found, representing the Chthonian Dionysos. They are very rare after the fifth century.

The purely divine and mythological types in the archaic period are very few in number. Of the Olympian deities few are represented, except in the conventional hieratic types, hardly to be differentiated one from another. But on certain sites are found representations of nature-goddesses, such as the Earth-mother with a child in her lap (Gaia Kourotrophos), or a nude goddess within a shrine, who may be a combination of Astarte and Aphrodite. These types are of Oriental origin, and are found in Cyprus, Rhodes, Naukratis, and Sardinia. They may represent offerings made after child-birth. Among the individualised deities we may point to figures of Hermes Kriophoros (from Rhodes and Sicily),^[426] of Herakles,^[427] or of the local nymph Kyrene, who appears holding the silphium-plant in a terracotta from Carthage.^[428]

Among miscellaneous feminine types are the hydrophoros or water-carrier, the woman riding on a mule, horse, or other animal, the musician, and the mother nursing a child. Some of these have their mythological counterparts, as in the Aphrodite riding on a goose, or the Earth-mother, already mentioned. Male types are curiously rare, the athletic influences, which are so strongly manifest in early Greek sculpture, not affecting terracottas. The most popular is that of the horseman, particularly in Cyprus. These figures are usually of a rude and primitive kind, especially in Cyprus and at Halikarnassos. The examples from Greece Proper show a more developed archaism, and are found at Athens and in Boeotia. Sometimes instead of a horse the man rides on a swan, mule, or tortoise.

Reclining male figures are sometimes characterised as Herakles or a Satyr; but this type is most fully developed at Tarentum, in numerous terracottas representing the well-known subject of the Sepulchral Banquet, associated with a cult of the Chthonian deities.^[429] There are also various types of grotesque figures, usually in a squatting or crouching attitude; some assume the form of a Satyr, and others are obviously derived from the Egyptian figures of Ptah-Socharis, with bent knees and protruding stomach.

In the fine and later periods, from the end of the fifth century onwards, the standing or seated feminine figures are still by far the most prominent. The change, however, which has taken place, from mythological to genre, has been described as an evolution rather than a revolution, brought about by artistic, not religious, considerations. The possible varieties of the feminine standing types may be best studied in the Tanagra figures (Plate IX.), which include women or girls in every variety of pose or attitude. In most cases the arms are more or less concealed by the himation, which is drawn closely across the figure; in others a fan, mirror, wreath, or mask is held in one hand, the other drawing the edges of the drapery together. Some lean on a column or are seated on a rock; others play with a bird or perform their toilet. Imitations of the Tanagra figures, but vastly inferior in merit, subsequently became popular all over the Greek world; they are found at Myrina in Asia Minor, in Cyprus, the Cyrenaica, and many parts of Southern Italy.

Among miscellaneous types of the Hellenistic period, many of the archaic ones already mentioned retain their popularity. Others appear for the first time, and are more in accordance with the spirit of the age, such as girls dancing, playing with knucklebones, or carrying one another pick-a-back. There is a beautiful group of two knucklebone-players from Capua in the British Museum (D 161). The dancing type is found widely distributed.

Figures of goddesses and mythological subjects are very rare at Tanagra, but fairly common on other sites, as at Myrina and Naukratis. Archaistic imitations of the archaic seated and standing goddesses are often found in the Cyrenaica and Southern Italy; but the Chthonian deities appear but rarely among the types of more advanced style. As in sculpture and vase-paintings, Aphrodite now becomes the most prominent among the feminine deities, and some of the later statuettes appear to be reproductions of well-known works of art, the Cnidian Aphrodite, the Anadyomene, or the crouching type of Aphrodite at the bath. Artemis and Athena are occasionally found, but Nike (Victory) is really the most popular figure after Aphrodite. She, however, plays little more than the part of a female Eros, a counterpart to whom the Hellenic artist felt to be a necessity. Formerly these winged female types were styled Psyche, but this was a conception of post-Hellenistic origin.

Among the male deities the conditions remain much as before. Zeus appears for the first time, and was especially popular at Smyrna, and Sarapis and Asklepios are also occasionally found. In Naukratis the influence of the Egyptian religion made itself felt in the production of numerous figures of Bes, Harpocrates, and the like. Hermes is not found so often as might have been expected, though there is a notable instance in the British Museum (C 406) of a caricature of the famous statue by Praxiteles, where a Satyr takes his place. Dionysos is only met with occasionally, as are Satyrs and Maenads; but masks of a Bacchic character are very common in Italy.

The one deity who really seems to have caught the popular taste is Eros, although at the time when most of the Tanagra statuettes were produced this popularity was hardly assured. The types of Eros standing, seated, flying, or riding on animals are innumerable and found all over the Greek world. The best examples come from Eretria in Euboea, but Myrina and Sicily have also produced large numbers. They vary from almost Praxitelean conceptions, like the Flying Eros from Eretria in the British Museum (C 199), to the veritable Pompeian amoretti from the same site and from Myrina. The riding types of Eros (on a horse, dog, swan, or dolphin) are chiefly found in the Cyrenaica or Southern Italy. In many cases the Eros types are used for ordinary unwinged boys.

Among the human male types a new feature is the introduction of the athlete, as he appears in many boyish figures from Tanagra, and later as a boxer among the somewhat coarse conceptions of the Roman period. Some years ago a remarkable copy of the Diadumenos of Polykleitos in terracotta was found in Asia Minor.^[430]

In the tombs of the Aegean Islands, Italy, and elsewhere, a class of ware has sometimes been found quite distinct from the ordinary fictile pottery and resembling the porcelain or enamelled ware of the Egyptians and Babylonians, such as the ushabtiu, found in the tombs of the former, and the enamelled bricks of the latter. For the most part they must be regarded as importations, of foreign manufacture, the medium of commerce being the Phoenicians, who not only introduced Egyptian objects of art, but themselves endeavoured to imitate them. Hence we must distinguish some as of Egyptian origin, others as made by the Phoenicians. As might be expected, they are most often found where Phoenician influence was strong, as in Rhodes and Sardinia. Egyptian perfume-vases have been found in the Polledrara tomb at Vulci (see Chapter XVIII.) and may be dated by the accompanying scarabs of Psammetichus I. as belonging to the end of the sixth century.

But these are by no means the earliest examples. In the Bronze Age tombs of Cyprus occasional finds have been made of plates of blue porcelain or faïence, with Egyptian designs going back to the eighteenth dynasty^[431]; and for several centuries other Egyptian objects in porcelain, or with enamelled glaze, continue to be found in the tombs of Cyprus, Rhodes, and Greece. And there is also a considerable quantity of such wares which is not Egyptian in character, although it may be to some extent imitative, and therefore demands notice. Of this the most remarkable examples are the rhyta, or drinking-horns, found at Enkomi in Cyprus, in 1896, and now in the British Museum.^[432] The two finest specimens are in the form of a female head surmounted by a cup (Plate X.) and a ram’s head respectively. Although found in tombs with Mycenaean objects, and therefore presumably of early date, the style and modelling are so far advanced—so purely Hellenic—that they may be compared with archaic work of the sixth century B.C. or even later.

In the tombs of Kameiros in Rhodes,^[433] along with Egyptian porcelain objects, were found many vases of this ware, of apparently Greek workmanship. This is further implied by the presence in one tomb of a figure of a dolphin with a Greek [Π]Υ[Θ]ΕΩ [Ε]ΜΙ, “I belong to Pythes.”^[434] It is quite conceivable that the Greeks of Rhodes (as of Naukratis: see below) knew and practised Egyptian methods. The finds include small alabastra with friezes of men and animals in relief, and flasks of a compressed globular shape similarly ornamented; also aryballi of various moulded forms, such as animals or helmeted heads (Plate X. fig. 3). The vase in the form of a head seems to be an early Phoenician idea; and this particular type of the helmeted head seems to have been adopted subsequently by Ionian artists in the Clazomenae sarcophagi.^[435] Similar vases and figures have been discovered in the tombs of Melos, Corinth, Cervetri, and Vulci, and also in Syria and at Naukratis in Egypt.^[436] Others again from the tombs of Kameiros and Vulci take the form of jars of opaque glass ornamented with zigzag patterns in white and dull crimson on a greenish ground.^[437] A specimen of somewhat similar ware was found in a Bronze Age tomb at Curium, Cyprus, in 1895,^[438] consisting of a tall funnel-shaped beaker of blue and yellow glazed ware with an edging of dark brown (Plate X.). The technique is superior to that of the later examples, and more on a level with that of the porcelain rhyta from Enkomi.

In Greece Proper there are altogether few traces of this enamelled ware, and after the sixth century B.C. it quite disappeared. But some very fine specimens have been found in the tombs of Southern Italy. A jug with delicate ornamentation in blue and white came from Naples, and a similar vase from the same site, but shaped like a kalathos and of a pale green colour, is now in the British Museum. Objects of this ware have also been found on the site of the ancient Tharros in Sardinia. Their glaze was a pale green, like that of the twenty-sixth dynasty wares, and with them was found a scarab of Psammetichus I, which shows them to be contemporaneous with the objects found in the Polledrara tomb. But the strong Phoenician element in Sardinia is sufficient to indicate that these fabrics are all of Egyptian importation.

In the Hellenistic period, when vase-painting had reached its latest stages, the fashion of glazed enamelled ware was revived; its chief centre was Alexandria, which would naturally have carried on the traditions of Egyptian porcelain or faïence. Specimens of glazed ware with reliefs or modelled in various forms have been found at Naukratis and in the Fayûm, including a fine blue porcelain head of a Ptolemaic queen (Plate X.). In a tomb at Tanagra were found a beautiful askos in the form of a duck on which Eros rides, and another porcelain vase,^[439] evidently imported from Alexandria, or some other industrial centre of Hellenised Egypt. Porcelain jugs, inscribed with the names of Arsinoe, Berenike, and one of the Ptolemies, have been found at Benghazi in North Africa, at Alexandria, and at Canosa in Southern Italy.^[440] They are of blue ware, with reliefs of Greek style attached. Fragments of the same kind dating from the first century B.C. were found at Tarsos in Cilicia,^[441] and in the Louvre there are glazed wares covered with yellow or green enamel from Smyrna and Kyme. The British Museum possesses similar vases from Kos and elsewhere, with wreaths and similar patterns in relief (Plate X.), but these are not earlier than the Roman period. Enamelled wares of early Roman date have also been found on the Esquiline, and the ware is common at Pompeii.^[442]

It does not appear that the manufacture of these enamelled wares was confined to one spot; they are found all over Asia Minor, Italy, and Gaul, and in other countries bordering on the Mediterranean. It seems probable, however, that there were three principal centres of the fabric, at least in the Roman period. The first of these was in Asia Minor, or the islands along its coasts, whence came the specimens found at Tarsos, in Ionia, and in the islands such as Kos. These are mostly small vases, of metallic form, especially in the treatment of the handles (cf. Plate X., fig. 5), the colour being usually a bluish green, though some examples are more polychromatic. These seem to have been exported to Italy, and viâ Marseilles to Gaul. Next, there are the wares made at Alexandria, of which the vases described above are examples. And, thirdly, there was a Gaulish fabric, which must probably be located at Lezoux in the Auvergne (see Chapter XXIII.), examples from which are found at Vichy, in the Rhone Valley, and at Trier and Andernach in Germany.^[443] Fragments of this ware are even reported to have been found in England—as, for instance, at Ewell in Surrey, at Colchester and Weymouth.^[444] These are of grey clay with yellow, green, or brown glaze, with ornaments of leaves, vine-branches, or scrolls, stamped in moulds; the shapes are jugs, flasks, or two-handled cups. A later variety is of white clay with a malachite-green glaze, the forms being again of a metallic type, and towards the end of the period imitations of glass with barbotine decoration (see Chapter XXIII.) appear. These two groups cover the first century after Christ.

Sometimes the ornamentation of the later glazed wares from Italy takes the form of small reliefs (emblemata), made separately and attached before the glaze was applied, and there are two or three specimens of this class in the British Museum. It was also not infrequently used for lamps, which, apart from the glaze, have all the characteristics of the ordinary kinds, and even for figures of gladiators, boats, and other objects. The glaze is of a thick vitreous character, and was not improbably produced by lead; at all events a French writer^[445] maintains, in opposition to the views of Brongniart and Blümner, that by a study of this ware he has established a knowledge of lead-glaze among the ancients.^[446]