As regards the traditions, even Strabo^[757] realised their absurdity, when he asked, “How could the wheel be the invention of Anacharsis, when his predecessor Homer knew of it?” On the other hand, Poseidonios adheres to the tradition, maintaining that the passage in Homer is an interpolation.^[758] Other allusions to the wheel are in the writings of Plato^[759] and the comic poet Antiphanes.^[760]

Among the Egyptians and Greeks the wheel took the form of a low circular table, turned with the hand, not as nowadays with the foot.^[761] The assumption that the wheel was turned with the foot is only supported by one passage in the Book of Ecclesiasticus^[762]; the evidence of Plutarch^[763] and Hippokrates^[764] tells decidedly against it. In 1840 some discs of terracotta, strengthened with spokes and a leaden tire, came to light on the site of the ancient potteries at Arezzo, and these had evidently been used as potter’s wheels.^[765] The process is also represented on two or three vases, as on a Corinthian painted tablet of about 600 B.C. (Fig. 65),^[766] on a kylix in the British Museum (B 433), on a B.F. hydria in Munich (Fig. 67 b, below), and on a R.F. fragment from the Acropolis of Athens (Fig. 66),^[767] which shows a man modelling the foot of a large krater, while a boy or slave turns the wheel, as on the Munich vase. On the British Museum cup the potter is seated on a low stool, apparently modelling a vase which he has just turned into shape on the wheel.

In making the vases the wheel was used in the following manner:—A piece of paste of the required size was placed upon it vertically in the centre, and while it revolved was formed with the finger and thumb, the potter paying regard not only to the production of the right shape, but to the necessary thickness of the walls. This process sufficed for the smaller pieces, such as cups or jugs; the larger amphorae and hydriae required the introduction of the arm. The feet, necks, mouths, and handles were separately turned on moulds, and fixed on while the clay was moist. They are often modelled with great beauty and precision, especially the feet, which are admirably finished off, to effect which the vase must have been inverted. The modelling and separate attachment of the handle is represented in more than one ancient work of art (see Fig. 66). In many cases the joining of the handles is so excellent that it is easier to break than to detach them. Great technical skill was displayed in turning certain peculiar forms of vases, and generally speaking the Greeks with their simple wheel effected wonders, producing shapes still unrivalled for beauty.

In the case of the earlier vases, which are made by hand, after the clay was properly kneaded the potter took up a mass of the paste, and hollowing it into the shape of walls with one hand, placed the other inside it and pressed it out into the required form. In this way also the thickness of the walls could be regulated. When raised or incised ornaments were required, he used modeller’s tools, such as wooden or bronze chisels. The largest and coarsest vases of the Greeks were made with the hand, and the large πίθοι, or casks, such as have been recently found in such numbers in Crete and Thera (p. 152), were modelled by the aid of a kind of hooped mould (κάνναβος): see ibid.). The smaller and finer vases, however, were invariably turned on the wheel. On a Graeco-Roman lamp from Pozzuoli, in the British Museum,^[768] a potter is seen standing and modelling a vase before his furnace, in the manner no doubt employed at all periods.

Certain parts of the ancient painted vases were modelled by the potter from the earliest times—e.g. on those of the Geometrical period horses are occasionally found on the covers of the flat dishes moulded in full relief, and in other examples the handle is enriched with the moulded figure of a serpent twining round it. This kind of ornament is more suitable to works in metal than in clay, and suggests the idea that such vases were, in fact, imitations of metallic ones. On vases of all periods moulded bosses and heads, like the reliefs on metal vases, are sometimes found; even in black-figured vases the insertions of the handles of hydriae and oinochoae are occasionally thus enriched. In the later styles modelling was more profusely employed; small projecting heads were affixed to the handles of jugs^[769] at their tops and bases, and on the large kraters found in Apulia the discs in which the handles terminated (see above, p. 171) were ornamented with heads of the Gorgon Medusa, or with such subjects as Satyrs and Maenads. These portions were sometimes covered with the black varnish used for the body of the vase, but frequently they were painted with white and red colours of the opaque kind.

A peculiar kind of modelling was used for the gilded portions of reliefs, introduced over the black varnish. When the vase was baked a fine clay was applied to the parts intended for gilding and delicately modelled, either with a small tool or a brush, a process similar to that adopted in the Roman red ware (en barbotine, see Chapter XXI.). It may indeed have been squeezed in a fluid state through a tube upon the vase, and then modelled. As the gilded-portions are generally small, this process was not difficult or important. A vase discovered at Cumae^[770] has two friezes executed in this style, the upper round the neck, representing the Eleusinian deities, delicately modelled, coloured, and with the flesh completely gilded; the lower one consists of a band of animals and arabesque ornaments. Several vases from the same locality, from Capua and from the Cyrenaica, have wreaths of corn, ivy, or myrtle, and necklaces round the neck, modelled in the same style, while the rest is plain.

But the art of modelling was soon extensively superseded by that of moulding, or producing several impressions from a mould, generally itself of terracotta. The subject was in the first place modelled in relief with considerable care; and from this model a cast in clay was taken and then baked. The potter availed himself of moulds for various purposes. From them he produced entire parts of his vase in full relief, such as the handles, and possibly in some instances the feet. He also stamped out certain ornaments in relief, much in the same manner as the ornaments of cakes are prepared, and fixed them while moist to the still damp body of the vase. Such ornaments were principally placed upon the lips or at the base of the handles, and in the interior of the kylikes or cups of a late style. A late bowl of black glazed ware in the British Museum (see Plate XLVIII.) contains an impression from one of the later Syracusan decadrachms having for its subject the head of Persephone surrounded by dolphins: it was struck about 370 B.C. by Euainetos.^[771]

The last method to be described is that of producing the entire vase from a mould by stamping it out, a process extensively adopted in Roman pottery. During the best period of the fictile art, while painting flourished, such vases were very rare; but on the introduction of a taste for magnificent vases of chased metal, the potters endeavoured to meet the public taste by imitating the reliefs of metal ware.

The most remarkable of these moulded vases are the rhyta or drinking-horns, the bodies of which terminate in the heads of animals, produced from a mould (see above, p. 192). By the same process were also made vases in the form of jugs or lekythi, the bodies of which are moulded in the shape of human heads, and sometimes glazed, while the necks were fashioned on the lathe, and the handles added. These were coloured and ornamented on the same principle as the rhyta, the vase-portion being generally covered with a black glaze, but sometimes with a white slip, after the manner of the terracottas. Besides the rhyta, phialae, or saucers, were also moulded; fine examples of which process may be seen on the flat bossed saucers, or phialae mesomphaloi, discussed in Chapter XI., p. 502.

Amphorae and other vases of late black ware, the bodies of which are reeded, were also evidently produced from moulds, and could not be made by the expensive process of modelling. Of smaller dimensions, but also made by moulding, were the vases known as gutti, or “lamp-feeders” (see above, p. 200). They have reeded bodies, long-necked mouths, and circular handles; and on their upper surface a small circular medallion in bas-relief, with a mythological subject. Such vases are principally found in Southern Italy and in Sicily, and belong to the second century B.C. (Chapter XI., p. 502). After being moulded they were entirely covered with a black glaze. Other vases again are entirely moulded in human or animal forms, with a small mouth or spout. These are found at all periods, but chiefly in the archaic Rhodian and Corinthian fabrics, and again reviving in the later stages of vase-fabrics in Southern Italy. Examples may be seen in the First Vase Room (Cases 33–34 and F) and Fourth Vase Room (Case B) of the British Museum: see also Plate XLVI. Others again retain the form of the jug or lekythos, with a figure or relief attached to the front of the body and coloured or covered with a white slip, while the back is varnished black. The whole subject is treated in fuller detail in Chapter XI.

Many vases of the fourth century and later are entirely covered with a coating of black glaze, while rows of small stamped ornaments, apparently made with a metal punch, have been impressed on the wet clay before the glaze was applied. These decorations are unimportant in their subjects, which are generally small Gorgons’ heads, tendrils, or palmettes, and hatched bands, arranged round the axis of the vase. This latter ornament was probably produced by rolling the edge of a disc notched for the purpose round the vase, in the same manner as a bookbinder uses his brass punch. When these vases came into use the potter’s trade had ceased to be artistic, and was essentially mechanical. They are found on almost all sites from Cyprus to Italy.^[772]

After the vases had been made on the wheel they were duly dried in the sun^[773] and lightly baked, after which they were ready for varnishing and painting; it is evident that they could not be painted while wet and soft. Moreover the glaze ran best on a surface already baked. It is also probable that the glaze was brought out by a process of polishing, the surface of the clay being smoothed by means of a small piece of wood or hard leather. At all events this seems the most satisfactory interpretation of a vase-painting in Berlin (Fig. 67a),^[774] where a boy is seen applying a tool of some kind to the outer surface of a completed vase (kotyle); that the vase is not yet varnished is shown by its being left in a red colour, while two others, varnished black all over, stand on the steps of an oven close by, probably to dry after the application of the varnish.

Many vases, whether decorated with designs or not, are varnished black throughout the exterior, except the feet and lips, and we cannot be certain whether or not any glaze had been previously applied to the surface; but in respect of the red-figured vases, it is clear from the method employed (see p. 221) that they were originally glazed throughout.

This lustrous glaze is, like the black varnish, now quite a lost art. Seen under a microscope it has evidently been fused by baking; it yields neither to acids nor the blow-pipe. It is remarkably fine and thin, insomuch that it can only be analysed with great difficulty. No lead entered into its composition. It is however far inferior to modern glazes, being permeable by water; but it is not decomposed by the same chemical agents. On the later R.F. vases it is of decidedly inferior quality, and often scales away, carrying the superimposed colours with it.^[775]

3. The Baking of Vases

The process of baking (ὀπτᾶν, coquere) was regarded as one of the most critical in the potter’s art. It was not indeed universal, as Plato^[776] distinguishes between vases which have or have not been exposed to the action of fire (ἔμπυρα and ἄπυρα), and Pliny^[777] speaks of fictile crudum (ὠμόν) used for medicinal purposes. But all the vases that have come down to us have certainly been baked. The necessary amount of heat required was regulated by the character of the ware, and in the case of most Greek fabrics it appears to have been high. Many examples exist of discoloured vases which have been subjected to too much or too little heat, and in which the varnish has acquired a greenish or reddish hue. On the other hand, in some of those that have been subjected to subsequent burning, the red glaze has turned to an ashen-grey colour,^[778] the black remaining unimpaired; but there are also instances of the varnish peeling off, the red colour alone preserving the outline of the figures.

Other accidents were liable to befall them in the baking, such as the cracking of the vase under too great heat; this produced an effect expressed by the term πυρορραγής or φοξός, words which seem to have some reference to the sound of a cracked pot.^[779] Or the shape of a vase might be damaged while it was yet soft, one knocking against another and denting its side, or crushing the lip through being carelessly superimposed. On a R.F. amphora in the British Museum (E 295) a dent has been caused by the pressure of another vase, which has left traces of a band of maeanders. This probably happened when the vases were in the kiln for the second firing. The quality of the baking was tested by tapping the walls of the vase.^[780]

These misfortunes were attributed to the action of malicious demons, whose influence had to be counteracted in various ways; thus, for instance, a Satyric or grotesque head was placed in front of the furnace and was supposed to have an apotropaeic effect against the evil eye.^[781] The pseudo-Homeric hymn addressed to the potters of Samos invokes the protection of Athena for the vases in the furnace, and mentions the evil spirits which are ready to injure them in the case of bad faith on the potter’s part. Among the names given are: Ἄσβεστος, “the Unquenchable”; Σμάραγος, “the Crasher”; Σύντριψ, “the Smasher”; Ὠμόδαμος, “the Savage Conqueror.”

The form of the oven probably differed little from those in use at the present day. No furnaces have been found in Greece, and our only evidence is derived from the painted vases; but they have been found at Ruvo^[782] and elsewhere in Italy, and also in France, Germany, and England. Those of Roman date are indeed by no means uncommon, but are discussed in fuller detail in the corresponding section of the work (Chapter XXI.).

As depicted on vases and elsewhere, the ancient furnaces seem to have been of simple construction, tall conical ovens fed by fires from beneath, into which the vases were placed with a long shovel resembling a baker’s peel. The kilns were heated with charcoal or wood fuel, and in some of the representations of them we see men holding long instruments with which they are about to poke or rake the fires (Fig. 68). They had two doors, one for the insertion of the vases and one for the potter to watch the progress of the baking. For vases of great size, like the huge πίθοι, special ovens must have been necessary; and we have a representation on a Corinthian pinax^[783] of such an oven, the roof of which resembles the upper part of a large pithos surrounded by flames.

On the lamp from Pozzuoli in the British Museum, referred to on p. 209, there is a curious subject in relief, representing a potter about to place a vase in an oven with a tall chimney; and on a hydria at Munich^[784] (Fig. 67 b) a man is about to place an amphora in a kiln, while other jars (painted white) stand ready to be baked. But for our purposes the Corinthian pinakes are even more valuable for the information they afford. There are several representing the exterior of the conical furnace, with men standing by watching the fires and tending them with rakes^[785]; in another we have a bird’s-eye view in horizontal section of the interior of an oven, filled with jugs of various forms (Fig. 69). Flames are usually indicated rising from underneath the ovens.^[786]

The Munich hydria (Fig. 67b) reproduces the interior of a potter’s workshop with such detail that a full description of the scene may be permissible.^[787] On the left of the picture a seated man seems to be examining an amphora, which has just been finished (it is painted black) and is brought up for his approval. Next is seen an amphora on the potter’s wheel, painted white to indicate its imperfect state; one man places his arm inside to shape the interior, while another turns the wheel for him. On their right another white amphora is being carried out, just fresh from the wheel, but without handles or mouth, to be dried in the open or at the furnace; next is another standing on the ground to dry. On the right of the scene stands the foreman or master of the pottery, before whom a nude man carries what has been thought to be a sack of coals for the furnace, which is seen on the extreme right.

Even more vivid and instructive, in spite of its careless execution, is the painting on a kotyle found at Exarchos or Abae in Lokris, and now in the Athens Museum (Fig. 70).^[788] The style is that of the imitation B.F. vases found in the temple of the Kabeiri at Thebes, late in the fifth century. We see represented the interior of a potter’s workshop, in which the master of the business sits holding up a kylix in one hand, while with the other he threatens a slave, who runs off with three kotylae ready for the furnace; three similar kotylae stand by the master’s feet, and behind him are two more vases on a shelf. On the right of the scene a workman sits at a table on which is a pot full of paint, with a brush in it; he holds up a newly-painted kotyle, admiring his workmanship. The picture is completed by a realistic representation of an unfortunate slave suspended by cords to the ceiling as a punishment for some offence, while another belabours him with a leather thong.

It would appear that the vases after the baking were often placed on the exterior of the furnace, either to prevent the too rapid cooling of the clay, or (as indicated on the Berlin cup) for the pigments to dry. Jahn and others have published a gem^[789] on which a small two-handled vase is placed on the top of an oven, and a youth is applying two sticks to it, perhaps in order to take it down without injury by the contact of the hand. A companion gem,^[790] on which an artist is painting a similar jar, shows a jug and a kylix standing on a kiln.

When the vases were returned from the furnace, the potter appears to have made good as far as possible the defects of those not absolutely spoiled; and if naturally or by accident any parts remained too pale after the baking, the defect was remedied by rubbing them over with a deep red ochre, which supplied the necessary tone.

4. Painting

We may distinguish three principal classes of painted pottery, of which one at least admits of several subdivisions:—

(1) Primitive Greek vases, with simple painted ornaments, chiefly linear and geometrical, laid directly on the ground of the clay with the brush. The colour employed is usually a yellowish or brownish red, passing into black. The execution varies, but is often extremely coarse.

(2) Greek vases (and Italian imitations) painted with figures. These may be subdivided as follows:—

(a) Vases with figures in black varnish on red glazed ground (see Frontispiece, Vol. II.);

(b) Vases with figures left in the red glaze on a ground of black varnish (see Frontispiece, Vol. I.).

(3) (a) Vases of various dates with outline or polychrome decoration on white ground (see Plate XLIII.);

(b) Vases (also of various dates) with designs in opaque colour on black ground.

Of these, the second group is by far the largest and most important, and the complicated and technical processes which it involved will demand by far the greater share of our attention in the following account of the methods of painting. In both the classes (a) and (b) the colouring is almost confined to a contrasting of the red glazed ground of the clay with a black varnish-like pigment, a contrast which perhaps more than anything else furnishes the great charm of a Greek vase.

This black varnish is particularly lustrous and deep, but varies under different circumstances. Great difference of opinion has always existed as to its nature, and the method by which it was brought to such perfection by the Greeks. The variations in its appearance are due partly to differences of locality and fabric, partly to accidents of production. It is seen in its greatest perfection in the so-called Nolan amphorae of the severe red-figure period; and at its worst in the Etruscan and Italiote imitations of Greek fabrics. On the vases found at Vulci it shows a tendency to assume a greenish hue, as opposed to the blue-black of the Nolan vases, while variations in the direction of red, brown, and (on late South Italy fabrics) grey are of frequent occurrence. It is probable that these gradations of quality are mainly due to the action of fire, according as a higher or lower temperature was employed. On the other hand, the ashen-grey hue which vases of all periods sometimes assume^[791] seems to be due to the direct action of fire in contact with them, and this may perhaps be explained by supposing that they had been burnt on a funeral pyre. This varnish also varies in the thickness with which it was laid on, as can be easily detected with the finger.

Although the chemical action of the earth sometimes causes the black varnish to disappear entirely, leaving only the figures faintly indicated on the red-clay ground, there has never yet been found any acid which has any effect upon it.^[792] Various opinions have been promulgated, from Caylus downwards, as to the elements of which it is composed.^[793] Brongniart^[794] has analysed it with the following results:—

It is unnecessary here to enter in detail into the numerous other theories of its composition, but so far it cannot be said that any certainty has been attained.

Turning now to the methods by which the black varnish was applied, we find it necessary to distinguish between the two classes of black-figured and red-figured vases; some vases, of course, are completely covered with it, having no painted design, but these do not enter into the question.

In the black-figured vases the figures are painted in black silhouette on the red ground of the vase, the outlines being first roughly indicated by a pointed instrument making a faint line.^[795] The surface within these outlines was then filled in with the black pigment by means of a brush, the details of anatomy, drapery, armour, etc., being subsequently brought out in part by further incising of lines with a pointed tool. In some of the finest vases, such as those of Amasis and Exekias (p. 381), the delicacy and minuteness of these lines is brought to an extraordinary pitch of perfection. After a second baking had taken place, the designs were further enriched by the application of opaque purple and white pigments, usually following certain conventional principles, the flesh of women and devices on shields, for instance, being always white, folds of drapery always purple. A third baking at a much lower heat was necessary to fix these colours, and the vase was then complete.

It should here be noted that there are really two subdivisions of these black-figured vases, which may be termed for convenience “red-bodied” and “black-bodied.”^[796] In the former the whole vase stands out in the natural red colour of the clay; whereas in the latter the treatment approaches more nearly to the red-figure method which we shall presently discuss. The whole body of the vase is in these examples covered with the black varnish, with the exception of a framed panel of red, on which the figures are painted. This distinction may be well observed in the Second Vase Room of the British Museum, where most of the vases on the east side of the room belong to the former or “red-bodied” class, while all those on the west side are “black-bodied,” with designs in panels.

In the red-figured vases the black varnish is used as the background, and covers the whole vase, as in the “black-bodied” B.F. fabrics, the figures not being actually painted, but left red in the colour of the clay. The process was as follows:—Before the varnish was applied the outlines of the figures were indicated, not by incised lines but by drawing a thick line of black with a brush round their contours. It is probable that a fine brush was used at first, especially for more delicate work, and then a broader brush producing a line about an eighth of an inch in thickness. The process, be it noted, is more akin to drawing than painting; and it was as draughtsmen par excellence that the red-figure artists excelled. The next stage was to mark the inner details by means of very fine black lines (corresponding to the incised lines of B.F. vases), or by masses of black for surfaces such as the hair; white and purple were also employed, but far more sparingly than on the earlier vases. In the late Athenian and South Italian vases a tendency to polychromy sprang up, but the main process always remained the same to the final decadence of the art. The figures being completed and protected from accidents by their broad black borders, the varnishing of the whole exterior surface was then proceeded with. This was of course a purely mechanical business. A fragment of a red-figured vase in the Sèvres Museum forms an excellent illustration of the method employed, as, although the figures are finished, the ground has never been filled in, and the original black border is plainly visible (Fig. 71).

The result of the second baking was to fix the varnish and cause it to permeate the surface of the clay in such a way as to become practically inseparable from it. The subsidiary colours, on the other hand, which were laid on over the black, are always liable to disappear or fade.

A very interesting representation of painters at work on their vases is to be seen on a hydria from Ruvo (Fig. 72).^[797] Three painters are seated at work with their brushes, of whom two are being crowned by Victories, while the third is about to receive a wreath from Athena, the protecting goddess of the industry. Their paint-pots are to be seen by their side. At one end of the scene a woman is similarly occupied.

From Blümner.

In class 3 (a), or vases with figures on white ground, we have to deal with the process of covering the naturally pale clay with a white slip of more or less thick and creamy consistency, on which the designs were painted. In the archaic period this process is fairly common, especially in the earliest vases of Corinth and of Ionia, and at Kyrene and Naukratis. It was revived at Athens about the end of the sixth century (see pp. 385, 455). But when once the white slip was laid on, the technical process differed little from that in use on ordinary red-ground vases, except for the general avoidance of white as an accessory; it merely results that instead of a contrast of black and red, one of black and cream is obtained. The method was one also largely practised in early painting, as we see in the Corinthian pinakes and the sarcophagi of Clazomenae (pp. 316, 362).

But there is another class of white-ground vases to which we must devote more special attention, namely, those on which the figures are painted either in outline or with polychrome washes on the same white slip. The earliest instance of such a method is in the series of fragments found at Naukratis, dating from the beginning of the sixth century (see p. 348), which technically and artistically are of remarkably advanced character, and combine the two methods of painting in outline and in washes of colour. In the fifth century the practice was revived at Athens as a means of obtaining effective results with small vases, and became especially characteristic of one class, the funeral lekythi, which are elsewhere described (Chapter XI.). This, however, must serve as the most convenient place for a few remarks on their technique.

The vases, after they had left the wheel and were fitted with handle, etc., were covered with a coating of white flaky pigment, in consistency resembling liquid plaster of Paris, or, when dry, pipeclay. They received this coat of white while still on the wheel, and then a second coating, of the usual black varnish, was applied to such parts as were not required for decoration. Usually the white covered the cylindrical part of the body, and the shoulder up to the neck; black was applied to the mouth, neck, handle, base of body, and stem. The clay, it should be noted, is of the ordinary kind, but two varieties have been distinguished, one of pale red, for light thin vases, the other of a blackish-grey, for thicker and heavier ware. The natural colour appears on the inside of the lip and foot. Before being removed from the wheel the vases were finely polished, which gave to the white coating a sort of lustrous sheen; they were then fired at a low temperature.

The method of decoration^[798] was usually as follows:—A preliminary sketch was made with fine grey lines, ignoring draperies (hence the lines of figures are usually visible through the draperies), but not always necessarily followed when the colours were laid on. This was done as soon as the first lines were dry, the colour being applied with a fine brush and in monochrome—black, yellow, or red—following the lines of the sketch more or less closely. In the later examples red was used exclusively, and at all periods at Athens; but in the vases attributed to Locri and Sicily, a black turning to yellow is used. This combination of black and yellow is also used on the best Attic vases for various details, such as eyes and hair. The outlines also served to indicate the folds of the draperies. For the surfaces of drapery and other details, polychrome washes were employed, the colour being spread uniformly by means of a large brush. All varieties of red from rose to brown are found, also violet, light and brownish yellow, blue, black, and green. Hair is sometimes treated in outline, sometimes by means of washes. It is noteworthy that in the later examples the wash-colours were often painted right over the red lines. On the bodies of the figures these washes are rare, but in some cases shades of brown are used for flesh colour, as on the figure of Hypnos on a lekythos in the British Museum (D 58).

At Athens this polychrome decoration was not indeed limited to the lekythi, but was extended to the kylix, the pyxis, and other forms, of which some beautiful examples exist in the British Museum and at Athens.^[799] In these, as in the best of the lekythi, the drawing of Greek artists seems almost to have reached perfection, and arouses our wonder yet more when we reflect that everything was done merely by freehand strokes of the brush. This technique is practically limited to the period 480–350 B.C.

The subsidiary ornamentation of the lekythi was put on either after the main design or before, this being immaterial. The lines above the design can be seen to have been painted on the wheel, as they go all round the vase; but the palmettes on the shoulder and maeander patterns above the design do not extend beyond it. After the colouring the vases appear to have been fired again, and in some cases the white slip was probably varnished. The details of their manufacture show that the lekythi were not intended for daily use; the shape is awkward for handling—the handles, for instance, are obviously not intended for practical use—and the delicate, lightly baked slip made them too porous for liquids. Everything tends in the direction of elegance and delicacy.

Our next sub-division consists of vases, chiefly of late date, in which the decoration is by means of opaque colours laid on the surface of a vase altogether coated with black varnish or glaze. The process is not indeed one absolutely unknown in earlier times, for there is the primitive Kamaraes ware of Crete (p. 266), and also a small series of archaic vases belonging to the early part of the fifth century (p. 393) in which this principle is adhered to, the designs being painted in opaque red or white on the black varnish. The latter seem to show a development from the black-figure period, to the end of which they belong, and may have been intended to rival the new red-figure method, but failed to attain popularity.

We next meet with the process in Southern Italy, where it again appears as the last effort of a worn-out fashion to flicker into life with renewed popularity. The centre of this revival, which follows on after the Apulian vases of the third century, was Gnatia (Fasano), on the coast of that district. The vases are partly modelled in relief, or have ornaments in relief attached; the decoration, in white and purple, is confined to one side only, and is very feeble and limited in its scope. An apparently local variety, perhaps made in Campania by native craftsmen, has the figures in opaque red, with details marked by rudely incised lines.

The Gnatia style was adopted by the Romans in the second century for a small series of vases inscribed with names of Italian deities, such as Juno and Vesta (p. 490), and it appears in the method of decoration known as en barbotine on the pottery of the Empire (see Chapters XXI., XXIII.).