From the use of this word sigilla (a diminutive of signum), for terracotta figures, the makers came to be known as sigillarii, or figuli sigillatores,^[2648] and a street in which they lived was known as the Via Sigillaria.^[2649] There was also a market for the sale of sigilla for the feast near the Pantheon.^[2650] Although the names of makers are constantly found on Roman lamps and pottery, as well as the tiles, they are very seldom found on statuettes, with the exception mentioned below of those found in Gaul. But the name of Q. Velius Primus, in a sort of mixture of Greek and Latin, is found in raised letters on a mask of a Satyr in the British Museum (D 177 = Fig. 197), and other names are occasionally found on the moulds. The social condition of the Roman potter seems to have been much lower than that of the Greek, who was often a person of respectable position; but this may be partly due to the fact that his clientèle was drawn mainly from the poorer classes. He was generally a slave, sometimes a barbarian, and even the masters of the potteries were only freedmen. As we saw in the case of the tile-makers, the potters often worked on the estates of wealthy or influential people, from which their clay was obtained. More details of Roman potters will be found in the sections dealing with tiles and lamps.

On the technical aspect of Roman terracotta figures little need be said. The processes were practically the same as those described in Chapter III. when dealing with the Greek terracottas. Large figures were made from models (proplasmata) and built up in several pieces on a wooden framework, known as crux or stipes.^[2651] A reference to this method may be traced in a fable of Phaedrus,^[2652] which describes Prometheus as having made human figures in clay in separate pieces, and, on returning from a supper with Bacchus, joined them together wrongly, so that the sexes became confused. The smaller figures were all made from moulds, by means of which they could be repeated with but slight alterations. Few statuettes seem to have been made after the second century of the Empire.

The range of subjects in Roman terracottas is much the same as in the Greek figures of the Hellenistic period. At Pompeii genre figures predominate, including such types as gladiators, athletes in the circus, slaves carrying bundles, and personages in Roman costume.^[2653] A favourite type at Pompeii is a mask of a youth in a Phrygian cap.^[2654] There is a decided preference shown for portraits and grotesques. Von Rohden,^[2655] in dealing with the question of the extent to which these figures represent Greek or purely Roman types, considers that although the influence of the former is still strong, yet they are marked by such wide differences that they must be ranked in the latter category. He dates them in the time of Vespasian, in which the decadence which had begun with the later Hellenistic age is in the Roman fabrics still more strongly accentuated. The style is negligent, the proportions faulty, and the art of colouring practically lost. They are only redeemed from insignificance by the taste for portraiture and the interest which attaches to the reproduction of motives borrowed from contemporary life.

The Pompeii figures may serve as typical Roman terracottas, but they are also found elsewhere in Italy, as well as in other parts of the Roman Empire; nearly all, however, are of inferior merit and execution. At Praeneste in 1878, on the site of the temple of Fortuna Primigenia, were found genre figures and votive objects,^[2656] and similar ex votos have come to light at Gabii.^[2657] At Nemi figures have been found which are obviously of Roman date, some of considerable size.^[2658] From time to time finds have been made in Rome, and there is a pretty little head in the British Museum found in the Tiber (D 383), which, however, may be of Greek workmanship. The industry also extended from Rome to the provinces, and even in Britain terracotta figures are sometimes found, as at Richborough^[2659]; at Caistor, by Norwich, a terracotta head of Diana, of fairly good style, is recorded.^[2660] There are also in the Guildhall Museum some terracottas in the coarse red clay which characterises most of the British examples: a Venus on a swan; a female head with turreted crown, of archaistic style, from Finsbury; and a large figure of Proserpina holding a fruit, of very fair style, from Liverpool Street.^[2661] A figure of a boy on horseback is or was in the Museum of Practical Geology.^[2662]

2. GAULISH TERRACOTTAS

In Gaul there appear to have been very extensive manufactures of terracottas, but not anterior to the conquest by Julius Caesar in 58 B.C. These statuettes were made for the Roman colonists, who introduced the types of their own religious conceptions, but the makers were local craftsmen. Potteries have been unearthed at Moulins on the banks of the Allier, and in Auvergne and other parts of France, and even in Germany, where one was discovered at Heiligenberg in Alsace, and others on the Rhine (see below, p. 384). The finds on the Allier, made in 1857, give a practically complete survey of the subjects; they are all now collected in the museums of Moulins and St. Germain, and were fully published at the time in a work by M. Tudot.^[2663] The figures found here are not from tombs, but were unearthed from the sites of the potteries and from ruins of buildings; they are all made in a peculiar white clay, whereas the figures of the Gironde district are grey or black, and those of the Rhine Valley reddish, like those of Britain. The technique resembles that of the Roman figures; there is no vent-hole, and they usually stand on a conical base; the modelling is very heavy, and the latest specimens are absolutely barbaric.

Until recently the subject of Gaulish terracottas had been greatly neglected; Tudot’s plates were useful, but his text unsatisfactory and devoid of method, there being no proper description of the plates. M. Pottier has given a good summary of his work, and M. Héron de Villefosse has also dealt with some aspects of the subject.^[2664] But they had not been treated as a whole and in relation to the subject of ancient terracottas in general until 1891, when an important memoir by M. Blanchet appeared, in which a complete survey of the Gaulish terracottas was given.^[2665] This must of necessity form the basis of the present account.

In dealing with the technical character of the terracottas found in Gaul, M. Blanchet points out that the white clay of which many are made (e.g. those from the Allier valley) is not universal; some are made of red or grey clay, which has turned white in the baking, apparently by a process analogous to that used by the Chinese for porcelain, others are actually covered with a white engobe like the Greek terracottas. This appears to have been done with a view to subsequent colouring, which in nearly all cases has quite disappeared; but statuettes with remains of colouring, made of purely red clay, have recently been found in the neighbourhood of the Moselle and in Germany.^[2666] M. Blanchet quotes an example in the Museum at Angers, with the name of the maker, P · FABI · NICIAE, which is coated with a lead glaze like the enamelled wares described in Chapter III. He considers that the moulds from which they were made were often of bronze, and that bronze models were used as copies; but that they were also of terracotta is clear from the numerous examples given by Tudot. A terracotta mould for a figure of Venus Anadyomene, found at Clermont-Ferrand, is in the British Museum, and another from Moulins is for the back of the head of a similar figure, with hair elaborately coiled.^[2667] From the numerous moulds which have been found it may be seen that the figures were cast in two pieces, longitudinally, the arms being added afterwards, together with the circular plinth. The mould in the British Museum may be cited as an example of one for the back part of a figure; probably only the upper part was modelled.

Potters’ names are exceedingly common, not only on the figures, but also on the moulds,^[2668] and form two distinct classes, those on the exterior of the moulds, and those on the figures or interior of the moulds (which are obviously the same thing). The distinction is that the former were merely for the identification of the moulds, while the latter indicated the creator of the type and made him known to the world, a feature which, as will be noted in Chapter XXIII. (p. 511), reappears in the pottery of Westerndorff in Germany. Tudot gives an example of a mould with the name ATILANO on the exterior and IOPPILLO on the inside.^[2669] Many of the names are identical with those of the makers of vases,^[2670] but the types and subjects are quite distinct from those on the Gaulish terra sigillata. Those on the exterior of the moulds are usually in a scrawling cursive type, whereas the other class are in capital letters^[2671]; the cursive characters resemble those in use at Pompeii, but are not necessarily contemporary; they are, however, not later than the second century. The influence of this cursive character seems to have extended to the other class; for instance, in the inscription given in Fig. 198 below, not only are the G and S of cursive form, but E appears in the form II. Otherwise the letters are in the ordinary Roman alphabet (with the exception of A, which is sometimes 1514Attic alpha the forms E and II seem to have been used indifferently in Gaul at all periods. The “signature” sometimes combines the two names, as in the form

which has been taken to mean Sacrillos fecit forma Caratri, “made by Sacrillos from Caratrius’ mould.”^[2672] Among the Roman names which occur are Attilianus, Lucanus, Pistillus, Priscus, Taurus, and Tiberius; among the Gaulish, Abudinus, Belinus, Camulenus, and Tritoguno.

A large majority of the existing statuettes were, as we have seen, made in the valley of the Allier; these show more conspicuously than any others, the influence of transplanted Graeco-Roman art. Curiously enough none have been found at Lezoux, one of the chief pottery-centres of Gaul, although there is abundant evidence that the vases and statuettes were made in the same workshops (see above).^[2673] M. Blanchet considers that there was a large and important manufacture in Western France, which may have been inspired by the Allier workshops, but mainly exhibits native characteristics; he also notes the scarcity of these figures in Southern Gaul (Narbonensis), which may perhaps be explained by the preference there shown for bronze statuettes and vases with medallions (p. 530).^[2674] Other centres were Cesson, Meaux (where Atilanus and Sacrillos can be located), Bourbon-Lancy in Saône-et-Loire, and St. Rémy-en-Rollat (see p. 516), where vases also were made of the local white clay. M. Déchelette has been able to assign to the last-named pottery a date between A.D. 15 and 50. Another fabric was in the neighbourhood of Liège, and in Germany there were centres at Salzburg, and at Cologne, where the maker Vindex can be dated in the reign of Postumus (A.D. 260-270).^[2675] An important maker, Pistillus, had a pottery at Autun; his statuettes are found all over Gaul,^[2676] and the name appears on vases and coins, and also in an inscription.^[2677] Julius Allusa had a workshop at Bordeaux. In West and North-West France statuettes are found with the name of Rextugenos; they are all of peculiar and original character, with highly-ornamented backgrounds to the figures, and easily distinguished. The specimen given in Fig. 198, representing Venus Genetrix, was found at Caudebec-les-Elbeuf in Normandy (Seine-Inférieure); it bears the inscription RIIXTVGIINOSSVLLIASAVVOT, Rextugenos Sullias auvot (sc. fecit).^[2678]

An interesting find of terracotta figures was made at Colchester in 1866,^[2679] consisting of thirteen figures presenting exact analogies to the Gallo-Roman terracottas of the second period both in type and style. One very poor specimen represents Hercules with club and lion-skin; another a bull, and a third a bust of a boy (perhaps a portrait of Nero or Britannicus); four are recumbent figures. The rest are more or less grotesque, including caricatured seated figures holding books or rolls, and a buffoon. With them were found vases in the form of animals of yellow-glazed ware. Figures of suckling goddesses (see below) have been found in Britain, and similar finds of Gallo-Roman types in white clay in London, among them a Venus holding a tress of her hair.^[2680] Votive offerings of parts of the body and figures of the goddess Fecunditas were found near the source of the Seine, in a temple of Dea Sequana, the local river-deity.^[2681] Other finds have been made in Touraine, Anjou, La Vendée, Brittany, and Normandy, brought by commerce from the Allier potteries; and in Germany at Heddernheim and on the Rhine. Part of a group of some size in purely Graeco-Roman style from the Department of Marne is now in the British Museum (Morel Collection).

Tudot originally classified the Gaulish terracottas chronologically in three periods according to style, and in this he has been followed by M. Pottier. But M. Blanchet^[2682] has pointed out that the former’s method was altogether unscientific, that he trusted too much to the evidence of coin-finds, and that he was altogether wrong in conceiving the possibility of any being anterior to the Roman conquest. On the whole the chronological data are exceedingly vague, and can only be accepted in isolated instances, as in the case of the finds at St. Rémy-en-Rollat (A.D. 15-50) or Cologne (A.D. 260-270), or where a resemblance in the coiffure of the feminine figures to those of Roman ladies can be traced. Some figures may probably be dated about A.D. 100 on the latter ground, the head-dress recalling those of Domitia and Julia the daughter of Titus. But it can only be laid down with certainty that the manufacture of statuettes was introduced into Gaul with the terra sigillata or ornamented red pottery at the beginning of the Imperial period. Where there is a question of decadent or barbaric style, as is undoubtedly often the case, it does not necessarily imply a late date, but only that the inferior work is due to the incapacity of some local artist, and figures of varying style must frequently be contemporaneous.^[2683]

In dealing with the types of Gaulish terracottas, their origin and signification, M. Blanchet divides the subjects into three classes, of which the first is not only the largest but the most interesting: divinities, subjects from daily life, and animals. The deities are not those we should expect from Caesar’s statement^[2684] that Mercury, Apollo, Mars, Jupiter, and Minerva represent the scale of popularity in Gaul, for they are mainly variants of one type, that of Venus. Many of these Venus figures reproduce types familiar in Greek and Graeco-Roman art, such as the Anadyomene, and the Cnidian or Pudica type; but in the majority she is frankly recognised as a Nature-goddess (Aphrodite Pandemos or Venus Genetrix), and hence we find numerous examples in which the old Oriental conception of the nude Aphrodite-Astarte with pronounced sexual characteristics, so common in the primitive terracottas of Chaldaea, Phoenicia, and Cyprus,^[2685] once more reappears, as in Fig. 198. Of almost equal frequency is the seated type of the Mother-Goddess or Κουροτρόφος, suckling a child^[2686]; this is not peculiar to Gaul, but is found in the terracottas of Southern Italy.^[2687] We may compare also the Fecunditas types on Roman coins.^[2688] Blanchet thinks that the goddess Rumina may be here intended, but prefers to adopt the general term of Mother-Goddess.

Among other mythological types the Ephesian Artemis, Pallas, Mercury, Epona (Fig. 199), and Abundantia occur; and among genre subjects the most interesting type is that of the Spinario, or boy extracting a thorn from his foot, familiar in Greek sculpture. Slaves, caricatures, and busts of ladies (see above) or children wearing the bulla, vases in the form of heads, and busts affixed to plates, also come under the latter category. Many of these are exceedingly rude and barbaric; children are transformed into coarse grotesques, and animals look (says M. Pottier) as if they had come out of a Noah’s ark.

The artistic origin of the Gaulish types has been discussed by M. Blanchet,^[2689] who points out that although the modern tendency is to restrict the rôle played by Alexandrine art of the Hellenistic period in influencing that of Rome,^[2690] yet its effect on Gaul cannot be altogether ignored. That Egyptian cults found their way into Gaul is well known,^[2691] and in the terracottas such types as Isis and Horus appear, while comparisons may frequently be made with the late terracottas found in the Fayûm and at Naukratis. But there was also a stream of influence from Southern Italy, especially Campania, whence, as we have seen, the Mother-Goddess types were largely derived.

As regards the uses for which these terracottas were made, much that has been said on that head in Chapter III. will apply equally to Gaul. They have been found not only in tombs, but in wells and rivers, and on the sites of sanctuaries^[2692]; but they do not seem to have had any special funerary significance. The majority were probably used for various domestic purposes in the houses, the figures of animals, for instance, as toys, and were then buried with their owners. Those found in wells or rivers may be regarded as votive offerings, as it is well known that the Gauls were fond of throwing votive figures into rivers or springs.

3. MISCELLANEOUS USES OF TERRACOTTA

It is impossible to enumerate all the purposes to which the Romans applied terracotta, but a few peculiar uses deserve special notice. The excavations at Pompeii have yielded several examples of its application to the decoration of a puteal, the circular structure which protected the mouth of a well; the core is of tufa or other hard material, and round this are laid curved slabs of terracotta decorated with reliefs.^[2693] They are all of comparatively early date; one has triglyphs and bulls’ heads in relief, and is stuccoed over. Instances are also found at Pompeii of its use for table-legs, in the form of figures of kneeling Atlantes,^[2694] like those supporting the entablature in the Thermae (p. 374), but sculptured in the round. Small altars, or stands for holding lamps or for burning incense, supposed to have formed part of the furniture of the domestic shrines, have also been found in this material.^[2695] Varro tells us that the dolia or large jars made by potters were used as cages for dormice which were being fattened for the palates of Roman epicures^[2696]; and Columella gives instructions for the use of clay tiles in making beehives.^[2697] Porphyry implies that it was customary to hive bees in kraters or amphorae of clay.^[2698] Tickets (tesserae) for admission to the circus or amphitheatre were also occasionally made of clay, and on them were stamped letters or numbers referring to the position of the seat, or representations of the animals exhibited. Two from Catania in the British Museum^[2699] have an elephant on the obverse and the letter A on the reverse, showing that they were for admission to a spectacle in which those beasts were shown. There are also possible instances of tesserae frumentariae, or tickets for the supply of cheap corn in time of necessity.^[2700] Moulds of terracotta for making counters, with masks or figures of Fortune and Isis, have also been found; there is an example in the British Museum from Arezzo (E 46).^[2701]

Herr Graeven, in a very interesting article,^[2702] has recently collected all the known examples (numbering some fifty) of money-boxes in terracotta used by the Romans. There is no mention of such objects in Latin literature, but it is probable that they were known as loculi, and were made in imitation of the metal Θησαυροί used for keeping money in temples. Of this there is a clear instance in a specimen recently found at Priene in Asia Minor,^[2703] in the form of a small shrine with a slit in the top. Graeven states that there is evidence of their having been placed on a cornice which ran round the walls of the rooms in the houses. This box has an additional hole at the back for extracting the money, but the Roman specimens have only one opening. An example of a clay treasure-box from Western Europe is one in the form of a chest, 12½ inches high, with a bust of Apollo on the top, found at Vichy, and now in the Museum at Moulins.^[2704] It may have been placed in a sacellum or chapel for the offerings of those who visited the medicinal springs.

Of the Roman money-boxes proper four main types may be distinguished. The first, of which examples have been found at Pompeii,^[2705] is in the form of a small chest or coffer (arca), and may have been known by the name arcula. The second type is that of a money-box in the form of a vase.^[2706] The custom of hoarding money in jars (ollae, p. 470) was universal in Roman times, as we know from the Aulularia of Plautus, the plot of which turns on this practice,^[2707] and from the numerous finds of coins in jars in our own day. None of these have any ornamentation; they have been found in Germany, and there is a small specimen in the British Museum from Lincoln,^[2708] of spherical form with a knob at the top. Aubrey records the finding of a similar one in North Wiltshire.^[2709] These appear to be of very late date.

The next two types are of much greater interest, not only from their ornamentation, but from their form and the inscriptions which they bear. In the one the box takes a flat circular form, closely resembling the body of a lamp (the shape is that of Fig. 207), with a design similarly placed in a medallion. One actually has a figure of Victory with a shield, which reproduces the type of the New Year lamps described on page 413 (B.M. No. 309), and has a similar inscription.^[2710] It may be supposed that these boxes were carried round on New Year’s Day to solicit contributions, just as is done (says Herr Graeven) by boys in Rome at the present time. Others have figures of Fortune and Hermes in a shrine,^[2711] the latter deity being of course specially associated with money-making. These two examples have their respective makers’ names on the back, C IVN BIT and PALLADI, names which are also found on Roman lamps,^[2712] another detail which shows the close connection between these two classes of objects.

The last type to be described is shaped like a bee-hive, or, as in Fig. 200, like a circular temple, forms which were found convenient for the then favourite design of a deity in a shrine. Among the examples quoted by Graeven^[2713] is one of the latter shape with Fortune (Fig. 200), now in the Bibliothèque Nationale. Of the bee-hive form three may be mentioned as presenting interesting features. One with Hermes in a shrine has the maker’s name, PAS AVGV, which also occurs on lamps^[2714]; another, found on the Aventine, and now at Gotha,^[2715] has on the front the figure of a victorious charioteer, on the reverse a slit for the coins, and the maker’s name, AEL MAX. D’Agincourt suggested that this type of box was carried about by victors in the games to receive donations. Lastly, there is one recorded to have been found in the Baths of Titus in 1812, but now lost, which contained coins of Trajan, and was inscribed FISCI IVDAICI CALUMNIA SVBLATA. The evidence points to the dating of these two classes in the first century of the Empire, or slightly later.

Terracotta moulds for false or debased coins of the Imperial period have frequently been discovered in different parts of the Empire.^[2716] None, indeed, have come to light in Italy, but they occur in Egypt, Tunis, France, on the Rhine, in Switzerland, Lower Austria, and Britain. They were first noted by A. le Pois in 1579 at Fourvières, where moulds were found of coins of Septimius Severus and his successors. In 1697 and 1706 more of the same period, of local clay, were found at Lingwell Gate, near Wakefield,^[2717] in 1704 at Lyons, and in 1764 at Augst, near Basle. In 1829 and 1830 further finds were made at Wakefield, and again in 1869 at Duston, Northants.^[2718] Numbers have been noted from time to time in the museums of France and the Rhenish provinces, the most interesting find being that made in 1829-30 at Damery, near Épernay, in the Department of Marne. In 1859 a find of 130 moulds contained in a jug was made at Bernard; they appear to have been hastily placed there and left by forgers. At Bordeaux in 1884 finds were made in the ruins of a pottery, and others more recently at Autun and La Coulouche. In 1899 thirty-four moulds were found at Susa in Tunis. The British Museum has a collection of moulds of denarii from Egypt, mostly found at Crocodilopolis (Arsinoe) in the Fayûm; they are of a deep brick-red local clay, but a great number are burnt black.

Nearly all these moulds fall between the reigns of Septimius Severus and Diocletian, but some of those at Bernard go back as far as Trajan, and there are isolated instances of coins of Domitian at one end, of Constantius II. and Julia Mamaea at the other. Caracalla and Elagabalus are frequently represented, and those in the British Museum include Albinus, Crispus, Constantine, Galerius, Licinius, and Macrinus. The Damery find included thirty-nine moulds, comprising types of the coins of Caracalla, the elder Philip, and Postumus; 2,000 pieces of base silver coin, chiefly of Postumus; 3,900 bronzes of Constans I. and Constantius, all evidently made together; chisels and remains of other tools, and groups of moulds still containing the metal, and also lumps of metal which had overflowed from the moulds.

The way in which these moulds were used is as follows. The complete mould was composed of two shallow round boxes with hollow impressions respectively of the obverse and reverse, obtained by impressing the designs from genuine coins into the soft clay. The depth of the hollow was so calculated that when the two were placed together the space represented the required thickness. To cast the coins, a number of these moulds were placed one on the other, and luted with clay to prevent the liquid metal from escaping between the two pieces of each mould; down the side of the column formed by the pile of moulds a hollow cutting was made, at the base of which holes were pierced corresponding to the cavities where the metal was to enter. The metal was then poured into the hollow, and ran in through the holes as required.^[2719] Sometimes the columns were joined in groups of three 2529image for which a single column served; of this there is an example at Damery, where each rouleau contained a dozen moulds (thirteen discs). In the Cabinet des Médailles at Paris there is an example of one of these rouleaux of moulds, found at Lyons in 1704 (Fig. 201),^[2720] with the basin in which they were placed for the casting. At Susa the moulds were fitted slantwise into a bronze tube.

It is not absolutely certain whether these moulds were all used for fraudulent purposes by forgers; the find at Damery, for instance, was made on the site of Bibe, an important station on the road from Rheims to Beauvais, which would be too prominent a place for forgers to have selected. It is much more likely that in such a case they were used to make coins of inferior alloy, perhaps in some instances for the issues of usurpers who, being at a considerable distance from the capital, were unable to fill their military chests except with hastily cast coins. The distant parts of the Empire in which these moulds are found lend some colour to this theory. It will also be remembered that they mostly date from the time when a debased coinage was current throughout the Empire, beginning with the reign of Septimius Severus; this was put an end to by Diocletian in 297. We may therefore suppose that they represent, so to speak, officially recognised forgeries, emanating from a kind of local mint for producing coins hastily for provincial use. Hence the rapid spread of base money in the third century, which was not only forced upon the State, but was also readily taken advantage of by forgers.