Chapter 11:
Portable Works of Art



Introduction


Although portable works of art frequently echo or transmute motifs and styles from larger pieces, when of sufficient quality to offer lessons to would-be imitators, the art historian avidly searching for sources has usually ignored all except coins and ivories, with the result that there is little comprehensive information on how they were used. Toby Yuen, however, has started to redress the balance with a detailed examination of their effect on the art of Giulio Romano, Giovanni da Udine and Raphael (Yuen 1979). Although this book is not about the influence of antiquities on contemporary art, it bears stating that the mediaeval propensity for producing its own small works to the nearly total exclusion of more monumental pieces makes the imitation of small-scale antique exemplars as part of that process the more likely: Adhémar, for example (1939, 86), sees the church treasuries which received antique works from the rich Gallo-Roman families of France as no less than `veritable museums of Gallo-Roman art'. It follows, therefore, that such work may be of prime importance for revivals of antique art, most of which would be created for or in association with the Church.  

Most small objects were no doubt found singly and casually, but some spectacular finds helped encourage the belief that antiquities meant treasure or demons, if we read Benvenuto Cellini on his adventure in the Colosseum, or the tales about  Virgil the Magician - for, as Oldoni says, mediaeval romance literature `is punctuated by grottoes, machines, statues and spells' (Oldoni 1977/80, 2.605ff.). Many tales freely combine the two elements of treasure and magic, as in that of the famous statue in the Campo Marzio in Rome with HIC PERCUTE on the head: an ingenious interpretation of the inscription led to the discovery of an underground cave, with walls and ceiling of gold - and treasure. The tale, which appears in William of Malmesbury, was transposed to other localities such as Apulia, where no less a hero than Robert Guiscard performed the deed (ibid., 2.548ff.). Thus does the basest of all motivations for the exploration of antique sites blend with the more pressing need for building material, during the search for which it is logical to assume that most finds of treasure were made. It is easy to see why the survival rate of such objects is near to zero: when not fragile, like pottery, or ornamental, like gems, most finds were melted down for their bullion value, converted into more up-to-date designs, or broken down into their constituent saleable parts.



Manuscripts


With sarcophagi or `idols', there was good reason for the Middle Ages to preserve the one and destroy the other. But with manuscripts, the easiest thing to prove about their survival was that it was haphazard, as a study of any textual tradition will show (e.g. Billanovich 1951, 149ff., 199ff.): `how much do we know,' writes Pächt (1943, 64), `of the survival of classical manuscripts in late mediaeval libraries? Indeed, the darkest mystery surrounds the fate of pagan manuscripts between the demise of the public and private libraries of the antique world and the renewal of interest in their contents several centuries later; this is reflected in the fact that so few MSS were copied between 550 and 750, and many parchments had their writings washed off for re-use. The Carolingian revival was the first (excluding the English and Irish outposts) to interest itself in classical literature, and therefore ancient MSS were once more copied.   

Furthermore, the suggested rationale for the copying of manuscripts leaves much to be desired - namely that MSS were copied to preserve the text (and illustrations), because the originals were too fragile or decayed to survive. We have the copies, but why have so few of the originals survived, even as fragments? Or were many of them lost in post-mediaeval times? Can the acquaintance with an isolated piece of illumination have sufficed to convey to the northern painter the artistic secrets of a distant epoch and a foreign milieu?' The question is important, given the belief of some that manuscripts played a crucial part in the development of later mediaeval styles; thus Emile Mâle's influential L'Art réligieux du 12e siècle en France begins with a chapter entitled `The birth of monumental sculpture: influence of manuscripts' and with the gloss that `sculpture disappears at the end of classical antiquity ... Under the influence of the miniature, sculpture reappears in southern France in the late eleventh century.' Mâle is referring to mediaeval manuscripts, of course, and many of his comparisons are convincing; but he leaves no place for considering the possible influence of actual antique artefacts in the development of the new style. Given his belief that Carolingian ivories are themselves influenced by manuscripts, the omission of a discussion of surviving classical manuscripts is unfortunate.


Destructions


What were the conditions in scriptoria relevant to the preservation and transmission of ancient manuscripts? The copying of manuscripts was essential for the preservation of the text when the material used was as fragile as papyrus and parchment. The latter was particularly vulnerable to damp and insect attack, because it was virtually untreated leather and most nutritious; once damp, it would have been very difficult to dry with the writing preserved. Horror stories regarding the fragility of manuscripts are therefore common (Lesne 1938, 28). Parchment was also expensive, and skins were therefore liable to be scraped clean for re-use, so that old codices in script which was either old-fashioned or difficult to read would be broken up (an example in Billanovich 1951, 184). However, there is no evidence that re-use of skins was a deliberate attempt to blot out the pagan past, as has sometimes been surmised; and perhaps this whole area is of little consequence for the preservation of classical works, for Lesne (1937, 475) maintains that there are no cases in which such palimpsests conserve a complete work, and also that the majority are Christian texts. Simple destruction of manuscripts was also easy, either through neglect or design: we are told that the great library of Alexandria went in 638 to feed the fires of the baths of that city, on the excuse that `if these writings of the Greeks agree with the book of God, they are useless and need not be preserved: if they disagree, they are pernicious and ought to be destroyed' (cited by Gibbon 51).


Availability


The extent to which antique manuscripts (particularly illustrated ones) were available in the Middle Ages is disputed - a natural reaction given the small size of most known collections (Lesne 1938, 767ff.), but unadventurous given that most of the classical texts surviving today do so thanks to mediaeval copies. One centre where the copying of antique works of art, or their illustration, did take place more or less without a break is Constantinople, which `must still have been a treasure house of classical art objects, including precious manuscripts, before its destruction in the year 1204 AD' (Weitzmann 1949, 160). But it is not considered that a similar quantity of classical sources was available in the West: thus Rosenthal will admit a direct late antique source only for Terence and Aratus manuscripts, noting for the rest that classical elements were `imitated' indirectly, and `only after having been transformed during the late Antique and early mediaeval centuries. Classical elements, therefore, were not `revived'; rather they entered Carolingian art as they had `survived' in their transmuted forms' (1953, 85). He does not discuss later works, such as the Vivian and Moutier-Grandval Bibles, which must surely have been copied directly from fifth-century manuscripts (Beckwith 1964, 52ff.). Indeed, for many of the sumptuous manuscripts with purple leaves and gold and silver capitals, it is surely not unreasonable to suggest either fifth- or sixth-century manuscripts as direct exemplars, or Roman lapidary capitals (cf. McGurk 1962, 23, 30f.). Certain texts do indeed survive in large numbers of mediaeval manuscripts, reflecting the interest they aroused: a notable case is that of Vitruvius - surely one element in the Carolingian revival of antique forms, to be considered along with survival of actual buildings (Frankl 1960, 86ff. and n. 16). Did the MSS of this work available to the Carolingians still have their illustrations?  

From reflections in later works, then, it can be inferred that some early illustrated manuscripts were indeed available to scriptoria during the Middle Ages (cf. the discussions in P[um] acht 1943 and Bloch 1950). The majority must have survived in monastic or at least religious collections (cf. Lesne 1938, 31ff. for French examples): we have no record of royal or civic collections in the West from the fall of the Empire to the time of Charlemagne. For example, it is well known that the Vienna Genesis (or the Cotton variant) was used for the Trecento mosaics in S. Mark's, Venice, and inspired the creation of further manuscripts (Buchthal 1971, 47ff., 63f.; Pächt 1943, 65f.). Again, the Calendar of 354 (Vat. Barb. Lat. 2154) is one of a group of copies based on a Carolingian intermediary; and the fourteenth century Egerton Genesis `must have preserved to a high degree the late antique style of the original if indeed it was not itself the original' (Pächt 1943, 63, and 64ff. for general remarks). Another example is the Paris BN Manuscript of the third decade of Livy, the influence of which can be documented through the copies which were made, and the travels of which from Avellino to Corbie and then Tours are thereby reconstructed (Lesne 1937, 483f.). Proof of availability is only firm in such cases as this, where we still possess both `original' and `copy'. In the majority, however, the classical originals are  lost: with many Carolingian manuscripts, the classical forms and (sometimes) classical subject matter point to sources of the fourth and fifth centuries being available in the scriptoria: thus Delogu (1977, 66, n. 196) sees the illustrations in the Utrecht Psalter as a reflection of the antiquarianism and interest in columns and capitals manifested at Aix. The same epithet of `antiquarian' could, of course, be applied to the very paleography of many Carolingian manuscripts, which imitate the epigraphic style of the second and third centuries - rather than earlier manuscript styles (Deschamps 1929, 14).


Manuscripts as Models


The reflections of ancient MSS in later productions are indications that several at least were `re-used' by being copied or imitated in the Middle Ages. Unfortunately, this is an area where archaeology can be of little help, for it cannot advise on availability. Nor should we assume that any MSS available would automatically be used as models - as Beck reminds us (1980, 33) by recounting the fiasco of Cardinal Bessarion's gift of 500 manuscripts to the State of Venice in 1468. These were made available only four times between 1472 and 1476, and put into packing cases in 1485; subsequently it was decreed that no codex could be borrowed without the consent of three-quarters of the senators! `It is hard to comprehend', remarks Beck, `that all of this took place in a period we call the Renaissance.' Why do we not possess a more complete picture of the fate of antique manuscripts in mediaeval scriptoria? Because, replies Lesne (1937, 475f.), monasteries lacked the means, the mission and the need to conserve pagan texts; and our view must be a partial one because the monastic schools did not adopt the same range of interests as the pagan schools.  

If, as at Tours, there were monks who were lapidary masons as well as copyists of manuscripts (Adhémar 1939, 149, n. 4), then perhaps it was not unusual to take inspiration from different media. Certainly, the matter of the transmission of classical forms is problematic. Where, for example, did illuminators find those complicated acanthus capitals with which so many manuscripts are decorated? Jalabert (1965, 37f.) suggests that some would be pure invention, but never considers that full-size antique exemplars might have been available when the books were illuminated. Similarly, she sees some Romanesque forms as inspired by manuscripts (ibid., 45). The matter is complicated by the manifest connections between manuscript and full-scale representations in the ancient world: compare, for example, the similarities adduced by White (1956, 92f., and pl. 29b and c) between Cavallini's work in S. M. in Trastevere, the Vatican Virgil and the probable appearance of the large fifth-century frescoes in S. Paolo fuori le Mura. Indeed, so antiquarian were some productions that it is sometimes difficult to decide whether certain manuscripts are late antique `originals' or Carolingian `copies'; the latter view assumes the availability of late antique originals to imitate. In many cases, of course, the availability of exemplars cannot be in doubt - as in Köhler's demonstration of the sixth-century model behind a mere fragment of work from the Ada School (Köhler 1952), or Pächt's remarks on the use by the illustrators of Frederick II and Manfred of ancient models, and not corrupted mediaeval copies (1950, 24). It is not surprising that the centres which prized `genuine' manuscripts, and imitated them, were precisely those in which antiquity in other materials was also highly regarded. Such an interest in `authenticity' meant, in the case of the Carolingians, a focus on the city of Rome and imitation of some of her celebrated features: for those manuscripts containing vine columns (deriving from the columns in Old S. Peter's), the source was probably sixth century Roman manuscripts (Rosenbaum 1955, 1ff.). A parallel example is the Lindisfarne Gospels, which have been shown by Saxl (1943, 17f.) to be connected with surviving Roman sculpture as well as with imported manuscripts and other small-scale work: their artist `must have become aware of the relation between these imported documents and the old Roman stone work still everywhere visible, and probably in particular profusion in Northumbria. It does not seem illogical to assume that his inherited instincts and the Romano-British models still available led him to evolve the new style of the Lindisfarne Gospels.' A third example shows the circularity of some of the arguments on the survival of ancient manuscripts: Werckmeister (1976, 543ff.), investigating the possible sources for the Bayeux Tapestry, considers both Trajan's Column and the possible survival of a picture roll of the Joshua Roll type, itself a work of the mid-tenth century. However, it is impossible to decide whether the Joshua Roll is the last of an antique tradition or, rather, a new invention inspired by viewing the columns of Constantinople (ibid., 545 and nn.).


Coins and medals


Antique coins have always been available in varying quantities, as laws on treasure trove reflect (Morrisson 1981 for the East; Hill 1933 for the West), and they have always been imitated for their iconography, motifs and even lettering (on which cf. Morison's suggestion that the wedge serif in Irish and other manuscripts could well come from Imperial coins and medals: Morison 1972, 155ff.). And because the most important characteristic of coinage is to reflect stability in value, the mediaeval world had much knowledge of `external' coins through trade, so that we not only find Arabic coin types imitated in Europe, but also Roman and Byzantine types imitated in the Arab world (Al-Ush 1971).  

Most antique coins were surely found in hoards, and there are some curious examples of re-use (rather than imitation) of antique coins in the mediaeval period, ranging from a whole series of first century coins overstruck and re-issued in the late fifth century, and Roman denarii turning up in tenth- and eleventh-century Germanic hoards, to Roman coins still in circulation in the France of Napoleon III, and Roman, Byzantine and Moslem coins in use in Asia Minor later in the same century (Morrisson 1983, especially n. 5). There are even cases of `official' re-use, as when coins of Domitian and Theodosius were found countermarked, dated 1636, and in use as contemporary Spanish coinage (Merson 1975). Perhaps we may conclude that, where hoards contain some coins much earlier than the majority, these were still in use as currency.


Coins and political status


An important reason for the continuing popularity of antique coinage was the power of the messages it could bear. Because the Christian iconography of rule to be seen on coins evolved from pagan ideas, rather than replacing them, and the pictorial répertoire was reduced rather than extended, there was no viable alternative to a basically pagan iconography. Thus although some abstractions such as Spes and Providentia were not used in coinage after the fourth century, personifications such as Victoria, Roma and Constantinopolis did survive (Grierson 1976); by the fifth century, we find the Cross supported by Victory - a hybrid which was amended rather than replaced, for Victory became an angel, or S. Michael. In the East, Victory disappeared by 629; in the West, the Franks used it up to c. 575, the Visigoths to c. 580, and the Lombards in Italy to c. 690 (ibid., 622ff.). Such an iconographic tradition was also a strong argument in favour of imitation: the Lombards imitated Roman coins in both legend and iconography, just as the Celts had imitated Greek ones (Belloni 1980); and the Carolingians Charlemagne, Louis the Pious and Lothair I all had themselves portrayed on some of their coinage after the fashion of Roman pieces of the Empire (Colin 1947, 91, n. 1) - surely a political rather than simply an aesthetic decision. Indeed, the imitation of Roman coins by barbarians begins during the Empire itself (Belloni 1981), and was widespread thereafter (Baum 1937, 54ff. and figs. 26-43): bracteates, as used for example by the Longobards, turned classical coin-images from money into jewellery (Werner 1972).  

The mediaeval regard for antique coins is best known from Charlemagne's famous pastiche, so frequently used to prove just how antiquarian he was. However, this example is misleading, for not only is its style exceptional in the coinage of his reign, but it may not have gone into circulation, being used simply as a medal for presentation to friends - a practice for which there are parallels for other `special' issues (Grierson 1954, 1062f.); perhaps Einhard reminded his master that Augustus himself had done likewise (Suet. Aug. 75: `he would now give gifts of clothing or gold and silver; again coins of every device, including old pieces of the kings and foreign money'). Nor was Charlemagne's coin the first to imitate the antique, for a silver denarius of Ethelbert, king of East Anglia, of about 790, shows the wolf suckling Romulus and Remus - a use equally as `political' as that of ancient marble or statues (the same might be said for the Early Christian manner of the Salerno ivories: Bergman 1974, 172f.). But whoever began the fashion, it was adopted by some (but not all) later rulers. Thus we find in Pavia a lead seal with the head of Lothair `all' antico', with laurel crown, and the legend DN HLOTHARIUS AUGUSTUS, with GLOR EGNI on the reverse, framed by a laurel crown with a fillet. The gold solidus of Louis the Pious was much imitated (Grierson 1951). The coins of Gisulf I (946-77) are antiquarian as well, for the reverses have Roman-type perspective scenes (e.g. a follis with a view of Salerno, and the legend OPVLENTIA SALERNV): what is more, the coins are individual to his reign, being very different from those of his predecessors and his immediate successors. But Gisulf II (1052-77), the last Longobard prince of Salerno, imitated Roman coins so precisely for some of his money supply that the whole point would have been lost had not originals been available for comparison (discussion in Delogu 1977, 169ff.; Grierson 1956).  

In no case can we do more than guess what the impact such implied `messages' on coins might have been: but this is also the case with their sources for, in a rich discussion of Roman attitudes to their coin types, Michael Crawford (1983, 53) believes that `we cannot conclude ... that the Greeks and Romans often noticed the programmatic coin types with which they were confronted'. But the growth in popularity of such imitations during the later Middle Ages may indicate a growing availability of antique coins and medals. Frederick II's famous augustales were imitated from a specific issue of Augustus, and were actually put in circulation; as Wentzel has shown (1952), they were by no means the only example of antique inspiration in the coinage of that period. Even earlier, we find William II of Sicily issuing (from about 1184) a trifollaro with a lion-mask, which is taken directly from the ancient coinage of the area, particularly Rhegium and Messina. Such imitations may indicate a political identification with the history of the area, and a desire to demonstrate continuity; and, as the author who has recently published the detailed comparisons points out, `we must recognise and now accept evidence for a still earlier date in mediaeval history of the collecting and study of ancient coins - almost certainly an aspect of the collection of other classical art and artefacts' (Breckenridge 1976, 282). Further research (in areas like the coinage and seals of the Staufer) may well uncover more examples of direct imitation of the antique.  

William must have found the Greek coins he imitated on Sicily itself, but there are examples of coins from Greece and its islands probably available in Western Europe in the thirteenth century. The most curious were known as the thirty pieces of silver paid to Judas, which have been studied by De Mely (1899): one of these is a medallion of Syracuse, surrounded by a gold disc on which is engraved QUIA PRETIUM SANGUINIS EST (Babelon 1901, 80); others are drachmae of Rhodes bearing the head of Helios and, on the reverse, the legend RODION - perhaps taken as a reference to Herod. It would be idle to try to  decide whether these coins reached Europe after the fall of Constantinople, or through the Knights of Rhodes, who are connected with at least three of the coins (De Mely 1899, 503, 505f.) - especially when it is clear that the coins of Rhodes were famed for their beauty in Antiquity, and eagerly collected even then (Babelon 1901, 68; 79, n. 3). Robert de Clari is recorded as bringing `numismata' back from Constantinople (Riant 1875, 206), and the Corbie reliquary he also brought back contained a Roman medallion (Babelon 1901, 81), but we have no further details. Grierson (1955) suggests that the Rhodian coins were found as a hoard on that island in the early fifteenth century.


Coins which survived as jewellery


Some coins did survive by being converted from money into ornaments and, although silver ones were treated this way as early as the first century, the practice generally used gold. The choice of that metal for such jewellery (mainly brooches and pendants) is easily explained by its intrinsic value: many Romanised barbarians preferred to trust metal, and perhaps the Imperial images the coins bore were simply a bonus. Such conversions became popular from the third century, and medallions were used as well when they came into vogue (Zadoks-Jitta 1953). Coin pendants and rings seem to have been especially popular in the seventh century, as examples found in Merovingian graves testify (ibid., 456). Of course, more humble coins were frequently used as ornaments as well, perhaps having the power of amulets: and the pagan Anglo-Saxons sometimes had Roman coins buried with them (Meaney 1981, 213ff.). Nor was this rare in Italy: compare the medallion necklace from Lodi, now in Pavia, where a gold solidus of Theodosius I (379-95) was set into its mount not later than the end of that century. The practice continued well into the ninth century (Aix la Chapelle 1965, 168ff.).  

The popularity of such `antiquarian' decorations, and consciousness of their aesthetic value, can be judged from the frequent use of imitated - rather than genuine - Roman coins, perhaps indicating that even by Merovingian times (but only in some areas?) antique coins were not only scarce, but were prized more for what they represented than for their metal: thus a coin in the Hague (ibid., 457) is pseudo-Roman and bears the Roman eagle with IOHANNIS - an example of the Christianisation often seen in Bishops' rings which are made from gems. Indeed, this supposed scarcity is supported by Zadoks-Jitta's description (ibid., 459) of a presumably seventh century fibula with a much-worn aureus of Antoninus Pius, which the proud owner converted into a pendant when the pin from the brooch broke off. And the suggestion that antique artefacts were prized for what they represented is seen in Bruce-Mitford's argument (1961) that the Anglo-Saxons had an intelligent attitude toward Roman culture, reviving both art and techniques `not in the confusion of invasion and settlement and the breakdown of the Roman system, but when the Saxon Kingdoms were consolidated and achieving wealth and status'. Thus Roman designs were sometimes used not simply as trinkets, but for conscious political or social ends, as we have seen. The argument is strongly supported by the attention given to gems, which had no bullion value: these, not always of late Imperial date, were widely employed in Migration Period brooches, and have been found in fifth- to seventh-century graves in Britain and on the Continent (Henig 1974, 162ff.; 196, and cats 111, 518).



Collecting


 

Turning coins into jewellery is the proof that the aesthetic value of ancient coins was recognised early - linked, probably, to both metal value and political overtones. Coins also had historical value, for both their iconography and inscriptions, and it would be interesting to know whether those states which proudly displayed full-sized inscriptions also collected antique coins which turned up on their territory. Unfortunately, the earliest collection of antique coins of which we know is that of Petrarch (Weiss 1969, 37), who purchased medals brought to him in Rome by peasants, and deciphered their inscriptions with emotion (Babelon 1901, 83). But Petrarch was not alone: Giovanni Mansionario drew several coins about 1320, a splendid series appear in a Fermo manuscript of about 1350 (cf. Degenhart 1968ff., 2, cat. 640), and the 1351 inventory of the treasures of Doge Marino Faliero included a box `with fifty coins of marvellous antiquity'. There was therefore a market for such coins by the mid-fourteenth century, as is confirmed by the acquisition by Oliviero Forzetta in 1335 of fifty medals in Venice; he bought them from a Magister Simeon, and they must have been antique (Babelon 1901, 82f.). Roman coins would of course be readily available on Italian soil, but the collecting of Greek coins from the fifteenth century (Babelon 1901, 85ff.) was probably possible only through new imports - largely, perhaps, from Venetian possessions or trading posts; and this was, indeed, the path whereby one of the Rhodian coins reached Europe, being given to Cosimo de' Medici by the Greek Patriarch at the Council of Florence. However, with the fall of Constantinople and the ever-increasing grip of the Turks, the importation of spolia of any kind from the Eastern parts became much less easy than during those centuries when the West traded freely with the East, and ruled over parts of it.  

By the fourteenth century, then, Roman coins and medals were not only prized and imitated, but probably common, if we are guided by their reflection in contemporary art work, and by the inspiration they provided for coins and seals (Wentzel 1952; 1955). The increased pace of urban development, as well as purely intellectual reasons, may have played some part in the process. Francesco Novello da Carrara (reigned 1390-1405) was perhaps the first Western ruler to use more than the occasional antique image on his coins and, in so doing, he began a fashion (Weiss 1969, 53f.).



Coins sought as bullion


The attitude that works of art in valuable materials were `bank accounts', well known from Louis XIV and his famous suite of silver furniture, has never disappeared: it was not unusual in the Middle Ages for institutions to maintain wealth in ostentatious form (conspicuously as reliquaries or other sacred ornaments) which were melted down when pressing occasion arose (Sumption 1975, 155f.). Money had spread throughout the Roman Empire in trade - but also as tribute, if we may judge from a decree of 374 forbidding the furnishing of gold to barbarians, who used Roman gold coins and Roman jewellery as bullion (Belloni 1980). This may be some reflection of the barbarian fashion for burying valuables with the dead (Bloch 1933, 9), as well as an early hint of the impoverishment of supplies of precious metal which would so affect the Middle Ages; for, as Bloch reminds us (1933, 11), these dwindled thanks partly to the import of luxury goods from the East. There were no gold mines in Western Europe in the Middle Ages, and the shortage of gold (which had virtually disappeared from Germanic Europe by the eighth century) surely contributed to the thesaurisation of treasure, including perhaps that of Alaric (Dochaerd 1949, 35f.), although it has been maintained (Lesne 1936, 238f.) that some pieces survived for centuries because of their beauty.  

Evidence from hoards indicates that an assiduous mediaeval grave-robber could sometimes collect interesting material. One hoard at Corinth had Roman coins from the 340s to the 470s, Byzantine coins of the sixth century, coins from Sikyon (fourth to second centuries BC), and coins of Philip II of Macedon (Dengate 1981, 153ff.). Dengate lists several other late hoards containing Greek coins, which are not rare (157, n. 18), and surmises that the collector of the Corinth hoard simply scavenged around the city and its graves, for he also collected droplets of bronze from the earlier bronzeworking areas. Indeed, some of his worn fourth-century coins may have been collected directly from circulation - that is, having been in use for about two centuries (ibid., n. 21 for list of later hoards containing them). However, most antique coins were surely appraised simply for their metal value, without any consideration of their iconography or `artistic' interest. Digs at, for example, Silchester and Leicester, have produced residues from the melting down of Roman coins. Analysis of those from the Forum at Leicester has found them to consist of silver, copper, lead sulphide and calcium phosphate: the absence of tin means that base silver coins, not bronze ones, were destroyed for their metal. This also happened to a large treasure found before the year 1004 during reconstruction of the cathedral of Orleans, by the masons who were checking the firmness of the earth for the new foundations (Mortet 1911, 3f.).  

Indeed, since the original point of metal coinage was its bullion value, it is not hard to see why so much extraneous coinage was melted down, or (supposedly) imported from the Arab world (critique in Grierson 1954): Henry III of England bought large quantities of Muslim gold coins - not because he needed them in their bought form, but because it was a convenient way of obtaining the metal. Nor, indeed, was there any lack of extra-European or even non-Christian money circulating in the Middle Ages, including Saracen, Syrian and Byzantine coins; but whether Roman money survived in use for long after the fall of the Empire is a moot point (Bloch 1933, 13, and n. 5). It would, however, be interesting to know whether antique coinage had any recognised barter value in the Middle Ages.


Overstrikes


Overstriking is a footnote to the use of ancient coins as bullion. The practice was not very rare, probably because the quality of ancient metals was recognised as superior to that of modern efforts. One example is a proof coin from the reign of Francesco da Carrara, where an actual reverse from the reign of (probably) Antoninus Pius, has had the bust of Francesco looking like Vitellius added to the obverse: his mint had evidently filed the antique head off a Roman coin, and then struck that of the modern ruler in its place (Cessi 1965). A century later, the forger Cavino may well have done the same, for some of his medals are struck in yellow metal, perhaps worn Roman sestertii: `thus, the variations in thickness and diameters of such medals might be explained by how worn the coin was and how much force was used to strike the coin' (Klawans 1977, 11-12).  

Overstriking continued in later centuries, as Louis Savot writes in his Discours sur les médailles antiques, Paris 1627, 308f.): `For because we today have lost the means of making the beautiful yellow metal of the Ancients, those who wish to imitate it take an old medal of this metal, worn smooth, which they strike with a new die made in imitation of the antique, so that thereby they convince the more easily that the medal is antique.' Such care was exceptional, and overstrikes which have not obliterated the image underneath are frequent, as for example the large number of tenth-century Byzantine coins thus treated found in the excavations of Corinth; in mediaeval Salerno they are a valuable guide to relative chronology, because the original coins are modern, not antique (Grierson 1956, 40f.). Overstriking was clearly practised for the sake of thrift - but also, perhaps, because the ancients had done likewise (Babelon 1901, 938ff.).



Gems and work in pietra dura


One of the oldest art-forms with a continuous history (Zazoff 1983), gems may well have been the first small artefacts to be collected: Julius Caesar gave his collection to the Temple of Venus Genetrix, because she was the ancestress of his gens; and Marcellus, the nephew of Augustus, gave his collection to the Temple of Apollo on the Palatine (Wace 1949, 28) - proof, perhaps, of a vogue at the beginning of our era. All the peoples of mediaeval Europe seem to have treasured antique gems, and Cagiano de Azevedo (1970, 250f.) reminds us that the mediaeval interest in splendour and glitter is the legacy of late Antiquity. Thus, unlike most categories of antiquities, gems were indeed collected during the Middle Ages, and much prized: for many antique pieces are preserved in mediaeval settings and, indeed, gem-making continued during the Middle Ages, with Christian motifs introduced and, sometimes, pagan gems altered (Wentzel 1963; Zazoff 1983, 374-86). The availability to the Middle Ages in miniature of the main motifs of classical art is therefore certain (e.g. Horster 1970). Gems also had plenty of secular uses: the Carolingians, for example, made seal-stones of antique intaglios, and other seals with imitations of such antiquities (Colin 1947, 91f.). Thereafter the practice became common, as with the seal impression of Richard, son of Drogo, on a diploma of 1090, made by an antique intaglio of two men facing, the one standing hand on hip, the other seated and holding a shield; perhaps this example reflects how readily the Norman conquerors of Southern Italy took up indigenous fashions.  

Because gems are among the least destructible of antique artefacts, `there is every indication that some of the glyptic masterpieces which have come down to us were never buried but have served to delight their owners through twenty centuries' (Henig 1974, 21); certainly, they have been found all over the Roman world (Zazoff 1983, 261ff., 307ff.). Perhaps most remained unburied because of the burial practices of the Christians who, unless they were Merovingian, tended not to indulge in grave-goods. It follows that, since pagan jewellery settings were apt to be melted down, converted to a religious use, or perhaps modernised (Lesne 1936, 175ff.), what gems were available during the Middle Ages came either from non-Christian burials (such as the Frankish disc-fibula in Darmstadt: Steingräber 1957, 17f., which includes a gem with a Medusa head), from a few burials in Britain (Henig 1974, 65f.; Meaney 1981, 228ff.), or were preserved, frequently by monarchs - because `jewels became part of the ceremonial of the revived king-worship of antiquity' (Steingräber 1957, 17). However, although some treasuries with antique gems survive (such as that at Aachen), later disturbances have dispersed most of them: we know from English records, for example, that both St Alban's and St Paul's possessed very rich jewellery collections - including, no doubt, antique pieces - which were lost at the Reformation (Henig 1974, 199). Part of such collections may have been made during the investigation of ancient sites, especially those with religious significance: compare the recent work in the Sulis spring at Bath, which has yielded no less than thirty-four (Henig 1974, 61f.). Indeed, gems cannot have been scarce in Britain, for they laid the foundations for some British coin-types, and were later used in seals (Henig 1974, 2.106ff. lists 34). An early description survives of one gem, presented to King Aethelred, which was later drawn by Matthew Paris (Dodwell 1982, 109 and pl. 25).  

Jewels also had a secure place in the trappings of Christian ritual, for `every precious stone was thy covering, the sardius, topaz and the diamond, the beryl, the onyx, and the jasper, the sapphire, the emerald, and the carbuncle, and gold' (Ezekiel 28.13) - a passage used by Abbot Suger illustrating (and justifying) his exposition of the riches of his church (Frankl 1960, 19). Indeed, the mediaeval propensity for viewing gems in terms of their mystical Christian associations (as in Alexander Minorita's Expositio in Apocalypsim ch. 4, in MGH Quellen zur Geistesgeschichte 1.51) explains how it was that classical gems - even those which could admit of no Christian explication - were given to religious foundations as ex votos (Baltrusaitis 1973, 51ff.). Occasionally a Christian interpretation is indeed possible if tendentious: it seems likely that commissioners and jewellers alike knew the general origin of the pieces they bought and used, but this did not prevent some preposterous misinterpretations (Müntz 1887, 48) - any more than it did the use of apparently unsuitable pagan gems as clerical seal-stones (Adhémar 1939, 106ff.; Baltrusaitis 1973, 287, n. 86; 287, n. 95). Nor was it unusual to mount pagan cameos and gems in Christian contexts, two famous examples being those set in the crown of the Miraculous Virgin of Notre Dame du Puy (Bachelier 1956), and the surviving sections of the Ecrin de Charlemagne in the Louvre, one of which has an antique cameo in a setting of about 860-70 (Steingräber 1957, 19ff., and fig. 3); indeed, one of the most splendid collections of gems was that at Saint Denis (Montesquiou-Fezensac 1975), and one of the finest objects therein (apart from the Ecrin ) was the Tabernacle des Corps Saints built, perhaps predictably, by Abbot Suger.  

One popular Christian destination for gems was as decorations in sumptuous covers to religious books, frequently those used in the services. The practice began early: S. Augustine inveighs in a letter (PL 22.418) against women whose books are in gold on purple vellum, and their covers decorated with gems. Beginning under the Empire for secular books, it was taken over, apparently exclusively, by the Church (cf. Ezekiel 28.13, quoted above). Are there any examples of the covering to secular texts from the Middle Ages decorated with engraved gems?  

Antique gems and cameos, like ivories, were enthusiastically imitated throughout the Middle Ages (Snijder 1933; Wentzel 1962) - a proof of their continuing popularity to place beside that provided by their continuing use in liturgical contexts well into the fifteenth century (Montesquiou-Fezensac 1975, 143ff., for a gift by the Duc de Berry, a great collector in the medium). Intaglio gems had natural scarcity because they were not made in large quantities after about 200 AD, and the scarcity of originals must have been partly filled by imitations: the Longobards, who also reset originals, such as the fibula from Benevento in the Ashmolean Museum, Oxford (Rotili 1975, 124f.), were making glass imitations of antique pieces by the early seventh century, if Snijder's dating of the Brescia Cross is accepted (Snijder 1933, 121; Wentzel 1963); and Leo IV certainly (847-55) decorated the high altar of S. Clemente with a crown embellished with glass gems (LP 2.125) - which helps explain Eraclius' recipe De gemmis quas de Romano vitro facere quaeris (1.iii, in Merrifield 1849, 1.183ff.). Indeed, as Wentzel (1952) and Sauerländer (1961), among others, have shown for the thirteenth century, gems were the inspiration for work in other media (to which Panofsky 1970, 63, n. 5 adds statuettes), just as antique coins were the inspiration for new seal-stones in the thirteenth and fourteenth centuries (Wentzel 1955). Indeed, Wentzel's examination of French mediaeval gems and seals (1953) leads him to conclude that a taste for the antique was highly fashionable in the thirteenth century, and their imitation a guide not only to this, but also to the scarcity and high cost of the originals. If Wentzel is correct (ibid., 342), then were antique gems responsible for the reintroduction of the profile bust into European art, about 1220 AD?  

The apparent increase in interest in gem stones and references to them (Faral 1967, 351ff.) in the later Middle Ages may have been helped by the Crusades: Geoffrey of Monmouth specifically mentions large quantities of gems `of such rarity that they are unknown in our lands'  as part of the booty of the First Crusade (Gesta Regum Anglorum 4.371), and the succeeding expeditions must have increased the numbers considerably; some of the imports provided inspiration for contemporary gem-cutters (Steingräber 1957, 33). Similarly, the most curious examples of gems treasured during the Middle Ages are surely those preserved in a pre-fourteenth century ebony and ivory box in the Treasury of the Capella Palatina at Palermo: attached labels identify them as a collection of `precious' stones brought back by a pilgrim from the Holy Land, with their provenance noted (`de sepulcro Ismael' on one, for example). The value of all of them was sentimental and religious - with the exception of a Mesopotamian cylinder seal of the third millennium BC, which is so worn that it may well have been hung around the neck as some kind of talisman; it was clearly not made in Palestine, so either it or the pilgrim must have travelled (Rocco 1981). The rest of the stones are, on our scale of values, worthless, but the power of association no doubt made them collectable: a famous parallel is that of the Holy Grail, convincingly suggested to have been a marble altar table, but proclaimed in Wolfram von Eschenbach to have been of precious stone. Probably Wolfram embroidered on the fact that several such early tables are in porphyry - a precious stone among marbles (Barb 1956, 49f.).  

The import of gems from the East, then, was probably not rare, for there are references to gifts to church treasuries from Palestinian, Arabic and Byzantine sources, and these may have included antiquities as well as contemporary work (Lesne 1936, 180, n. 2). Some of the uses of gems in Western ceremonial may, like fashions in marble, be inspired by Byzantine practice - and we should always keep in mind the scale of the Byzantine export trade of its own, sometimes antique-inspired, productions (Nickel 1978). The late tenth-century pectoral worn by the Empress Giselda at her wedding to Conrad II at Easter 1027 (Steingräber 1957, 22f., and fig. 8) contains several Roman intaglios, and the whole form derives directly from the dress of the Byzantine Emperors. The use of Byzantine gems was a distinct vogue in thirteenth century France where, in later settings, they appear as ecclesiastical clasps, and also as models for antiquarian sculptors (Coche de la Ferté 1966). Others probably arrived at various times as gifts from one ruler to another; this, for example, is Colin's explanation (1947, 107ff.) for the Cup of the Ptolemies now in the Cabinet des Médailles, given by Charles the Bald to Saint-Denis; and he makes powerful arguments for the onyx vase given to the English monarch Aethelstan in 927 (which has not survived) as contemporary, not antique. We have mediaeval accounts of other vases, too, which clearly were antique (Lesne 1936, 227f.), but confirmation is impossible because they have in most cases long since disappeared. Many surviving works in pietra dura were probably (like many gems) never buried - although some were surely taken from tombs. The Norman court of Sicily collected such work (and probably made new pieces: Giuliano 1980, 24f.); secular and religious treasuries (conspicuously that of S. Mark's, Venice) did likewise (survey in Gasparri 1979), large numbers probably arriving as a result of the 1204 sack of Constantinople.  

The use of antique gems as talismans was common, and one obvious reason for their popularity, for both designs and materials were believed to have magical or medicinal properties (cf. refs. in Maue 1982): one of the most famous of the antique gems in the fifteenth century, a chalcedony now lost, being rescued by Niccol[gr]o Niccoli from a child's neck. The lists of talismanic gems proclaim their magical properties: they frequently name the antique subject matter of gem types in particular stones, and their length reflects the large numbers of stones available (cf. Wright 1844, 449ff. for two lists of 27 and 22 items respectively). Another and parallel token of the wide availability of gems are the treatises on their medical and other properties, being imitated in English mediaeval art (Higgitt 1973, 12, and n. 2): the Gesta Abbatum of S. Albans (1.84) notes that the majority of the Abbey's cameos went in the early twelfth century to ornament the shrine of the saint - except for a cameo noted for the aid it gave in parturition. Such lists of the properties of gems provide tangential evidence that they were not only prized when found, but actively and assiduously sought, for the formula is generally If you should find a stone in which there is ... or In whatever stones you should find ....


.st Portable Works: Ivories

Ivories


Since prehistoric times, ivory has been a favourite medium for small three-dimensional figures and, especially, for relief work: it takes detail well, and can receive a high polish - characteristics which, together with its rarity, befitted it for the luxury market of containers and decorative panels. I know of no examples of the re-use of ivory statuettes in the Middle Ages; but plaques were prized, whether made up into caskets or used as Consular Diptychs - that is, bivalve tablets prepared by officials for presentation to the Emperor on the occasion of their investiture.  

Diptychs were were easily translated into a Christian context, for bishops replaced consuls, and religious scenes replaced pagan ones of the hunt or the circus games; and several notices survive of the conservation of pagan and Christian examples in church treasuries (Lesne 1936, 233ff.). For such diptychs could themselves be re-used - as for example the Anastasius leaf in the V and A, London, which has lists of names on the verso which are perhaps one or two centuries later than its creation in 517 AD; the Clementinus Diptych (Liverpool), which bears an inscription indicating re-use in Magna Graecia in the eighth century; or the Probianus Diptych (Berlin, Staatsbibliothek), which was used as covering to an eleventh-century manuscript (other examples in Adhémar 1939, 88f.). Indeed, decorated bookcovers have frequently preserved gems, ivories and metalwork from earlier periods (Steenbock 1965, passim). The fashion might have been imported from Constantinople - witness Halitgaire's gift to Charlemagne, after an embassy there, of `tabulas eburneas quibus libri cooperti' (Lesne 1936, 178, n. 6).

 

There are therefore several reasons for the prizing of antique ivories during the Middle Ages, ranging from their scarcity value and connections with the Empire, to their use, alongside  Christian counterparts, as exemplars for contemporary art work. In centuries when full-scale sculpture was rarely made (partly, at least, because it was not portable), ivories with an established liturgical function were popular enough to be imitated, sometimes so exactly that supposed copies are accepted as such only with difficulty. This is the case with several ivories of the Carolingian Court School (Fillitz 1965-6), the sources for at least four of which are late antique, perhaps fifth century. Saxl (1943, 13) has shown that an ivory is the source for one scene on the Ruthwell cross; and it need not surprise us that ivories rather than larger works are arguably the source of some motifs in later mediaeval panel paintings and manuscripts: for example, the physiognomical types of the Ada Gospels (ibid., 235f.). In this case, however, it was work from the court of Justinian which was imitated, rather than Western styles.  

The existence of antique ivories at mediaeval courts is therefore used by some scholars as a sufficient explanation for the sources of Carolingian and Ottonian antiquarianism; and it is indeed true that they were one channel whereby knowledge of large scale antique works was maintained (e.g. Graevan 1902) - such as the iconography of Athena, transmitted to the court of Charlemagne via Byzantine work (Lewis 1980, 89ff.). Thus for Colin (1947, 113f.), the representation of a Gallo-Roman tomb in the Sacramentary of Drogo, and other similar structures in ivory panels, come directly from late antique ivories - one possible source, of course, for Villard de Honnecourt's drawing of a `Saracen' tomb. For Jalabert (1965, 46ff.), ivories are the key inspiration for the altar table at Saint Sernin and for the cloister at Moissac - apparently to the exclusion of surviving Roman forms, which she never considers. Weitzmann (1971) is more flexible for, while demonstrating that in Byzantium ivories did indeed form models for marble sculptors (he gives such a parallel for a twelfth-century relief in the sacristy of Ferrara Cathedral), and referring to Goldschmidt's demonstration that ivory models were thus in thirteenth-century Hildesheim and Halberstadt, he also suggests antique text illustrations as models for certain ivories, and does not totally rule out the use of Early Christian marbles during the Middle Byzantine period.  

On the other hand, Schwartz (1960, 158) believes that elements in some Carolingian ivories (in the group attributed to Metz, Liége and Cologne) are themselves directly inspired by antique sarcophagi and tomb structures which, he suggests, the ninth-century artists must actually have seen; and Rosenbaum (1955, 7 and pl. 4) also suggests sarcophagi (albeit with circumspection) as a source for figure groupings in Carolingian canon tables. For Bayle (1980), studying Romanesque sculpture in Normandy, manuscript illumination and ivories together, in some cases of two centuries previously, provide the main but certainly not exclusive channels of inspiration. In some instances, of course, such as the Ionic-style capitals in the crypt of Saint-Germain at Auxerre, it is clear that neither ivories nor other small works of art were not sources: the inspiration for these must derive from Provence, where the monk Heric records the constructors of that church having gone for material.



Metalwork


Metalwork, like coinage (Grierson 1954, 1060), is bullion re-shaped as art. Where the metal value of objects was more highly prized than either their associations or their artistry, it is easy to see why few objects could survive from antiquity, and we have no indications that objects of precious or semi-precious metals (except perhaps for a few coins) survived above ground through the whole of the Middle Ages. The centre about which most is known is inevitably Constantinople, where little was preserved from the oldest periods, for it suffered modernisation, conversion into coinage, or iconoclasm (Lombard 1974, 128ff.). Exceptions are the lavish gifts which reached Western rulers from the Emperor: for example, those to Chilperic included a gem-incrusted salver weighing fifty pounds, and medallions weighing one pound each (HF 6.2). To such imports we must, once again, add the antique material in the religious booty brought back from the Fourth Crusade (Riant 1875). Unfortunately, references in the West are far from explicit, so just what was available is a matter for argument. This is illustrated by the records of about 875 noting the gifts of silverware to churches in the city of Auxerre by Bishop Didier (603-21): Colin (1947, 92ff.) believes not only that this material was seventh century Byzantine, not Roman; but also that it was probably melted down early in the ninth century. He refuses to acknowledge, what is more, that one of the bribes offered to Bishop Theodulf of Orleans on a visit to Provence in 798 included antique art; instead, he believes the vase with twelve scenes from the life of Hercules which is mentioned to have been contemporary Byzantine work (ibid., 103ff.).  

In the West, the centres captured by the Moors suffered the destruction of those treasures which had not already gone to earlier invaders: that of the Visigoths at Toledo, in 711; or that of the Persians at Ctesiphon in 637 - the sale of the precious metal causing inflation and a drop in the price of gold (Lombard 1974, 195ff.). One technique used by the Moors to extract money and goods was high taxation: thus Syrian and Egyptian churches lost their treasures in the eighth and ninth centuries for, writes Lombard (ibid., 198), `like the treasures of the temples in Antiquity, `the treasures of churches and monasteries played, in the Musulman world from the seventh to the ninth century, the role of reserves of metal for turning into money, and on which the sovereign draws from time to time.' The same device was common in Europe as well. Bloch (1933, 12) cites the case of the gold crucifix offered by Archbishop Willigis to the Cathedral of Mainz, one foot of which later went to pay the Pope his pallium, and a second to pay the expenses of a local war. In Gaul, precious objects were sold in times of hardship: accounts survive of liquidation to feed the poor and to pay ransoms (Lesne 1936, 164ff.), and the same happened in England (Dodwell 1982, 7), where the Reformation completed the devastation. Calabi Limentani (1970, 259) reads a similar interest into the noticing of objects of bronze during the later Middle Ages - as, for example, in Magister Gregorius' mention that he saw a bronze wolf and a (small) bronze tablet in the Lateran; nevertheless, some works survived (cf. Gazda 1970 for refs). Metal was also used as payment for goods - as can be proved from the seventh century onwards. Thus Bede tells us ( Hist. Eccl. 3.6) that King Oswald cut up a silver dish to distribute as largesse: there is no hint that an antique piece was involved, but the very deed might suggest that antique objects could also be viewed as bullion, not art. David Herlihy (1957) has shown, by studying the contracts of the period, that treasure was used as substitute money on an increasing scale from 960, reaching a peak in the later eleventh century; over the period studied (960-1139), forty per cent of contracts involved payment in something other than money (and cf. Grierson 1951, 3ff. for the Carolingian period). Similarly, bronze statues in Constantinople were supposedly melted down for coinage at the 1204 Sack (Gazda 1970, 247).  

What objects survive today come largely from modern excavation, from chance discovery, or as imports from the East. But we do have a few mentions of Western discoveries: Gregory of Tours thought it worthwhile to record the casual find of a bronze snake and rat while the Parisians were clearing out a blocked drain (HF 8.33), and also the finding of bronze and iron objects by some monks in 563 (HF 4.31) which, since they are not described, are presumably mentioned as an indication of the value of the metal. The story has a moral for, in their lust for gain, the monks were careless, and the hillside collapsed and buried them. Gregory's monks were not alone in searching for vessels, and the proof that these were indeed dug out throughout the Middle Ages is provided by the various benedictions which allow their re-use in Christian ritual: one, the Benedictio super vasa reperta in locis antiquis calls upon God to cleanse `these vessels made with the skill of the pagans' (cf. Wright 1844, 439f.). Perhaps the vase found at Gilton (Kent) in the nineteenth century is an example of such Christianisation, for it had been repaired in the eighth or ninth century with strips of metal decorated with Saxon figures.  

We also have plentiful archaeological evidence that small items, often of metal and especially of bronze, were robbed from graves, most of which would have been found by chance. At Herdonia, near Foggia, ancient tombs were searched in later Antiquity or the Middle Ages: some were cleared out, but in others pottery was left behind - an indication that metals had been extracted (Greenhalgh 1982, 8). In both Britain and Gaul small metal objects, like gems, seem to have had the power of talismans; and broken keys, Roman brooches, knife-handles, belt-plates, a spoon with a broken handle, and even Neolithic necklaces, arrowheads and flints have been found in graves; fragments of Roman glass are also frequent, reflecting the sophistication and rareness of the medium amongst both Anglo-Saxons and Franks (Meaney 1981, 222, 227f.). To this list should be added many finds of glass beads; although what proportion are Roman (as opposed to Germanic, Prehistoric or contemporary) is apparently difficult to say in the present state of knowledge (ibid., 192ff.).  

Some `bulk' bronze also survived the Middle Ages - such as some of the architectural decoration of the Pantheon. For although its bronze roof tiles were taken to Constantinople in 663 by the last Emperor to visit the city (just as, earlier that same century, in 629, the pope had stripped the Basilica of Maxentius of its gilded bronze tiles for use on S. Peter's), the great bronze roof truss from the pronaos lasted until the seventeenth century, and was then taken, largely to turn into cannon. The (original?) bronze doors still survive, but the bronze cladding of one of the pilasters was melted down in the sixteenth century to make the S. Peter for the top of Trajan's Column. Genseric's sack of Rome in 455 AD had wrought similar damage on the bronzes of the city, including taking over half the bronze cladding of the roof of the Temple of Jupiter Capitolinus (Procop. Wars 3.5); but antique bronze doors survived in sufficient quantities to serve as models for some mediaeval sets (Mende 1983, 18ff.).



Pottery


Although most pottery is no candidate for re-use, its sheer quantity ensured that it was noticed - for it was a sign, like bricks and tiles, of previous settlement on a site. But while a survey of mediaeval pottery would show the survival of a few basic shapes, there is no evidence of a thoroughgoing imitation of Roman forms, or indeed of any collections of pots before the late fifteenth century - excent for evidence suggesting that the Saxons prized and collected the fabric of terra sigillata for use as medecine (N. Hammond in The Times, 30 August, 1984). Occasionally, Roman ware may have been noticed and preserved, however: nearly all the pottery found when excavating the Torre Civica at Pavia was residual; but one first-century black-gloss bowl might have been taken into the tower in the eleventh century, judging from its stratigraphy (Blake 1978A).  

The antique pottery which attracted attention in the later Middle Ages was Arretine and other samian wares, because of its decorative motifs: it was well known at Arezzo itself, probably elsewhere in Italy, and perhaps further afield as well. In a short description of that city, Giovanni Villani mentioned it as the main feature of the place, thinking it impossible that it was the work of mere human beings (1.47). Such was its fame that Ristoro d'Arezzo devoted a whole chapter (if a short one) to it - the Chapter on ancient vases - and, in the long description he gave of the motifs to be found used as decoration, proved a good knowledge of its varieties. It was readily available, and Villani wrote that `when today one digs for any reason inside the city or up to two miles around it, large quantities of these vase pieces are found, more in some places than in others ... it is assumed that they have lain underground for about a thousand years'. What is more, they were prized by artists: `when one of these pieces came into the hands of sculptors or draughtsmen or other knowledgeable people, they considered them holy objects, and said they were divine work. And it was thought that the subtle nobility of such vases, which were carried over almost all the world, was a gift of God Himself.' How could Ristoro have known of the export of such vases, were not reports of their finding elsewhere available to him?  

In Villani's day, then, kilns and their spoil-heaps were known, probably outside the Roman city and hence in the suburbs of its mediaeval successor - so that, again, urban expansion apparently helped uncover antiquities. Their location is illustrated by the discovery of one kiln with four whole vases and many broken ones, said by Vasari to have been discovered by his ancestor Lazzaro Vasari (died 1452) near the Ponte alla Calciarella. The toponym could be a corruption of `Carciarelle', because the furnaces looked like little prisons, or of `Calcinelle', `the little lime-kilns' (Gamurrini 1890, 65). Indeed, one suburb was called `of the furnaces' (Fabroni 1841, 18; 60). Much earlier, we have `Orciolaia' - `the place of little vases' - in a document of 1354; and another furnace site just outside the Porta Buia, in a deed of 1315 (Gamurrini 1890, 63ff.). Indeed, one puzzling word in Ristoro's account must indicate that kilns were rediscovered in his day, and not simply pots; for he states that `the vases were made of two colours, light blue and red, but more of the red'. Arretine pots are always in tones of red, so the `blue' ones must be rejects spoilt in the kilns and thrown away (Ballardini 1964, 70ff. for technical details); or just possibly, as Fabroni suggests (1841, 36), a bluish tint given to the black glaze under certain light conditions.  

Such hints make it probable that motifs on Arretine pottery were used as models by mediaeval artists, as they certainly were by those of the High Renaissance (Yuen 1979). They are mentioned several times by Vasari, who also refers to the availability of Etruscan pottery, painted and plain (e.g. in his life of Battista Franco: Vasari-Milanesi 6.581f.). A thorough search would no doubt show that other categories of antique vase were known and imitated in the Middle Ages: Gombrich, for example, has recently suggested (1976) that the decorative palmettes on Bonaventura Berlinghieri's S. Francis panel in Pescia shows motifs which demonstrate knowledge of antique Greek originals, and adduces an Apulian situla for comparison. He is surely correct to point to Pisa and Lucca as places where antique spolia provided inspiration, and to maintain that `a mason of early thirteenth-century Lucca would hardly have been surprised to see a painter deriving similar inspiration from a painted fragment of an ancient vase'. But the difficulties involved in `placing' a ubiquitous motif like a palmette into one century are surely too great for our present state of knowledge, and it is perhaps  to look to the vigorous tradition of architectural friezes (all but one of Berlinghieri's friezes are indeed architectural) conceived in Italy in terracotta as well as stone and marble and maintained throughout antiquity.