12
HELLENISTIC MOSAICS
Ruth Westgate
Mosaic is a hellenistic invention: the modern form of mosaic, made with cut cubes (tesserae) of stone or glass, which has adorned Roman baths and houses, Christian churches, and today’s public spaces, was first created in the hellenistic world. This paper is an outline of the evolution of mosaics in the hellenistic period, setting it against the historical and social background that produced this invention and shaped the early development of the medium. I hope in the process to draw attention to a form of Greek art which has been unduly neglected. Martin Robertson, in an article which has become the standard work on Greek mosaics (Robertson 1965), ended his narrative with the introduction of tessellated mosaic, dismissing the products of the late hellenistic period as ‘not negligible but really rather horrible’ (p. 89). Since he wrote, a considerable amount of new material has been discovered and published,1 and it is now possible to give a somewhat clearer – and, I hope, more favourable – account of these early mosaics.
However, although we now know of several hundred mosaics from the hellenistic world, it is still not possible to give a straightforward history of their stylistic and technical development. Too few pavements can be dated with any precision, and many of the dated examples are clustered in both place and time, making it difficult to distinguish stylistic change from regional variation. So, rather than trying to give a conventional history, this paper considers the surviving mosaics as a product of the social and economic conditions of the period, looking at the ways in which mosaics were used, and what they meant to the inhabitants of different parts of the hellenistic world.
The most important fact about mosaics in the hellenistic world is that they were made principally for private houses: about 80 per cent of tessellated mosaics from the period have been found in domestic contexts. As everyone instinctively knows, houses reveal a great deal about their occupants: people decorate their homes in a style that they feel to be desirable or fashionable, creating a living environment that reflects not only their status, but also their aspirations; in short, they try to present themselves as they would like to be seen. The stylistic and technical development of mosaic can be understood as a response to these very personal and yet very public needs. Mosaics were not of course the only means by which this was achieved: wall paintings, sculptures and textiles would also have contributed to the total effect. But textiles decay, sculptures are carried off by looters and lime-burners, and painted plaster crumbles when walls collapse; floors are often all that is left. By looking at mosaics, therefore – at their designs and the ways in which they were used, and, where possible, at the architecture and other forms of decoration that accompanied them – we can attempt to see their owners as their contemporaries might have seen them, and thus gain some insight into the tastes and aspirations of people living in the hellenistic period.
At the beginning of the period, mosaics of natural pebbles were the most common form of decorative paving. Pebble mosaics too are mainly found in houses, most frequently in the dining room, or andron, and its adjacent anteroom, which were the principal areas for entertaining guests. Their standard composition, consisting of concentric bands of black and white decoration framing a central motif, seems designed to present a satisfying view to the diners reclining on couches around the walls of the andron (Westgate 1997–8, 94–7, 102). The introduction of mosaics in the late fifth century was one element in a general trend towards the elaboration of private houses; the formal dining room itself was also part of this trend, along with painted wall plaster, purpose-built bathrooms, and occasionally peristyles (Walter-Karydi 1994). Many of these elaborations were clearly intended to impress outsiders visiting the house, particularly for the symposium, and this growing willingness to invest in private display seems symptomatic of a shift in priorities, away from the community and towards the interests of the individual, which is the beginning of one of the most characteristic aspects of the hellenistic period (as identified by Pollitt, 1986, 7–10).
In order to understand the invention of tessellated mosaic, it is first necessary to consider developments in pebble mosaic in the early hellenistic period, which seem to set the tone for what follows. Like so many strands in hellenistic history and art, the development of hellenistic mosaics starts in Macedonia, with the encounter between Greek and Macedonian cultures. Two large peristyle houses at Pella, dated to the last quarter of the fourth century, have yielded a series of spectacular pebble mosaics (Salzmann 1982, nos. 94–104, pls. 29–37; Makaronas and Giouri 1989). In the House of Dionysos (insula I.1) two anterooms (A and D) with black and white geometric mosaics led to andrones with central figured scenes, the eponymous Dionysos in room B, and two hunters attacking a lion in room C. The nearby House of the Rape of Helen (I.5) had mosaics in three large andrōnes: room B was decorated with elaborate floral scrolls, room Γ with an enormous scene of Theseus abducting Helen, and room Δ with a stag hunt (Fig. 1); a fourth mosaic, depicting an Amazonomachy, was found in an anteroom (I), which led to two plainer andrōnes.
Fig. 1. Pella, House of the Rape of Helen: Stag Hunt mosaic in room Δ. Photo courtesy of Archaeological Receipts Fund.
Like the earlier pebble mosaics in Greek cities, these decorate dining rooms, but they are designed for banquets on a much larger scale: the smallest rooms would have accommodated 11 couches, and the largest 20, compared to the usual seven-couch size in earlier houses. Not only are the rooms, and thus the mosaics, much larger than anything that preceded them, but the houses themselves are many times bigger than even the largest known Greek houses,2 and they each have multiple dining rooms – as many as five in the House of the Rape of Helen. The date in the last quarter of the fourth century makes it likely that these lavish houses and mosaics were financed by the spoils of Alexander’s eastern campaigns, which must have made many of the Macedonian aristocracy wealthy beyond imagination.
But it was not simply the massive influx of wealth into Macedonia that generated these mosaics: they were also a product of the very different culture and values of the Macedonians. Their monarchic and aristocratic society encouraged individual self-promotion, and they were much less inhibited than the Greeks about private luxury and ostentation. Long before this, for example, literary sources report that King Archelaos employed the painter Zeuxis to decorate his palace (Aelian, Varia Historia XIV.17), presumably with figured paintings – a form of decoration which, to the Greeks, belonged in public and religious buildings; likewise, the fact that the only surviving examples of Greek figured painting from this period are in Macedonia is a consequence of their use in a private context, in elaborate subterranean chamber tombs, rather than in public structures above ground.
The Macedonians adopted the forms of the Greek symposium, the characteristic architectural setting of the andron, decorated with painted plaster and mosaics, and the equipment, the couches and the drinking-vessels, often in luxurious precious metal (and again more frequently preserved in Macedonia as a result of their use as grave-goods); but they used the symposium for a rather different purpose. The intention seems to be not to foster a sense of intimacy between the participants, but to impress the guests with the host’s wealth and power.
Both the cultural fusion that produced these mosaics and the ostentatious entertainments that they adorned are typically hellenistic.3 Looking more closely at the mosaics themselves, it can be seen that they foreshadow later technical and stylistic developments too.
The most obvious difference from earlier mosaics is the strikingly three-dimensional appearance of the scenes depicted, which is achieved through perspective and shading. A few earlier pavements use small touches of colour for details, but this colouristic modelling is a novelty. It is usually attributed to a desire to imitate painting (e.g. by Bruneau, 1987, 51–4) – perhaps, arguably, to copy specific works – but it is also a function of the scale of the pavements: the floors are very large and the pebbles are relatively small, which enabled the mosaicists to achieve a much more realistic effect than was possible within the area of a conventional seven-couch andrōn. The cost of obtaining the pebbles must have been enormous, as they are carefully graded by both size and colour: although grading by size could in theory have been done mechanically, perhaps by sifting, grading by colour could only be done by eye, one pebble at a time. It is clear that these pavements must have cost far more than any of the earlier mosaics known to us,4 and it seems that the particular economic and social conditions in Macedonia at this time must have created the environment in which this illusionistic style became realizable.
The other major innovation at Pella is the introduction of artificial materials in addition to pebbles. In several of the mosaics thin strips of terracotta or lead are used to outline the figures and to represent details: the strips are all terracotta in the Lion Hunt, while in the Dionysos and the Rape of Helen both lead and terracotta are used (Fig. 2), lead mainly for profiles and fine details, and terracotta where a bolder line was required. Lead was presumably more practical than terracotta, as it could be shaped and reshaped as the design developed, and it is not surprising that only lead strips are found in later mosaics elsewhere. The strips create sharper outlines and finer details than could be achieved with pebbles alone; they form part of a ‘linear’ style characterized by calligraphic lines and simple grey or brown shading, in contrast to the more ‘painterly’ style of the Stag Hunt and Amazonomachy pavements (Fig. 1). The three ‘linear’ pavements also contained other artificial elements: the eyes were inlaid, perhaps in some precious material, which has disappeared leaving empty sockets (Fig. 2); and green-painted clay beads and pieces of blue glass were used to supply colours that could not be obtained in natural stone.5 Again, a desire to imitate painting may lie behind these innovations. But they also point to some of the limitations of pebbles as a medium, and this is the background to the invention of tessellated mosaic.
Fig. 2. Pella, House of the Rape of Helen: detail of the Rape of Helen mosaic, room Γ.
The origins of the tessellated technique are still a mystery: it is not clear exactly when it was invented, or where, or how. Pebble mosaics continued to be made throughout the third century (and, sporadically, in the late hellenistic period and beyond), but by the mid-second century tessellated mosaic was clearly established as the standard form.
The pavement that is most commonly identified as the earliest known tessera mosaic is in a house at Morgantina in Sicily, and depicts Ganymede being carried off by Zeus’s eagle (Fig. 3). When the mosaic was discovered, the American excavators believed that the house was abandoned after the city was sacked by the Romans in 211 BC (Phillips 1960, 243); a coin of Hieron II, found under the threshold of room 16, gave a terminus post quem of c. 260–250 BC for the construction of the house, so they concluded that the mosaic must have been made in the mid-third century. The technique of the mosaic is rather crude, with details such as Ganymede’s toes and testicles represented by pieces of stone cut to shape, rather than being built up from individual tesserae; two other mosaics in the same house are similarly rough (Tsakirgis 1989, 396–9, nos. 1 and 2, figs. 1–9). It was therefore seen as an early step in the development of tessellation, predating a number of mosaics with similar cut pieces, which Phillips dated to the second century (1960, 247–53). This fitted into a plausible historical narrative, placing the invention of tessellation in the sphere of Syracuse at its peak; and it seemed to be supported by Moschion’s description, quoted by Athenaeus (Deipnosophistai V.207c), of a luxury ship sent to Alexandria by Hieron II as a gift to Ptolemy III Euergetes, which is said to have been decorated with mosaics.
Fig. 3. Morgantina, House of Ganymede, room 14: Ganymede and the eagle. Photo courtesy of Department of Art and Archaeology, Princeton University.
However, it is not certain that the mosaics mentioned in this passage were made of tesserae rather than pebbles, and in any case it is a highly elaborate ekphrasis of uncertain date and questionable reliability (Dunbabin 1979, 265 n. 5; 1994, 26 n. 2). Moreover, as the excavations at Morgantina proceeded, it became clear that the House of Ganymede was reoccupied along with the rest of the city in the mid-second century BC. Some houses were rebuilt entirely, but others were simply refurbished, and several had mosaics installed. Some of these new mosaics are securely dated to the second or early first century by material sealed underneath them (e.g. in the Pappalardo House: Tsakirgis 1989, 405–6, no. 13; 413); others are dated by style alone (in the House of the Arched Cistern: Tsakirgis 1989, 401–3, nos. 5–9; 413). The House of Ganymede is still thought to have been built in the third century, before the sack, but the foundations of the mosaics contained no datable material, leaving open the possibility that they were part of a refurbishment after the sack. The excavators have stuck to the third-century date (Tsakirgis 1989, 412–13), but this now appears to be largely dependent on the assumption that the crude technique is an early attempt at tessellated mosaic. However, in the four decades since the mosaic was first published, our understanding of early tessellated mosaic has improved considerably, and the pavements cited by Phillips as the next step in the evolution of the technique are now thought to be late rather than mid-hellenistic. A third-century date is still possible, but it seems equally plausible that the unusual technique and general crudity of the mosaics in the House of Ganymede are simply the mark of a rather incompetent local mosaicist or workshop.
Four decades of new discoveries, however, have not produced evidence to support a clear model of the development of tessellated mosaic to replace Phillips’ account. What follows is a summary of the current state of our knowledge.
Two possible sources for the tessellated technique have been identified, which are not necessarily mutually exclusive. In Greece and Asia Minor, an alternative technique already existed alongside pebble mosaic in the fourth century, using irregular chips of stone. Some of these chip mosaics are plain,6 but two at Olynthos contain simple decorative elements in pebbles.7 In some cases chips of a contrasting colour are introduced to form a pattern: few of these pavements can be reliably dated, but one at Athens, decorated with a simple rosette in red and grey chips on a white ground, is thoughtto have been made in the late fourth or third century (Salzmann 1982, no. 151, pl. 78.1–2); a fragmentary mosaic from Aphrodisias (Fig. 4), part of which probably depicts a dolphin, has a terminus post quem of c. 261–246 from coins of Antiochos III found underneath it (Salzmann 1982, nos 144–5); and a black and white pavement at Euesperides must pre-date the desertion of the city in the mid-third century (Lloyd et al. 1998, 150–7, figs. 5–9). In another pavement at Euesperides (Salzmann 1982, no. 156, pl. 91.1–3) the pieces are more regular, and could be described as rough tesserae rather than stone chips; a more elaborate pavement of this type at Assos, now lost, had a terminus post quem in the fourth century from a coin sealed beneath it (Salzmann 1982, no. 150, pl. 84.3–4); and there is even a tantalizing reference in a footnote (Robinson 1933, 1, n. 4) to a fragment of irregular tessellated mosaic found at Olynthos, which must date from the period before the city was sacked in 348; but unfortunately the fragment is lost and was never photographed, so its technique cannot be verified.
Fig. 4. Aphrodisias: mosaic of irregular stone chips from under the Temple of Aphrodite.
Katherine Dunbabin (1979) suggested very plausibly that these chip mosaics may have played a part in the development of tessellated mosaic, and Salzmann, in his study of pebble mosaics (1982), also argues that chip mosaics represent an intermediate stage between pebbles and tesserae. Another category of pavements, in which pebbles are combined with various grades of irregular and regular tesserae, may also belong to this transitional phase. However, as the dating evidence for most of these ‘intermediate’ pavements is inconclusive or non-existent, it is still not clear how this evolution might have taken place.
On the other hand, Dunbabin (1994) has also pointed out that in Sicily and southern Italy squared tesserae were in use long before the appearance of tessellated mosaics, to decorate pavements of opus signinum (a type of red mortar containing crushed terracotta). This technique was borrowed by the western Greeks from their Punic neighbours: pavements of this type were made at Punic sites in North Africa at least as early as the fourth century, and some small areas of continuous tessellation have been found in contexts which pre-date the first known tessellated mosaics at Greek sites.8 Some of the earliest Greek tessera mosaics are associated with opus signinum: it was common to mark thresholds with a denser concentration of inset materials, and sometimes these form a true mosaic, as in the baths at Megara Hyblaia, which must date before the destruction of the city in 214 (Fig. 5). A threshold from Gela with a meander pattern in tesserae (Salzmann 1982, no. 157, pl. 92.4) may be even earlier, as the city was destroyed in c. 282; the rest of the room was paved in stone chips rather than signinum, but the principle is the same. It is possible, therefore, that the tessellated technique evolved independently in different parts of the Greek world.
Fig. 5. Megara Hyblaia, baths, room g: opus signinum pavement with tessellated threshold.
But it is still difficult to explain the reasons behind the development of tessellated mosaic. Most of the ‘intermediate’ pavements of stone chips and irregular tesserae are very simple or of very poor quality; they hardly seem to represent the cutting edge of innovation – or to be successful attempts at imitating painting, which is the motive most frequently suggested for the invention of tessellation. Moreover, floors of this type continued to be made in the late hellenistic period, long after the transitional phase, apparently as cheap substitutes for tessellated mosaic.9 It is possible, then, that the new technique developed initially as a cheaper alternative to pebble mosaic: although pebbles appear to be a free material, produced by natural forces, in fact making a pebble mosaic may have entailed more than simply collecting stones from the nearest beach or river-bed. In order to achieve the sharp contrast between black and white that is characteristic of all but the most incompetent pebble mosaics, it may have been necessary to obtain materials from areas where the geological conditions generated pebbles of a consistent colour, while selecting a large quantity of suitably coloured and evenly sized pebbles must have been a laborious task. Stone chips, on the other hand, might have been waste from a stonemason’s or sculptor’s yard; even obtaining a block of stone and breaking it up into tesserae might have been cheaper than using pebbles – and of course the resulting tesserae would be of a uniform colour and size. In other words, the invention of tessellated mosaic may have been driven by demand from people who wanted the distinction of a mosaic but could not afford one. Such an explanation would account for the poor quality of most of the ‘intermediate’ pavements.
Other advantages of the new technique would have quickly become apparent. Using cut stone pieces probably made it easier to obtain a wider range of colours, and artificial materials broadened the range still more: terracotta was a cheap and easily available source of reds and yellows; and bright blues and greens, which are difficult or impossible to obtain in natural stone, could be supplied by glass and faience (Guimier-Sorbets and Nenna 1992).
In the present state of our knowledge, this is only speculation. All we can say with any certainty is that various embryonic tessellated techniques existed in both east and west in the third century, and possibly earlier; Baldassare has rightly stressed the importance of considering each of these early manifestations in its regional context, rather than attempting to force them into a single sequence of development (1994; see also the discussion of Arpi below). No regular tessellated mosaics with figures or complex designs can be securely dated earlier than the second century, and it is not at all clear how the simple tessellated floors of the third century developed into the accomplished products of the late hellenistic period. However, an obvious place to look for the refinement of the technique is one or other of the hellenistic royal courts. By a lucky chance we happen to have two groups of mosaics which are likely to have been royal commissions, and which give us an indication of the quality of the best products of the period.
Because of its central role in hellenistic culture, Alexandria has always been a favourite candidate for the place where the tessellated technique was created. It has yielded a series of mosaics representing almost the entire spectrum of techniques from pebble mosaic to tessellation, including mixtures in various proportions, which has been seen as reflecting the process of invention.10 However, this is based on a hypothetical model of gradual stylistic and technical development which is not underpinned by dates from external evidence. As little is known about the context of most of the mosaics, it is hard to prove that the Ptolemaic court was the driving force behind the development, but its claim has recently been strengthened by two spectacular new finds from the Royal Quarter of the city (Guimier-Sorbets 1998b).
One pavement shows an endearingly life-like dog (Fig. 6), framed by a coloured guilloche and an exquisite border of lion-head ‘spouts’ in trompe l’oeil relief; the other, more fragmentary, depicts a black and a white youth wrestling (Guimier-Sorbets 1998b, figs. 7 and 8). Both are made in the very fine mosaic described by modern authors as opus vermiculatum, in which the tesserae measure 5 mm or less. These obviously represent a stage at which the tessellated technique has been virtually perfected, but they share a technical feature with pebble mosaics which suggests that they might date fairly early in the evolution of tessellated mosaic. In a few late pebble mosaics, lead strips are used to represent the outline of figures, as described above; but by the late second century, strips in tessellated mosaics are only found in geometric patterns, where they apparently function as a technical aid rather than as an artistic element in their own right (Fig. 7).11 In the Alexandrian mosaics, lead strips are used in both the figured and the geometric parts of the design, which suggests that they might represent a stage of development between the mosaics at Pella and those of the late hellenistic period. A date in the early- or mid-second century has been suggested (Guimier-Sorbets 1998b, 289), but there is little external dating evidence, and it remains possible that the technique is just a peculiarity of an individual workshop. In fact, the only other mosaic that shares this feature was also found in Alexandria, on the edge of the Royal Quarter (it is not clear whether inside or out). It depicts three Erotes hunting a stag, framed by a border of wild animals (Fig. 8).
Fig. 6. Alexandria, Royal Quarter: dog mosaic. Photo courtesy of Centre d’Études Alexandrines/A. Pelle.
Again, there is no independent evidence for the date, and estimates based on style and the topography of the area have ranged from the late fourth century BC to AD 50, but because the mosaic includes some pebbles along with the tesserae, and because of its obvious stylistic similarity to pebble mosaic and specifically to the Pella Stag Hunt (Fig. 1), it has most often been seen as an early, experimental tessera mosaic, whose creator had not fully moved away from the pebble technique (Daszewski 1985, 75–6, 106–9, with references to earlier literature). However, the pebbles are used only for a specific purpose, to represent the texture of hair. Elsewhere in the pavement similar shades of red and brown are rendered in tesserae (the black and white photograph masks a colour range which is rather richer than most pebble mosaics). Other features of the technique seem more advanced, especially the use of opus sectile (cut stone patterns) on the threshold, which has no parallels as early as the fourth century: it becomes common much later, in the late second or first century BC. I would prefer to see the mosaic as a sophisticated archaistic work, imitating the style of pebble mosaic in a playful mood which is very characteristic of the hellenistic period; the resemblance to the Pella Stag Hunt is probably superficial, produced by the use of stock figure-types in a formulaic composition scheme (Westgate 1999, 22–3).
The technical similarity with the dog and the wrestlers suggests that all three mosaics could be products of the same workshop, which perhaps worked for the Ptolemaic court; another feature shared by this group is the setting of tesserae in diagonal rows, which is very rare in mosaics from the hellenistic east, although it is common at western Greek sites (Westgate 2000a, 258–9, n. 14).
Fig. 7. Lead strips in tessellated mosaic (Palermo, Museo Nazionale, inv. no. 2288).
Fig. 8. Alexandria, Erotes mosaic from Shatby. Photo courtesy of DAI Cairo.
A more securely dated group of relatively early tessellated pavements has been found at Pergamon, which also happens to have been the location of the only mosaic that was famous enough to be recorded in ancient literary sources. Pliny the Elder (HN 36.184) describes a mosaic made by Sosus of Pergamon, called the ‘Unswept Room’, which showed the detritus of a meal, apparently dropped on the floor; on the same pavement or nearby was a scene of doves drinking from a basin. This mosaic survives only in copies,12 but we do have some very fine pavements from Palaces IV and V. Only fragments were found in Palace IV, but the ground floor of Palace V yielded two fairly complete pavements.13 The larger one, in the North-West Room (Fig. 9), had concentric bands of decoration, including crenellated towers, guilloche, waves, meanders and a floral scroll peopled with tiny Erotes, birds and insects, framing a large central field which was divided into four panels, all now missing. The smaller, in the Altar Room (Fig. 10), had borders of coloured squares and bead-and-reel enclosing two friezes decorated with garlands and three central panels, of which the sole survivor depicts a big green parrot; two further panels flanking the altar showed dramatic masks.
Both mosaics are entirely covered with decoration, with no plain bands between the patterned borders; the technique is very regular and careful, all the patterns are outlined with lead strips, and expensive materials – glass, faience and perhaps even crystal or mother-of-pearl – were used in addition to stone. The figured and floral elements were made in the finest opus vermiculatum, with tesserae as small as half a millimetre; the work is exquisite, with delicate shading and illusionistic effects (Fig. 10). Their high value is indicated by the fact that most of the panels and one entire border had been prised out of the floor. The mosaic in the North-West Room was signed by the artist, Hephaistion, on a trompe l’oeil card stuck down with blobs of red wax – a motif in the spirit of Sosus’ ‘Unswept Room’, and perhaps, as Robertson wryly suggested (1965, 88), carefully left behind by the looters because they wanted to pass off the mosaic as Sosus’ work.
The mosaics were laid in the first half of the second century, probably towards the end of the reign of Eumenes II (197–159) or during that of Attalos II (159–138); rejected blocks from the Great Altar built into Palaces IV and V give a terminus post quem, and recent soundings underneath the mosaics produced no material later than the middle of the century (Salzmann 1995, 109–10). Another mosaic in a very similar style was found in the Temple of Hera Basileia, which was dedicated by Attalos II (Dörpfeld 1912, 262, 326–8, pls. 17, 18, 22a, 27); it was probably made by the same craftsmen, who were presumably working for the court.
Fig. 9. Pergamon, Palace V: mosaic from the North-West Room.
Fig. 10. Pergamon, Palace V: detail of mosaic from the Altar Room.
These are the earliest tessellated mosaics that can be dated by external evidence rather than stylistic guesswork, and, like the Alexandrian ones, they represent the best work from the period. It is tempting to suggest that the rough tessellated technique of the third century was first refined into this elegant form by the court craftsmen of Alexandria or Pergamon as a suitably luxurious decoration for a royal residence, displaying not only the skill and labour entailed in creating such meticulously detailed work, but also the wealth of materials needed to produce the required range of colours. In addition, the subject-matter and painterly style of the figured elements may have had prestigious cultural and artistic resonances: the effect of the Pergamene mosaics in Salzmann’s new reconstructions is strikingly reminiscent of a gallery of paintings (1995, 106–9, pl. 21, foldout plan 2).
This model cannot be proved on the evidence currently available, but it does not seem unreasonable to imagine that the kings were leaders of fashion, representing a standard of living to which the inhabitants of the hellenistic world aspired. From about 150 BC onwards there appears to be a substantial increase in mosaic production, part of an upward spiral of private luxury and ostentation which dates back to the introduction of pebble mosaic in the late fifth century, but which accelerated dramatically in the late hellenistic period, fuelled by economic prosperity on an unprecedented scale (Westgate 1997–8, 111–15). This increase in luxury can be seen most vividly on the island of Delos, where there are many well-preserved houses dating to the late second and early first century (Chamonard 1922; Trümper 1998). Along with mosaics came wall-plaster in the Masonry Style, moulded and painted to look like stonework; stone statues (Kreeb 1988); comforts like built-in lavatories and heated baths (Trümper 1998, 63–8); and features borrowed from monumental architecture, notably the peristyle, which was apparently so desirable that it often seems to be squeezed into the houses at the cost of practicality.
However, most late hellenistic mosaics are less elaborate than those from the royal capitals, with fewer borders, and more white space between the decorated areas (Figs. 12 and 13). Their decoration consists mostly of geometric patterns, especially waves, three-dimensional meanders and perspective cubes, various types of guilloche, and architectural mouldings such as dentils or egg-and-dart. Relatively few – no more than 16 per cent – depict humans or animals, compared to about 40 per cent of pebble mosaics, which is very surprising in view of the common assumption that tessellation was invented to make it easier to produce realistic figured scenes. The plainer designs may reflect an aesthetic preference, but they were presumably also cheaper to produce, making mosaics affordable for a wider section of society. Moreover, the technical refinements seen at Alexandria and Pergamon are not widespread in later mosaics: figured scenes are often represented in standard-sized tesserae rather than intricate opus vermiculatum, the range of colours is rarely so rich, and delicate, expensive materials like glass and faience are relatively unusual. Lead strips are used in only about a third of the surviving mosaics, generally the better-quality ones; they disappear almost entirely by the mid-first century BC. These observations might support the hypothesis that tessellated mosaic was developed into a luxury art in elite circles, and then adapted and simplified to meet demand from the aspiring classes who wanted to emulate the lifestyle of the elite – a process that continued in the Roman period, when opus sectile seems to have been favoured as the most prestigious type of pavement, and mosaic was relegated to second-best.14
In some cases it is possible to observe these compromises between complexity of decoration and quality of execution. For example, in the House of the Masks on Delos, the owner commissioned a very large area of mosaic, covering four rooms and including a number of ambitious figural motifs, but apparently he could only afford to have most of the decoration done in coarse, rather irregular tessellation, which makes the scenes rather indistinct (Fig. 11). The only exception is the fine centrepiece of one of the floors, showing Dionysos riding a panther, which is an emblema, a panel made separately and inserted into the pavement – but the mosaicists failed to set it straight in the frame.15 In contrast, a customer on Samos opted for a smaller area of decoration, but of much finer quality (Fig. 12; Giannouli and Guimier-Sorbets 1988): the two mosaics have only two decorated borders each, in one room a band of coloured scales and a scroll with exquisite flowers, and in the other a meander and an unusual frieze of griffin-heads linked by a serpent-like body; the remainder of the floors is white. Others chose to economize further by having the outer parts of the floor done in irregular tesserae or stone chips (Fig. 13), or, in the west, opus signinum. A wide range of techniques was used, and there were clearly options to suit every budget.
Fig. 11. Delos, House of the Masks, room H: dancing Silenos (Delos 216).
Favourite subjects include Dionysos and his companions (Fig. 11); the theatre, most often in the form of comic masks; and other motifs which might allude to the Dionysiac sphere, such as drinking-vessels and ivy or vine scrolls. Marine motifs are very common, especially dolphins (Figs. 13 and 15), anchors and tridents; fish are a popular subject for fine emblemata (Fig. 16), perhaps reflecting their desirability as food.16 The natural world in general is a major source of images, such as birds (Fig. 10), animals (Figs. 6 and 8), and fantastic plant scrolls and rosettes (Figs. 9, 13 and 16). Another favourite theme is success and good fortune, represented by palm-branches, crowns and prize vases. Eros makes frequent appearances (Figs. 8, 9 and 16). Mosaic decoration seems designed to create an atmosphere of pleasure, grace and abundance.
Fig. 12. Late hellenistic mosaic at Pythagoreio, Samos. Photo courtesy of V. Giannouli.
The distribution of mosaics and other decoration in late hellenistic houses suggests that there had been significant changes in the use of domestic space since the classical period.17 In most classical houses wall and floor decoration is confined to a very limited area – usually just the dining suite – and the rest of the house is functional in appearance, whereas later houses often have wall plaster throughout, and mosaics in several rooms, both large and small, and sometimes in the courtyard or peristyle. On Delos at least, the best decoration of all seems to have been on the upper floors, which in many cases were used as separate dwellings.18 At more spacious sites, like Morgantina, some houses have a second courtyard, which is less elaborately decorated.
Most late hellenistic mosaics have a concentric design, of the type used in the classical andron, and there is often a plain outer border of suitable width for dining couches (Figs. 12 and 13). However, unlike classical andrones, rooms with mosaics no longer seem to be designed exclusively for dining: few are built to accommodate regular arrangements of couches, and many have multiple doors and windows which would have made such an arrangement inconvenient if not impossible (Trümper 1998, 146–7; Westgate, forthcoming). Symmetry often seems to take priority over the location of furniture or doorways in the design of mosaics: in Fig. 13, for example, the subsidiary panel which should mark the threshold is not aligned with the door, but with the main decorated ‘carpet’ of the mosaic. It seems that the concentric scheme had to some extent become conventional, and that rooms with mosaics were now intended for a wider variety of activities: no doubt many were still used for symposia, but they may have housed other social occasions, perhaps at different times of day. Lavishly decorated peristyles and carefully planned vistas give the impression that the house as a whole was on display to visitors.
Fig. 13. Delos, Quartier du Théâtre III N: mosaic in room I (Delos 261).
On the other hand, it is possible that the decoration was not simply marking out the more ‘public’ areas of the house, where guests were received, but also served to reflect distinctions within the household, between different categories of inhabitants, most obviously between the free occupants and their slaves, although age or gender divisions may also have been involved. Unfortunately the lack of reliable evidence for the activities that took place in individual rooms makes it difficult for us to understand exactly what the decoration meant, but it seems likely that there had been a change both in the boundaries between inhabitants and outsiders and in relations between members of the household. No less significant is the shift in attitudes implied by the spread of decoration: presumably the increased social mobility and insecurity of the hellenistic world made it more important to advertise one’s wealth, status and aspirations, which must have been a driving force behind the spiralling luxury that can be observed in the homes of the period.
Delos has yielded the largest group of late hellenistic mosaics, constituting almost half the total number known, which gives some idea of the phenomenal prosperity of the island in the decades after the Roman intervention of 166. Other major mosaic producers included the rival trading state of Rhodes and the royal capitals (although none have survived from Antioch); but hellenistic mosaics have been found all over the Greek world, from Spain in the west to Afghanistan in the east, as far north as the Crimea, along the African coast and far up the Nile, in contrast to classical pebble mosaics, which are concentrated on the Greek mainland and the coast of Asia Minor. Their distribution reflects the spread of Greek culture, and perhaps we can get an impression of what they meant to people by looking at their uses in some areas on the margins of the hellenistic world.
The site of Ai Khanoum, in modern Afghanistan, was a fortified town on the eastern frontier of the Indo-Greek kingdom of Bactria, and thus on the eastern frontier of the hellenistic world itself. The Greek rulers of the area built themselves a monumental Greek city, with all the trappings of Greek culture – gymnasium, temenos, theatre, and even a copy on stone of the Delphic maxims (Rapin 1990, 333–41). But there were also native elements, including two temples of Mesopotamian type, presumably dedicated to local deities. The huge palace, built in the second century, is also in a hybrid style, basically Achaemenid in plan, but with some Greek elements, including pebble mosaics. These are still decorated in the classical concentric style, with motifs that are familiar from the classical repertoire: one (Salzmann 1982, no. 2, pl. 70) has a border of wave pattern and a frieze of sea creatures, framing a central ‘star of Vergina’; another (Salzmann 1982, no. 3, pl. 71.1–4) is decorated with rosette and palmette motifs. But they are used in a different context from classical pebble mosaics, in bath suites rather than dining rooms.19
Nearer to the Mediterranean, the region of Kommagene in south-western Turkey broke away from Syria in the second century, under the leadership of a half-Greek, half-Persian dynasty. In the first century BC, King Antiochos I (c. 69–c. 36) constructed a cult centre at Arsameia on the Nymphaios in honour of his father Mithradates I Kallinikos, which included a building paved with several mosaics (Fig. 14; Salzmann 1982, nos. 146–9, pl. 86.3–5; Bingöl 1997, 106–7, figs. 71–4). The technique is rough and the colour scheme is restricted to black, white and red, but the multiple-banded compositions and the motifs, especially the border of crenellated towers, recall the mosaics in the palaces at Pergamon; the resemblance may well be deliberate. The royal palace at Samosata, recently rescued from the rising waters of the Atatürk Dam and not yet fully published, has produced similar mosaics, whose imagery alludes to Greek culture. One has a typically Greek design, with a border of fish and a central pair of dolphins flanking an amphora – perhaps, judging from its shape, a container for fine Chian wine (Bingöl 1997, 110, pl. 24.2); another fragment, from the centre of a floor, shows the mask of a character from Greek comedy, probably the Brothelkeeper (Bingöl 1997, 107, figs. 75 and 76, pl. 24.1). The mosaics are accompanied by wall paintings in the Second Style (Bingöl 1997, 111–18, figs. 77–84, pl. 25.1), and date either from the reign of Antiochos or that of his predecessor Mithradates (c. 96–c. 70 BC). Although some of the public monuments commissioned by the Kommagenian regime reflected the ruling dynasty’s mixed Greek-Persian origins, their choice of interior decor seems designed to stress the Greek side of their cultural identity.
Fig. 14. Arsameia on the Nymphaios, Hierothesion of Mithradates I Kallinikos: mosaic in Room II of the ‘Mosaic Building’. Photo courtesy of Seminar für Alte Geschichte, Westfälische Wilhelms-Universität Münster.
On the other side of the hellenistic world, at Arpi, near Foggia in south-east Italy, the native Daunian settlement has yielded two peristyle houses and a tomb decorated with mosaics which are clearly inspired by Greek prototypes. Some are made of black and white pebbles (Salzmann 1982, nos. 12–14, pls. 67–8, and Mazzei 1995, 192–3, nos. 1–4, figs. 112, 115); they have the characteristic concentric borders of geometric patterns or parades of animals and sea-creatures, around a central field decorated with further creatures (Fig. 15). Others are in rough tesserae, but are otherwise very similar in design (Mazzei 1995, 116–17, fig. 67; 192–6, nos. 5–7, figs. 113, 114, 118, 119); one house had mosaics in both techniques (Mazzei 1995, nos. 4–6). Greek-style pebble mosaics are rare in Italy, and their adoption in this particular region might be related to the existence of a local tradition of pebble mosaic, which is stylistically and functionally distinct from Greek mosaics.20 The houses also have Masonry Style wall plaster.
The excavators have dated the houses to the period of the settlement’s greatest prosperity, in the late fourth and third centuries. This seems surprisingly early to find such luxurious houses and such technically advanced mosaics; the archaeological evidence for the date has not yet been published in detail, but similar rough tessellated pavements elsewhere in southern Italy have been said to date from the same period,21 and it is possible that independent local improvisation produced an early form of tessellation, as suggested above. What is more certain is that these houses and their decoration are the culmination of a process of hellenisation in the region, led by the local elite, who had begun to build houses with Greek-style architectural decoration as early as the fifth century, presumably to set themselves apart from their less wealthy neighbours, who were still living in huts. By the early hellenistic period, more solidly-built houses had become the norm at all levels of society, but most were still relatively small and simple in plan, and the large peristyle houses found at Arpi and elsewhere in the region represent a significant improvement in living standards on the part of the wealthiest members of the community. Connections have been drawn with fourth-century houses on the Greek mainland, and especially in Macedonia, in view of the parallel adoption of Macedonian-style chamber-tombs in the region (Russo Tagliente 1992, 146–51). But the process of hellenization in Daunia goes beyond the imitation of Greek architectural forms and decoration to include the adoption of at least some elements of a Greek lifestyle. Several of the mosaics at Arpi are in Greek-style andrōnes, with raised borders for couches, and this, along with the presence of imported Greek drinking-wares in both habitation sites and burials, suggests that the custom of the symposium was adopted along with the architectural forms.
Fig. 15. Arpi: pebble mosaic.
In contrast, the inhabitants of late hellenistic Pompeii seem to have had somewhat different motives for acquiring Greek mosaics. The earliest mosaics at Pompeii are contemporary with the First Style of wall painting (itself a version of the Greek Masonry Style), in the late second and early first century; they appear in only a small minority of wealthy houses, at a time when the standard types of pavement were fairly modest concrete-like surfaces such as opus signinum, lavapesta (mortar containing crushed lava) and various grades of terrazzo.22 Unlike the Daunians, Pompeian customers did not opt for the classic Greek scheme of concentric decoration covering the whole floor; instead, they seem to have been interested solely in the figured panels, which are found in only a small minority of Greek mosaics. The earliest decorated mosaics consist of emblēmata set into otherwise plain floors (Fig. 16). The decorative framework which is the principal element of mosaics at Greek sites is omitted almost entirely (Westgate 2000a, 264–5).
Fig. 16. Pompeii: emblema with marine creatures, from the House of the Faun (Naples, Museo Nazionale no. 9997).
It is not difficult to understand the motives behind this very selective adoption of Greek mosaics. Almost all of the figured scenes are repeated in other mosaics from Pompeii and elsewhere in Italy, and they are often thought to be copies of famous Greek paintings. Although on closer examination many of them appear to be pastiches or genre scenes rather than literal copies, they were surely all acquired with the intention of displaying an appreciation of Greek art; they clearly form part of the Roman fashion for copies and imitations of Greek art in the late hellenistic period. Moreover, the subjects often seem to have been chosen to advertise the owner’s familiarity with other aspects of Greek culture, as they often depict mythology, drama – usually comedy – and even philosophy.23 It was already common for the minutely detailed figural panels to be prefabricated off-site for insertion into the pavement, but at Pompeii they seem more like independent pictures than ever before: mosaicists working for this Italian market started to make emblēmata on trays or backing-plates of stone or terracotta, which presumably made them safer to transport, sell and re-use (Westgate 1999; 2000a, 266–73).
Pompeii stands at the beginning of the long and rich Roman tradition of mosaics. The role of the Greeks in initiating this tradition has often been underestimated: mosaic was a medium through which the inhabitants of the hellenistic world advertised their wealth and taste, expressed their aspirations, reinforced their sense of Greek identity, or presented a fashionable image of cosmopolitan sophistication and culture.
Acknowledgements
I would like to thank Manchester University, the British Schools at Athens and Rome and the British Institute of Archaeology at Ankara for funding the research summarized here, and Roger Ling for reading a draft of the paper; any errors that remain are entirely my own. This paper was written during the tenure of a fellowship from the Arts and Humanities Research Board.
Notes
1 All the pebble mosaics discovered up to c. 1980 are catalogued and studied by Salzmann (1982); hellenistic tessellated mosaics are treated in regional surveys by Bruneau (1972), Baldassare (1976), von Boeselager (1983), Daszewski (1985) and Bingöl (1997), while Anne-Marie Guimier-Sorbets has published numerous articles on individual aspects of motif, iconography, design and technique. Katherine Dunbabin’s recent synthesis of ancient mosaics provides an excellent overview (1999, especially chapters 1–3). This article is based on my doctoral thesis (Westgate 1995), a comprehensive study of mosaics from their classical origins to the late hellenistic period, which is currently being revised for publication in the series Oxford Monographs on Classical Archaeology.
2 The House of the Rape of Helen covers 2350 m2, and the House of Dionysos 3160 m2; by comparison, the unusually luxurious House of the Mosaics at Eretria has an area of only 650 m2.
3 Compare Susan Rotroff’s explanation of the decline in the number of pottery kraters found in hellenistic contexts in the Athenian Agora: she argues that now lavish metal vessels would have served as the centrepieces of grandiose banquets (1996, 22–7).
4 These total about 70, nearly all from sites on the Greek mainland, notably Athens, Corinth, Sikyon, Eretria and Olynthos.
5 The green beads are used in Dionysos’ crown and thyrsos, and blue glass in the harness of Theseus’ chariot-horses. There are no artificial elements at all in the Stag Hunt and Amazonomachy.
6 Olynthos, House A5, andron a (Robinson 1930, 56–9, figs 153–161, pl. i); Vergina, palace, threshold of room E (leading to a pebble mosaic, Salzmann 1982, no. 130).
7 House of Many Colours, andron d, pavement of white chips with a border of black pebbles (Robinson 1946, 193, pls. 159, 160.2, 165); Sector 7, House B, pavement of polychrome chips bordered by a meander in pebbles (Robinson 1930, 102, figs. 237, 239; Salzmann 1982, no. 92, pl. 16.5).
8 A fourth-century house excavated between rues Didon and Arnobe at Carthage had a pavement of large terracotta cubes (opusfiglinum) decorated with a chequered stripe of black, white and red tesserae: Dunbabin 1994, fig. 15. A tiny fragment of red and white chequerboard from Kerkouane (Morel 1969, 499–500, fig. 28) must date before the site was abandoned in the mid-third century, although the fifth-century date claimed by the excavator seems doubtful.
9 On Delos, for instance, there are several pavements of stone chips with simple decoration in broken terracotta pieces (Bruneau 1972, nos. 41, 135, 221, 254, 260); Salzmann excludes these from his catalogue of ‘intermediate-type’ mosaics, although they are similar in technique to the examples from Athens and Aphrodisias cited above.
10 e.g. by Daszewski (1985, 98), although he stops short of attributing the invention exclusively to Alexandria. Recent excavations in the garden of the old British Consulate have uncovered fragments of a pebble mosaic (Guimier-Sorbets 1998a, 189), and two mosaics of mixed technique (Guimier-Sorbets 1998a, 188–9; 1998c, 227, fig. 5), at least one of which probably dates to the first half of the third century; two other mixed mosaics were already known (Daszewski 1985, nos. 1 and 2), as well as a number of tessellated mosaics (Daszewski 1985, nos. 5–7, 13–19).
11 It is clear from a mosaic on Samos, whose surface is partly destroyed, that the strips were slotted into guidelines incised in the layer of mortar below the surface, to serve as outlines for the mosaicist to fill in: Giannouli and Guimier-Sorbets 1988, 558, fig. 7. Lead strips are normally only used in figural motifs where a precise curve or straight line was required, such as the mast and yard-arm held by a female figure who probably represents Alexandria (Alexandria, Greco-Roman Museum, inv. no. 21739: Daszewski 1985, no. 38, col. pl. A), or the rings of an armillary sphere on a mosaic in the Casa di Leda at Soluntum in Sicily (von Boeselager 1983, 56–60, pl. xv.29–30).
12 Unless Donderer (1991) is right in arguing that the famous version from Hadrian’s Villa at Tivoli, now in the Capitoline Museum, is in fact the original.
13 Kawerau and Wiegand 1930, 53–65, figs. 39, 69–72; text pls. xxvi–xxxix; pls. viii–xix. Parts of the two complete pavements, heavily restored, are now in the Pergamon Museum in Berlin (Inv. mos. 68–71), along with some of the fragments; but recent re-excavation has shown that substantial remains of the mosaics are still in situ, and that the reconstructions are incorrect (Salzmann 1995).
14 A similar theory was proposed, for different reasons, by R. Vollkommer (1990), who argued that opus vermiculatum was invented first, and opus tessellatum was developed later as an imitation.
15 Rooms E, G, H, I: Bruneau 1972, 240–60, nos. 214–17, figs. 177–210. Blue glass is used in the figured scenes, but whereas glass tesserae are usually cut from purpose-made rods, here pieces of a broken bowl are used, which suggests improvisation (Guimier-Sorbets and Nenna 1992, 623).
16 Explored by Davidson (1997, 3–20). For fish mosaics, see Meyboom 1977; for birds, Tammisto 1997.
17 These arguments are presented in more detail in Westgate 1997–8; 2000b, 424–6; forthcoming.
18 Trümper 1998, 92–106. The fragments of paintings and mosaics from upstairs rooms tend to be finer and more colourful than those preserved on the ground floors (Bruneau 1972, 64, 67–8, 105–6).
19 The mosaics are in the cloakroom of Unit II and the bathroom of Unit IV respectively; a simpler pavement in the water-heating room of Unit IV, with a design of lines forming rectangles (Salzmann 1982, no. 4, pl. 71.1), is probably imitating paving of stone slabs.
20 e.g. Salzmann 1982, nos. 57–60, pl. 65, at Herdonia. The pavements are decorated with monochrome texture patterns formed by laying long pebbles on edge at different angles; the effect is rather like woven matting. It is an outdoor type of paving, most commonly found in cemeteries. The only other Greek-style pebble mosaics in Magna Graecia are at Taras (Salzmann 1982, no. 124, pl. 66.3), Metapontion (Salzmann 1982, no. S.1, pl. 66.2) and Motya (Salzmann 1982, no. 72, pl. 69).
21 For example at Volcei (destroyed 280 BC)and Elea.
22 The most comprehensive studies of the Pompeian pavements remain those of Blake (1930) and Pernice (1938); see also the recent summary by De Vos (1991). The figured panels in Naples Museum are catalogued in Pozzi 1989, 116–23.
23 Compare Ada Cohen’s reading of the most spectacular hellenistic mosaic of all, the Alexander Mosaic from the House of the Faun (Cohen 1997).
Baldassare, I.
| 1976 | ‘Mosaici ellenistici a Cirene e a Delo: rapporti e differenze’, Quaderni di Archeologia della Libia 8, 193–221. |
| 1993 | ‘La decorazione pavimentale: le tipologie più antiche e la introduzione del tessellato’, in R. Farioli Campanati (ed.) Atti del 1o Colloquio dell’Associazione Italiana per lo Studio e la Conservazione del Mosaico, Ravenna, 29 aprile–3 maggio 1993, Ravenna, 435–50. |
Bingöl, O.
| 1997 | Malerei und Mosaik der Antike in der Türkei. Kulturgeschichte der antiken Welt 67, Mainz. |
Blake, M.
| 1930 | ‘The pavements of the Roman buildings of the Republic and early Empire’, Memoirs of the American Academy in Rome 8, 7–159. |
Boeselager, D. von
| 1983 Antike Mosaiken in Sizilien. Hellenismus und römische Kaiserzeit, 3. Jahrhundert v. Chr. – 3. Jahrhundert n. Chr., Rome. |
Bruneau, Ph.
| 1972 | Exploration archéologique de Délos XXIX: Les mosaïques, Paris. |
| 1987 | La mosaïque antique. Lectures en Sorbonne 1, Paris. |
Chamonard, J.
| 1922 | Exploration archéologique de Délos VIII: Le Quartier du Théâtre, Paris. |
Cohen, A.
| 1997 | The Alexander Mosaic: Stories of victory and defeat, Cambridge. |
Daszewski, W.A.
| 1983 | Corpus of Mosaics from Egypt I. Aegyptiaca Treverensia 3, Mainz. |
Davidson, J.
| 1997 | Courtesans and Fishcakes: The consuming passions of Classical Athens, London. |
De Vos, M.
| 1987 | ‘Paving techniques at Pompeii’, Archaeological News 16, 36–60. |
Donderer, M.
| 1991 | ‘Das kapitolinische Taubenmosaik – Original des Sosos?’, Mitteilungen des deutschen archäologischen Instituts, Römische Abteilung 98, 189–97. |
Dörpfeld, W.
| 1912 | ‘Die Arbeiten zu Pergamon 1910–1911’, Mitteilungen des deutschen archäologischen Instituts, Athenische Abteilung 37, 233–408. |
Dunbabin, K.M.D.
| 1977 | ‘Technique and materials of hellenistic mosaics’, American Journal of Archaeology 83, 265–77. |
| 1994 | ‘Early pavement types in the west and the invention of tessellation’, in P. Johnson, R.J. Ling and D.J. Smith (eds.) Fifth International Colloquium on Ancient Mosaics, held at Bath, England, on September 5–12, 1987: Part I. Journal of Roman Archaeology Supplementary Series 9, Ann Arbor, 26–40. |
| 1999 | Mosaics of the Greek and Roman World, Cambridge. |
Giannouli, V. and Guimier-Sorbets, A.-M.
| 1988 | ‘Deux mosaïques hellénistiques à Samos’, Bulletin de Correspondance Hellénique 112, 545–68. |
Guimier-Sorbets, A.-M.
| 1998a | ‘Les mosaïques d’Alexandrie: nouvelles découvertes’, Revue Archéologique, 188–93. |
| 1998b | ‘Alexandrie: les mosaïques hellénistiques découvertes sur le terrain de la nouvelle Bibliotheca Alexandrina’, Revue Archéologique, 263–90. |
| 1998c | ‘Les mosaïques’, in M. Rausch (ed.) La Gloire d’Alexandrie, Paris, 227–30. |
Guimier-Sorbets, A.-M., and Nenna, M.-D.
| 1992 | ‘L’emploi du verre, de la faïence et de la peinture dans les mosaïques de Délos’, Bulletin de Correspondance Hellénique 116, 607–31. |
Kawerau, G., and Wiegand, T.
| 1930 | Altertümer von Pergamon V.1: Die Paläste der Hochburg, Berlin. |
Kreeb, M.
| 1986 | Untersuchungen zur figürlichen Ausstattung delischer Privathäuser, Chicago. |
Lloyd, J.A. et al.
| 1998 | ‘Excavations at Euesperides (Benghazi): an interim report on the 1998 season’, Libyan Studies 29, 145–68. |
Makaronas, Ch., and Giouri, E.
| 1987 | Οἱ Οἰκίες Ἁρπαγῆς τῆς Ἑλένης καὶ Διονύσου τῆς Πέλλας. Βιβλιοθήκη τῆς ἐν Ἀθήναις Ἀρχαιολογικῆς Ἑταιρείας 109, Athens. |
Mazzei, M.
| 1995 | Arpi: L’ipogeo della Medusa e la necropoli, Bari. |
Meyboom, P.G.P.
| 1977 | ‘I mosaici pompeiani con figure di pesci’, Mededelingen van het Nederlands Historisch Instituut te Rome 39, 49–93. |
Morel, J.P.
| 1969 | ‘Kerkouane, ville punique du Cap Bon: remarques archéologiques et historiques’, Mélanges d’Archéologie et d’Histoire de l’École Française de Rome, Antiquité 81, 473–518. |
Pernice, E.
| 1938 | Die hellenistische Kunst in Pompeji VI: Pavimente und figürliche Mosaiken, Berlin. |
Phillips, K.M.
| 1960 | ‘Subject and technique in hellenistic-Roman mosaics: a Ganymede mosaic from Sicily’, Art Bulletin 42, 243–62. |
Pollitt, J.J.
| 1984 | Art in the Hellenistic Age, Cambridge. |
Pozzi, E. (ed.)
| 1989 | Le Collezioni del Museo Nazionale di Napoli I.1: I mosaici, le pitture, gli oggetti di uso quotidiano, gli argenti, le terracotte invetriate, i vetri, i cristalli, gli avori, Rome. |
| 1990 | ‘Greeks in Afghanistan: Aï Khanum’, in J.-P. Descoeudres (ed.) Greek Colonists and Native Populations. Proceedings of the First Australian Congress of Classical Archaeology, held in honour of Emeritus Professor A.D. Trendall, Sydney, 9–14 July 1985, Oxford, 329–42. |
Robertson, C.M.
| 1965 | ‘Greek mosaics’, Journal of Hellenic Studies 85, 72–89. |
Robinson, D.M.
| 1930 | Excavations at Olynthus, Part II: Architecture and sculpture: Houses and other buildings, Baltimore. |
| 1933 | Excavations at Olynthus, Part V: Mosaics, vases and lamps of Olynthus found in 1928 and 1931, Baltimore. |
| 1946 | Excavations at Olynthus, Part XII: Domestic and public architecture, Baltimore. |
Rotroff, S.I.
| 1996 | The Missing Krater and the Hellenistic Symposium: Drinking in the age of Alexander the Great. Broadhead Classical Lecture No. 7, Christchurch. |
Russo Tagliente, A.
| 1988 | Edilizia domestica in Apulia e Lucania: Ellenizzazione e società nella tipologia abitativa indigena tra VIII e III secolo a.C., Galatina. |
Salzmann, D.
| 1982 | Untersuchungen zu den antiken Kieselmosaiken von den Anfängen bis zum Beginn der Tesseratechnik. Archäologische Forschungen 10, Berlin. |
| 1994 | ‘Zu den Mosaiken in den Palästen IV und V von Pergamon’, Studien zum antiken Kleinasien III. Asia Minor Studien 16, Bonn, 101–12. |
Tammisto, A.
| 1997 | Birds in Mosaics: A study on the representation of birds in Hellenistic and Romano-Campanian tessellated mosaics to the Early Augustan Age. Acta Instituti Romani Finlandiae 18, Rome. |
Trümper, M.
| 1998 | Wohnen in Delos: Eine baugeschichtliche Untersuchung zum Wandel der Wohnkultur in hellenistischer Zeit. Internationale Archäologie 46, Rahden. |
Tsakirgis, B.
| 1989 | ‘The decorated pavements of Morgantina I: the mosaics’, American Journal of Archaeology 93, 395–416. |
Vollkommer, R.
| 1990 | ‘Zum Ursprung des Opus Tessellatum’, Akten des XIII. Internationalen Kongresses für Klassische Archäologie, Berlin 1988, Mainz, 571–72. |
Walter-Karydi, E.
| 1994 | Die Nobilitierung des Wohnhauses: Lebensform und Architektur im spätklassischen Griechenland. Xenia 35, Konstanz. |
Westgate, R.C.