5. Doric Figured Supports:
Vitruvius’ Caryatids and Atlantes
A number of different figures used as architectural supports or creating the illusion of supporting can be found within the architectural sculpture of the Late Classical and Hellenistic periods. The simplest way of dividing these into types is by their pose rather than their gender.
Atlantes or Telamones provide support on the nape of their necks and on the forearms, their arms bent back with their elbows thrust forwards, and their pose expresses their function. Though the original iconography of the figures was male, from the early Hellenistic period females, as well as differently characterised males, began to appear; the form was adapted to suit the designers’ needs, most commonly serving to represent the adherents of Dionysos in the decoration of theatres.
Caryatids are often associated with figures carved in the round and replacing columns, which, as on the Erechtheion, were to be found only on Ionic buildings. They were always female, and the figures never expressed the burden of their supporting function (the Erechtheion maidens for example have one bent and ‘relaxed’ leg). At the end of the Hellenistic era a variant developed, where they became cistophoroi, as on the Inner Propylon at Eleusis; this form continued to be used in the Roman period.
Alternatively Caryatids have been identified as figures supporting superstructures with the poloi on their heads and one or two of their raised hands who were to be found within Doric contexts, although later this use was less rigidly applied. Extant examples, almost all female, are known from the beginning of the Hellenistic period and continue through to the time of the Antonines. Though generally depicted in relief there is an example of such figures carved in the round; they could be placed between columns or take their place.
Vitruvius, a Republican Roman architect with an enthusiasm for Hellenistic taste who wrote a treatise at the turn from the Hellenistic to Roman Imperial periods, described both Atlantes and Caryatids in his treatise on architecture. This would suggest that they were by then established and acknowledged forms of decoration, and that one should look to sculptures preceding his time to identify and examine their forms.
5.1 Vitruvius’ Caryatids
It has become common practice to refer to all female figured supports as Caryatids, the most famous examples being those of the South Porch of the Erechtheion in Athens. This assumption stems from a passage of Vitruvius (De Architectura I. 1. 5), but if one examines the text of Vitruvius it soon becomes apparent that he had a specific type in mind. There are relatively few mentions of the term in ancient literature, and its use as a generic term can be traced back to the eighteenth century, to Stuart and Revett, and to Winckelmann. A number of works on figured supports have appeared, roughly dividing the figures by sex, then by pose, and concluding that female figures or ‘Caryatids’ were to be found within the framework of the Ionic order, and male, known as Atlantes or Telamones, within the Doric. By reexamining these, and placing them in the context of recent discoveries, one can shed light on the controversy.
To begin one should distinguish between the two types of sculpted Caryatids in literature. The first are Vitruvius’ figures of women in submission used as architectural sculpture. The second are dancers that formed part of a group by Praxiteles, mentioned by Pliny N.H. XXXVI.23 as Caryatids or Thyades, the Delphic name for the Maenad followers of Dionysus, almost certainly a Choregic monument, and considered by some to be reproduced in the acanthus column dancers at Delphi. There are a number of sources for the Caryatids who served as dancers at the famous shrine of Artemis Caryatis. Pliny uses the term in reference to the decoration of the Pantheon of Agrippa built by Diogenes of Athens. For the first type of submissive Caryatid, Vitruvius’ account is the key text:
Architects ought to be familiar with history because in their works they often design many ornaments about which they ought to render an account to inquirers. For example, if anyone in his work sets up, instead of columns, marble statues of long-robed women which are called Caryatids, and places mutules and cornices above them, he will thus render an account to inquirers. Caria, a Peloponnesian state, conspired with the Persian enemy against Greece. Afterwards the Greeks, gloriously freed from war by their victory, with common purpose went on to declare war on the inhabitants of Caria. The town was captured; the men were killed; the state was humiliated. Their matrons were led away into slavery and were not allowed to lay aside their draperies and ornaments. In this way, and not at one time alone, were they led in triumph. Their slavery was an eternal warning. Insult crushed them. They seemed to pay a penalty for their fellow-citizens. And so the architects of that time designed for public buildings figures of matrons placed to carry burdens; in order that the punishment of the sin of the Caryatid women might be known to posterity and historically recorded.
[Vitruvius, De Architectura, 1. 1. 5: Loeb trans. F. Granger]
Vitruvius gives the story of the Caryatids as a digression on the necessity of architects knowing some History. Unfortunately he himself seems to have made mistakes, and thus the passage is open to interpretation. It has been suggested by some that Caria, a Persian Satrapy in modern-day Turkey, is meant as the origin of the women, which would better explain their Medising. Caria however is definitely not in the Peloponnese, and it is hard specifically to pinpoint the humiliation that the Greeks inflicted on it. The state is more likely to have been Caryae, a city in Laconia visited by Pausanias (III, 10, 7); one should then however point out the inconsistency in Vitruvius’ dating of its destruction to the years following the Persian war. What is known of Caryae is that, according to Xenophon, Hellenica 6.5.25 & 7.1.28, it was destroyed in 370/369 by the Spartans during Laconian-Theban wars after Leuctra.
Plommer pointed out that Vitruvius’ inclusion of the use of mutules, which are to be found beneath triglyphs, indicates that the figures were placed within the context of the Doric rather than the Ionic order. One can also read into the passage that the figures were ‘burdened’ and physically supporting the superstructure, presumably with their arms, atoning for their sins and not being honoured. They replaced columns and so literally supported the superstructure, but must also have been seen to do so in terms of their pose, bringing in an element of trompe l’œuil.
This makes improbable the inclusion within the Caryatid category of such figures as the Erechtheion maidens, called in the building inscriptions Korai, the figures from the Delphi treasuries, and those from the Heroon of Pericle at Limyra. These figures are honoured rather than dishonoured, and set into an Ionic framework, with the weight of the superstructure resting solely on their heads. They take the place of columns, perhaps as an evolution of columnae caelatae. The fame of the figures was such that had Vitruvius been describing the Erechtheion Korai, it is likely that he would have named them. The Cistophoroi from the interior of the Inner Propylon at Eleusis, built in the twilight of the Hellenistic era, were variants upon these, their raised arms supporting the baskets they carried on their heads, which in turn supported the entablature; they were copied in the second century AD for a building at Monte Porzio. The two sets of figures from the Villa of Herodes Atticus on the Via Appia, and their copies from a monument, probably to Demeter, found near the Mitropolis in Athens, are also to be seen as a development of such figures.
As an example of the pose of the figures believed to have been those that Vitruvius referred to as Caryatids, Plommer pointed out the existence of a Hadrianic marble relief, ht. ca. 0.87, Naples Museum inv. 6715 / no. 149, from Pozzuoli, that represents a Classical structure whose entablature is held up by two female figures using their poloi and the palms of their raised hands. In the centre is a mourning figure not dissimilar to the Persepolis Penelope, a type used in Roman art to represent conquered regions. The entablature is not represented as Doric, but this may be because the relief is schematised, or perhaps because by that time the Doric context was no longer applied. Its inscription, XX XXXXXX XX XXXXXXX XXX, would suggest that it represented the elusive Caryatid Monument, but the problem in identifying it as such lay in the fact that, as Plommer pointed out, other than the Athena on an early Classical Atlas metope at Olympia, no examples were then known of figures in this pose. Subsequently however three Doric tombs, and a number of other sculptures, have been identified, and these provide strong evidence in favour of the Caryatid Monument having had such figures, supporting the superstructure with their palms and poloi. The chronology of these tombs is uncertain, but they all belong to the Hellenistic period.
The first of these is a rock-cut tomb at Aghia Triadha, near Rhodes (fig. 60). A square ground-plan was formed by four klinai, at the heads of each of which stood statues of women, carved in the round, and crowned by the circular Doric entablature from which sprang a dome. The figures are highly fragmentary, but preliminary restoration would seem to indicate that they supported the superstructure with alternating raised palms and the poloi on their heads. Their dress, consisting of long peploi, also conforms to the implications of Vitruvius’ textual source. The presence of a dome leads one to assume a date in the later Hellenistic period, and publication of the archaeologists’ evidence for dating is eagerly awaited.
The second example is better preserved, a Thracian tomb near Svestari in modern Bulgaria. It dates from soon after 300 BC. The frontal figures, located in the main chamber, are in high relief, but otherwise they fit Vitruvius’ description. Cut into the limestone, they were fully painted, with much of the pigment remaining; details were picked out in ochre, dark brown, blue, red and lilac. The figures stand on ledges between Doric half-columns, and hold up a Doric entablature. The women wear long chitons, the high-girdled overfolds of which are heavily stylised and have been subject to local stylistic variations, turning into three acanthus leaves, below which the long skirts continue, moulding the legs; the shoes protrude from the bottom and are visible. There were ten of these figures in all, standing 1.20 m high: four on the north wall opposite the door, and three on the two side walls. The south wall, where the door was located, was undecorated.
The proportions of the women vary slightly but all are in roughly the same pose. The corner figures have only their inner arms raised, the central figures both. As on the Naples relief and in the Rhodian tomb, they bear the architrave on their hands and poloi. Of particular interest are the faces, with expressions that appear to represent pain and grief, which are quite unusual in Greek sculpture. Their features are all differentiated and highly individual, with a variety of ages, and they appear to be portraits rather than idealised figures. One might note especially the northernmost figure of the west wall, whose head is turned down in a look of great despair. The tomb was built under the influence of Macedonian art. Although no examples of Caryatids have been found amongst the rather limited architectural sculpture of these, they do make an interesting appearance on the throne in the so-called Tomb of Eurydice at Vergina, suggesting a possible chain of influence. Here there were figures around three sides of the seat, supporting the armrests and the painted panel that formed the back; male figures with raised right arms alternate with female ones whose left arms are raised, both using one of their palms to effect the support. A number of the figures are now missing.
Tomb N 228, Cyrene, is a rock-cut facade tomb, 6.93 wide, built some time during the later Hellenistic period, probably ca. 150-50 BC. At the corners of the facade there are Ionic quarter-columns engaged to pilasters; the whole is however crowned by an undecorated Doric frieze, making this a structure of mixed order. In the centre, between the doors, there are two Caryatids that supported the frieze with their poloi and both raised palms. The figures are carved in relief, and not fully depicted, turning into engaged half-columns with Doric fluting below the knees. The figures are highly Classicising; they seem to wear chitons and the way that these are represented, particularly the apoptygmata, is very much late Classical in style. The figures are quite worn, and the tomb has not yet been fully excavated, but the figures, other than in their dress, appear to conform to Vitruvius’ prescription.
The Sanctuary of Fortuna at Praeneste, modern Palestrina, is depicted in an engraving of Henry of Cleeves, who went to Italy in 1551 and saw the site before the Palazzo Barberini was built on it in 1640. A series of Caryatids, seemingly in relief, is depicted in the engraving, encircling the crowning tholos. Whether these figures were the artist’s fantasy or actually existed is debatable, with scholars tending to believe the latter. The figures as represented are surprisingly similar to those on the Vergina throne in the way that they stepped forwards, one arm raised, with billowing apoptygmata. This sort of energetic supporting figure, combining elements axiomatically linked to Nikai figures, was particularly popular in the depiction of Caryatids on Roman sarcophagi. The figures at Palestrina would most likely have backed on to piers, supporting the entablature with their heads and the palm of one hand, and were probably carved of the same white marble as the rest of the adornment of the sanctuary. It is unclear whether the spaces between the figures were solid or open. A sculptured figure with these characteristics is on display in Room II on the ground floor of the site museum; a female figure, inv. no. 135, ht. 1.60, she wears a crinkly chiton and chlamys. Much of the piece is missing, but from the angle of the shoulders one arm appears to have been raised. The back and sides are flat, as if inserted into a wall, and have square cuttings for dowels to have held it into place. She is dated stylistically by the curators to the end of the second century BC, a date not far removed from that of the building of the upper sanctuary, which is dated by Fasolo and Gullini to the mid second century BC.
A pair of Caryatids from Mylasa, in Caria, were found in conjunction with a Doric frieze and a frieze with masks, and are now in Istanbul. At the time, owing to the mask frieze that was found, this was thought to indicate the location of the city’s theatre; but a theatre has since been found elsewhere in the town, and it is now believed that they might have decorated an elaborate funerary monument of the later second or early first centuries BC. Both masks and figured supports tend to be part of theatrical iconography, but masks could also be used on tombs, and Vitruvian Caryatids almost exclusively were. The two figures formed a pair mirroring each other’s poses. The inserted heads, which were held in place by dowels, are lost. The arms are also missing, but their poses are easy to reconstruct; one arm was raised out to one side, the elbow bent and the palm held upwards, while the other arm was down. Carved of small-grained white marble, with a blue tint in places, the peplos-clad figures have flattened backs, showing that they were placed against a wall. The figures are almost certainly architectural, their heavy bodies, despite having had pieced heads, allowing for the illusion of bearing weight; both figures show signs of weathering. They are likely to have worn poloi, supporting the architrave on these and on one raised hand; it is tempting to suggest that the lost Doric frieze was from the entablature sustained, fulfilling the Vitruvian prescription.
A further type of Caryatid, with archaising and Severe style details, is known through copies surviving at Athens, Tralles and Cherchel; it appears to be an eclectic work of ca. 100 BC, rather than a Classical original, that was copied as late as the Antonine period. The type is known as the Tralles-Cherchel Caryatid, after its most famous known replicas. As with Vitruvian Caryatids the figures could have provided support on their poloi and the palm of one raised arm; those from Tralles may have been architectural, but the others were free-standing copies.
The oldest, and first discovered, figure was found at the Theatre at Tralles, and is now in the Archaeological Museum in Istanbul, no. 1189, ht. 1.86 incl. 0.03 plinth. A head from the figure’s pair in the Ecole Evangélique in Smyrna was destroyed in a fire and is now only preserved in a photograph; part of the right shoulder was preserved and showed that this was raised. Stylistically they date from the first half of the first century BC. Carved in a small-grained white marble, the back was summarily worked suggesting that the figure was placed against a wall; the left arm of the extant statue was pieced below the shoulder, and is lost. Also missing are the bottom of the plinth, a few fingers of the right hand, besides which there are numerous chips. It was highly painted, traces of red remaining on the face and blue on the polos. The figure wore a finely pleated short-sleeved chiton and a cloak folded in two and wrapped around her body and over the left shoulder. Although the figures were found in or by the theatre, there is insufficient evidence conclusively to state that they formed an integral part of its decoration; by this late date such Caryatids may no longer have been reserved purely for funerary structures, as they appear to have been earlier in the Hellenistic period, but the use of those that both pre- and postdated them would suggest that might still have been the case. Interestingly Apaturius of Alabanda, according to the testimony of Vitruvius, had shown statues of centaurs supporting architraves as part of the scenery he had designed for the small theatre at Tralles.
A version of the figure type was found in the palace of Juba II at Cherchel, the ancient Caesarea of Mauritania, and dates to the 20s BC. A variant rather than an exact copy, it was a cruder work, provincial and over-exaggerated. The head and upper neck are broken off, the arms missing below the shoulders. The left arm was raised. Two heads and the fragmentary pieces of their bodies, from Hadrianic or Antonine copies, were found on the Athenian Acropolis. The number of examples of the type, their wide geographical spread and the variations in their dates, would suggest that they are replicas of well known originals. Their style suggests that this was an eclectic work of the later Hellenistic period rather than the eponymous Caryatid Monument.
A number of high relief Caryatids from Miletos Theatre, that decorated the stage-front, are known; these appear to be divided into two sets, one Hellenistic and one Roman, which employed Caryatid-like figures, with one raised arm to support, in the place where figures in Atlantes-poses were generally found. One of the supporting archaistic figures from Miletos differs greatly from the other two in the Louvre, and Linfert would like to redate it to the third century, as a piece that was then in good condition and reused in the second century AD Roman rebuilding. This idea is much disputed, but the differences between Louvre Ma 2793 and 2794 are too great to be ignored. Although the consensus leans towards a Roman date for both, one can make a good case against it. The dating of the Hellenistic architectural sculptures is generally put at ca. 300 or in the early third century, so as to be contemporary to, if not preceding, the majority of the examples in the West; but the consensus is that the colonists were the innovators, and that the Miletos figures should be dated to a reworking of the mid-second century BC or after.
The figure believed to be Hellenistic, Louvre Ma 2793, ht. 1.65, is made up of three main fragments; the torso, the upper thighs, and the knees and calves. The arms, head and feet are missing. Her back is flat, with anathyrosis. She wears a chiton, with a roll of drapery over the torso, as on the Mylasa figures, emphasising it. A strong vertical line is created between her legs by two main folds and smaller ones on the sides, but it is not as pronounced as examples from Magna Graecia. Although archaising, it clings to many more Classicising forms. The position of the button on her left shoulder emphasises that that arm was raised; the top of the right shoulder shows that it was down. The lowered arm is on the side of the leg set forward. There is a severe lack of modelling on the legs, and the anatomy is generally nebulous. The vertical folds begin below the overhang. Linfert draws parallels between this figure, which he believes to be early Hellenistic, and the sculptured column drums from Ephesos, notably BM 1200.
The Roman figure, Louvre Ma 2794, ht. 1.41, has rather different drapery, its chiton much more wispy, almost transparent under the peplos. The roll of the cloak is different, and the vertical folds are much more emphasised: they jut out and begin higher up, from the level of the roll. There is also more contrast within the elements of the drapery, which is more deeply cut. She is made up of two main fragments, the upper torso and the rest, and her back is rougher; the arms, the head, the knees and below are missing. A socket also shows that the head and neck were originally inserted. Louvre Ma 2795, ht. 1.90, is its pair. The differences, as if resulting from the sculptor of one figure trying to emulate the other but failing, cannot be brushed off merely as characterisation.
Three more figures have been linked to the Hellenistic decoration of the theatre. A torso in the garden of the Basmane Museum in Izmir, inv. 74, ht. 0.90, appears to be of the same design as the Louvre Hellenistic figure, and forms a pair with it. A second figure in Izmir, in the Archaeological Museum, unpublished, is of a different design, and presumably one of the Roman figures. Heres has added a figure in the Pergamon Museum in Berlin to the corpus; the figure is of the same design as the Hellenistic Caryatids, and again would seem to form a pair to Louvre Ma 2793. The evidence suggests that at least two pairs of figures decorated the Hellenistic stage-front of the theatre at Miletos, and that these were reused, alongside contemporary figures imitating them in the Roman rebuilding.
The majority of these examples of Caryatids would appear to come from funerary contexts, so whatever the original pejorative intention of the figures when the Caryatid Monument was set up, the form soon took on overtones of mourning. As well as large-scale examples used as sculptural decoration on buildings, there are a number of small-scale depictions of the pose. One example comes from a Messapian tomb at Vaste, near Lecce, from the last quarter of the third century. A large female with both forearms raised, in the Atlantes pose, supports a battle frieze, ht. 0.30. The pose is of the theatrical type though in more sombre dress, and can be compared to the two mourning supports from a Tarentine naiskos. Of more interest is a small and badly worn low relief female figure at the end of the frieze, which wears a polos and peplos, and supports the moulding above the frieze with one raised hand, once again recalling the Neapolitan relief (fig. 59). A block of a Samian frieze is preserved, dating from the middle Hellenistic period. Five erotes at play are depicted, with on the right end an archaising Caryatid, providing support with her polos and the raised palm of her outer hand. Presumably there would have been a similar figure at the opposite ends of both friezes. Very similar small figures can be found on a number of late Hellenistic funerary stelae, now in Istanbul, and on either side of the doorway on a late fourth century or early Hellenistic rock-cut tomb at Limyra. The upper section of a Hellenistic Boeotian Stele preserves the pediment and entablature; the outer two metopes are filled with rosettes, the inner two with phialai, and in between them are two small figures in long robes, supporting the top of the metope with their head and raised arms. The figures are used in a funerary context, and although their location within the architectural framework has changed, in that they do not support the frieze but rather are placed within it, these would appear to be miniature, symbolic Caryatids.
Interestingly an unpublished late Republican box found at Baiae/Cumae and now on display in Naples Museo Nazionale, was decorated with ivory plaques depicting small Caryatid figures. The inner figures had both arms raised, supporting the frame of the reliefs with their heads and hands, whilst the women on the outer plaques had only one hand in the air. The box was found placed in a tomb.
Relief figures of women were carved on to the piers that supported the arches on the lower level of the Monument of Caius Memmius that stands on the corner of Kouretes and Domitian Streets at Ephesos. That the structure was built by Memmius, born in 70 and suffect consul in 34, is attested by an inscription. Possibly the repository of his ashes, but more likely to be a propagandist monument or cenotaph, it is a structure in the Corinthian order, built for a Roman following Hellenistic form, at the twilight of the period. The attic had relief panels, depicting men between pilasters; these are likely to have been ancestors of the eponymous, a grandson of Sulla. The figures carved on the supports on three sides of the building wore long dresses with Archaising omega folds. Their damaged torsos make it difficult to reconstruct the figures, but they appear to have had one or both arms raised, with the palm of the hand facing up, but these palms may not have reached to the lintel so as to give the illusion of supporting it. Even if one can argue that the figures did not meet all the criteria for being described as Caryatids, one can see their source of influence, particularly in the use of such figures on a sepulchral monument. Caryatids were used by the Romans, almost exclusively in a funerary context, and they decorated a number of Roman Sarcophagi. Caryatids were carved on the corners of a rock-cut sarcophagus at Cyrene which, though sometimes dated to the Hellenistic period, is more likely to be Roman.
Thus one can see that there are quite a number of sculptural examples of the Naples pose in the Hellenistic period, and conclude that this is the type that seems the most likely for Vitruvius’ Caryatids. A second text, Athenaios 241d quoting Lynceus of Samos, a contemporary of Menander, reinforces the theory. In it Eucrates speaks of dining in a dilapidated house of which he says:
when one dines here, one has to use one’s left hand,
as Caryatids do, to hold up the roof.
Although the pose does not preclude dancers, the idea of figures supporting a ceiling would be more in keeping with those described by Vitruvius.
The pose, of a female figure supporting with her head and one or both hands, was not new in Greek art, but its use in large-scale sculpture is only documented from the beginning of the Hellenistic period. My argument is that the designers of the Caryatid Monument adapted the figures to their own means, and provided the catalyst for their being copied in subsequent architectural sculpture for the decoration, at least initially, of tombs.
The question of where and by which Greeks the Caryatid Monument was erected arises. The Spartans would seem the most obvious answer, having erected structures with figured supports previously, the Amyklaeon and the Persian Stoa being the most famous examples. Plommer suggests that it may have been a table-like structure built inside the Persian Stoa, or possibly the alteration of the building to which Pausanias alludes.
An alternative is that the Monument was in a Panhellenic sanctuary. If this were the case it is strange that Pausanias nowhere mentions it. Of course it could have been destroyed by the second century AD, or he may simply have failed to include it in his description of the Persian Stoa, given his lack of interest in architecture, and particularly stoas. The fact that Vitruvius confuses the Caryatids with the Persian War, and that Artemisia of Caria who was defeated in it was amongst the figures represented on the Persian Stoa would seem to reinforce the link between the two structures. And after all, as Pausanias said in 3, 11. 1, there were many wonderful sights in the city of Sparta, as there were at Athens, and that he could not list them all.
5.2 Atlantes and Telamones
Atlantes or Telamones provide support on the nape of their necks and on the forearms of their raised and bent back arms. Their elbows were thus thrust forwards from and slightly above their heads; their necks, generally reinforced by the representation of a beard, were also extended in an attempt to create a vertical surface, as well as the illusion of burdened figures. The earliest extant examples of these are from the early Classical Olympieion at Agrigento; the structure is known to have been restored in the early Hellenistic period and may have acted as a model for subsequent imitators. Figures in the Atlantes pose, named after the mythical figure who held up the universe, could be male or female in the Hellenistic period. They were popular for the decoration of theatres, where their iconography was adapted to suit the context, and their attributes became those of the retinue of Dionysos. Bearded males that were not followers of the god appear to have been Vitruvius’ Atlantes rather than figures in the Atlantes pose. Since Vitruvius does not fully differentiate the term Telamon, I use them interchangeably. They were carved in high relief, and tended to incorporate a Doric frieze, which they supported. Kneeling figures acting as supports are also known, but few date from before the Roman Imperial period; those from the parodos walls of the theatrum tectum at Pompeii, ca. 80-75 BC, and from the theatre at Pietrabondante, late Hellenistic, were variants of the Atlantes pose, but most were not.
The most popular type of figured supports in the fourth century and Hellenistic period are not Caryatids, but Maenads and Satyrs in the Atlantes pose to be found in high-relief supporting stage fronts with their forearms. The figures involve a thematic change to fit the function of the building; they become, for example, part of the Dionysiac retinue at theatres. These supports were common in Magna Graecia and Sicily, with examples in the East at Delos and Athens, though the dating of the last is disputed and may be Roman. There is also residual evidence for them in painting and on small-scale objects from Etruscan culture, all funerary. The pose of the figures seems to develop from those decorating the exterior of the temple of Zeus at Agrigento. The examples whose order is known are all placed in conjunction with the Doric order. Such figures are described in De Architectura, where they are identified as Atlantes or Telamones.
Again, if statues of the male figure support brackets or cornices, we call them telamones, nor do we find in any treatises what they are and why they are so called. But the Greeks call them Atlantes. For, in history, Atlas is represented as sustaining the universe, because he was the first by his powerful intellect and skill to set forth to mankind the sun’s course and the revolutions of the moon and all the stars. And therefore because of this service he is represented by painters and statuaries as sustaining the world.
[Vitruvius, De Architectura, 6. 7. 6: Loeb trans. F. Granger]
Vitruvius, recording their use before the writing of his treatise, does not place the figures within an order, although the archaeological evidence suggests a Doric context, but rather includes them in his section on housing, after a digression on the Greek Xystos. An example exists of male figured supports used in such a context: the so-called Portico of the Hermeses, of the Xystos of Cyrene Gymnasium, for which Stucchi proposed a mid second century BC date, although others prefer the time of Commodus. The portico ran along the west of building, and was a wall articulated by alternating high-relief figures of Hermes and Hercules carved on to piers, with a Doric entablature linking them above. Between the pillars there are window-like openings. As recommended by Vitruvius, the Xystos is smaller in size than the stadium.
The earliest securely dated figured supports come from Monte Iato theatre. Monte Iato was a highly Hellenised city, inland in Western Sicily, given the status of a polis in the middle of the fourth century. This led to much construction, including the theatre for which one can note three main phases of work: during the last years of the fourth century, ca. 200 BC, and later rebuilding. The sculpted supports, two Maenads and two Silenoi, which are now in the Museo Civico Ietino, San Cipirello, are attributed to the stage front or parodoi of the first phase. That they belong to the theatre is attested by their find spots within it. Schmidt feels that their details indicate a position above eye-level. Two large contemporary lions, lying down on bases along the line of the parodoi, facing outwards, were located in the front rows of the auditorium at the level of the thrones. This unusual place for sculpture, and the high overall quality of the detail, shows how decorated the theatre must have been.
Here the use of supporting figures is largely illusionistic, for though they seem to have an active role in carrying the weight of the building, and one can note the strain depicted within the musculature of the Silenoi’s chests, the burden is on the uncarved rear of the blocks. Each of the high relief figures is sculpted on to the front of three super-imposed blocks of local stone and was covered in a light layer of stucco. They can be seen as figured ‘pilasters’, with broader rears applied to or set within the walls. The poses of both sets of over-life size figures are similar, ‘supporting’ with their forearms and napes, legs straight. There are slight differences between the ‘pairs’; some would see this as indicative of different dates, though here it is probably just an attempt at characterisation. Such a Dionysiac retinue was appropriate for a theatre.
The Maenads are particularly archaising, though less so than many later examples (fig. 63). One can see this expressed in their hair, and in their high-girdled symmetrical peploi, with ‘omega’ folds at the overhang and hem. Long hair falls over the front of their shoulders, and on their heads are ivy and berry wreaths around poloi; one polos has an edge preserved that may be architectural. The Maenads are especially well preserved. The Silenoi are whole, but less well preserved (fig. 64). They have long beards, and wear fleece loin-cloths that create a less symmetrical appearance than the Maenads’ drapery. Traces of one leaf garland can be seen by their bestial ears, and of another across their muscular chests. One can note a similarity between these Silenoi and the one from Athens that is sometimes assigned to the Lycourgan building.
To the patronage of Hieron II of Syracuse (269-215 BC) a number of figured supports can be ascribed. A Maenad and a Satyr, both of local stone and now in Syracuse Museum have been attributed to the architectural decoration of Syracuse Theatre, dating to the period of his reign. Both figures are like the Akragan Atlantes in pose. The upper part of a Maenad, carved in-the-round, was found in the orchestra, so is secured as part of the decoration. She wears an ivy wreath, and has more Classicising drapery than the Iato Maenads, but still with Archaic motifs. The rear, which was not intended to have been visible, is roughly worked, and the continuation of some details suggests that it was against a wall, but perhaps not attached to one. The figure is quite bulky, and may well have served as a support. An oddity is the extension of the back of the neck, which may have joined it to the wall behind.
The Satyr, of which only a damaged upper torso with a head and part of a fleece-covered pelvis survive, is without provenance, but its structure is quite different, resembling more the carved front of a high-relief pilaster. Though similar in general outline and made of the same kind of stone covered in stucco as the Maenad, the two pieces are diametrically opposed in mood. The difference in the treatment of the back, along with the popularity of male counterparts to Maenads in theatres, leads Schmidt to assign him to a different part of the theatre from the female support, but still within it. Yet a pair of feet, from a Satyr or Silenos, and attached in a similar way, can still be seen at the near-by Altar of Hieron II, and I propose to assign it there instead of the theatre. The altar, measuring 194.5 x 20.85 m with an original height of ca. 10 m, is the biggest known. There were staircases at each end, and fragments of two different sizes of Doric entablatures have been found, one of which crowned the figures. Jannot sees the Altar as deliberately copying the earlier temple at Agrigento, to create links with the past.
What is particularly interesting about Hieron II is that Athenaios 208b, quoting Moschion, describes a ship the tyrant gave to a Ptolemy, the Syracusia, later renamed the Alexandris. The passage is better known for its description of mosaics, but also mentions the exterior of the ship. At the top of this there was a row of Atlantes, either in relief or possibly painted, six cubits high, or ca. 2.70 m, with a triglyph frieze above them, thus indicating once again the association of these figures with the Doric order.
The upper bodies of two terracotta Silenoi, one with a garland on his head, the other without, were found in Syracuse. They are close enough to the Satyr linked to the Altar of Hieron II, or to the theatre, to be believed to have been copying it; their scale suggests that they were part of the decorative scheme of a Late Republican house. A similar phenomenon can be seen at Agrigento. A Hellenistic terracotta vessel from here had small relief Atlantes figures that acted as handles. A Telamon is also depicted on a red figure vase from the town. The figures depicted on both of these are close enough to the monumental Atlantes of the Olympieion for one to argue that they were imitators; if one wished to speculate further one could view them as souvenirs produced for visitors.
Delos theatre, from the late third century, is the only theatre outside the West in which it is certain that the supporting figures are Hellenistic originals; these are two Silenoi, life-size, Delos A 4175a-b. They backed on to a square pillar with sides of 0.20-0.22, which is approximately the size of the pillars of the proscenion, though these were constructed using a different method. It is likely that the figures supported the ends of the stage. These bearded figures are characterised by rolls of flab on the stomach, and bald heads crowned with ivy wreaths; a cloth was wrapped around their waist and upper thighs. They are quite damaged; one is preserved from the head to the bottom of his thighs, ht. 0.95, the other as a torso, ht. 0.62. They ‘supported’ with their head and shoulders. The two figures would have been symmetrical, with one Silenos having his weight on the right leg, his head turned towards that side, and the other figure in the opposite pose. Another supporting figure, Delos A 4177, may also have been part of the decoration of the theatre; it is on the same scale and technically similar to the other figures, although its features are those of a barbarian rather than a Silenos.
Attempts have also been made to redate a Satyr from the Theatre of Dionysos in Athens, inv. 2302, to the ca. 330 Lycourgan restoration, on the grounds that it differs greatly from the other figures, which are generally dated as Hadrianic. This has been controversial, and Schmidt dismisses the idea, although she admits that there are no concrete grounds why the piece should not date to the early Hellenistic period. Although poorly preserved, the figure is different in style and execution from the Roman examples. Neither theory has found extensive support, but should not be dismissed. Interestingly the figure is close to those from Monte Iato, which may have derived from it; this figure type was much copied in the Roman period, suggesting that the originals were in a rather more illustrious location than an obscure theatre in the mountains of Sicily.
The link between Atlantes and the decoration of theatres is so strong that the publishers of a tufa Telamon figure found built into a wall of the late Mediaeval Palazzo Dardes in Venosa saw it as evidence for a Roman theatre in the town. Although the pose is that of a standard Atlas figure, supporting with forearms and the back of its neck, the rendition of the corpulent naked body is unusual. The figure probably dates from the Late Republican or early Imperial period. Atlantes continued to be used for the decoration of theatres in the Roman period, for example the Telamones from Nîmes and from Falerio, but to a lesser extent. At some theatres the figures may have replaced or supplemented Hellenistic figures, as appears to be the case with the Athens figures, and the Caryatids at Miletos, whilst at others they were new creations that followed the Hellenistic manner of decorating theatres.
Segesta theatre, ca. 200-175 BC, a classic example of a Hellenistic theatre converted into a Roman one in the first century BC, had high-relief figured supports of Pan decorating the ends of the parascenia. The Hellenistic figures, ht. ca. 2.80, with their heads preserved, are related to figured columns: they supported a basket on their heads with one raised arm, which in turn supported the entablature. The Pans, though not in the Atlas pose, continue the theme of followers of Dionysos as decoration in theatres, and the use of supporting figures within these. A number of copies of the figures were made during the Roman period, all to the same size. Two figures in the Cortile of the Palazzo dei Conservatori, nos. 5 & 23, come from the Theatre of Pompey in Rome. The figures are heavily restored, and it is uncertain from which phase of the building they originate; Schmidt dates them to the second century AD, though they need not be so late. A single figure is now in a private collection on Capri. Another pair of figures comes from the Canopus of Hadrian’s Villa at Tivoli; there they were arranged in a setting with copies of the Erechtheion Kore type, suggesting knowledge of a thematic link between the two types, both originating in architectural sculpture where they were used as figured supports. The Velletri sarcophagus also makes such a link, with depictions of buildings with different types of figured supports.
Jannot states that there were no funerary examples of Telamones known in the west Greek world, in contrast to the larger role they play in Etruscan funerary iconography. Some examples from Southern Italy of female figures in the Atlantes pose do however exist: one pair is definitely from a tomb, while the other is likely to be from a funerary naiskos. In contrast the majority of Hellenistic Caryatids come from funerary structures in Greece and the east Mediterranean; this may be either because the Caryatid Monument was originally a cenotaph, or because the figures themselves became symbols of mourning. A late fourth century Messapian tomb at Vaste di Puglia, near Lecce, has yielded four archaising female figures in the Atlantes pose; these did not serve an architectural function, but rather lined one wall, and supported a continuous frieze. None of the pieces are now whole, but between them one can reconstruct the figures entirely, and note that rather than being in mirror-image pairs, which was the norm, they were almost identical. Interestingly, as previously noted, the small Amazonomachy frieze at one end has a depiction of a Vitruvian Caryatid. Two Tarentine figures, now in Geneva, are likely to have decorated a naiskos of the late fourth or early third century BC (fig. 65). The figures back on to pillars. The drapery has nice swallow-tail folds, but overall they are not overtly archaising. Differences within the two sculptures’ details were probably due to the carver rather than to any attempt at differentiation; again neither figure is whole.
A Telamon previously linked to a theatre has now been reassigned to the so-called temple beneath S. Leucio, Canosa; the structure is highly unusual, and may in fact have been a tomb or monument. The figure, a young satyr clad in a fur loincloth, was made up of numerous blocks of stone, of which three survive, and was coated in stucco, traces of which remain; the colossal figure was originally ca. five metres high, and presumably one of a pair. The structure was of mixed order, but the figures supported a figured Doric frieze; its date is uncertain, but is it believed to be from the middle Hellenistic period.
Although no monumental sculptures of Etruscan Atlantes used as architectural decoration survive, there is enough secondary evidence to show their popularity in funerary contexts. Four Etruscan cinerary urns decorated with Telamones flanking the scenes depicted are extant, all believed to have originated in Volterra. No. 82 in the Museo Guarnacci, Volterra, with a firm provenance from Volterra, has nude, bearded and heavily muscled men on the ends of the urn, flanking a scene depicting the departure to the underworld. It dates from the first half of the third century BC. Museo Guarnacci no. 315, dated ca. 150-130 BC, has similar figures around a scene of a husband bidding farewell to his wife. It was made by L’Atelier des rosettes et des palmettes, a group whose eponymous friezes are a derivation of the small-scale Doric friezes that were popularly depicted on Etruscan cinerary urns. Above it is a row of dentils. On Museo Guarnacci no. 273, ca. 150-130 BC, the Telamon figures, the right hand one of which is missing, flank a Centauromachy. Along the base of the alabaster urn runs a Doric frieze. Florence MA inv. 7784495(?) is a very poorly preserved urn with similar figures again positioned around a Centauromachy, and dated to ca. 150-130. A Doric frieze runs along the base. It would appear to be from the same workshop and a replica of the previous piece. On a fifth urn, Museo Guarnacci no. 186, depicting the Seven Against Thebes, the Telamones are replaced with kneeling figures on brackets within the entablature, whose upper bodies are in the Atlas pose and would appear to be captives. The alabaster urn is very late Etruscan, dating to the beginning of the first century BC.
Related paintings are known from two Tarquinian tombs. In the Tomba del Trifone, Tarquinia, late second century, two of the pillars are decorated with the eponymous figures, who have the upper bodies of Atlantes but are anguiped. There is also a female figure, but with wings. The paintings from the Tomba Tartaglia at Corneto, believed to date to the second half of the third century BC, are no longer extant; they were fortunately preserved in an eighteenth century engraving. The tomb was found in 1699, and had paintings that ran along the upper halves of the walls; the decoration of two of the walls was preserved at that time, with ten figures divided into scenes by three Telamones. The Telamones are believed to have stood on ‘ground lines’, possibly bases, not shown in the engraving. They supported rounded forms below the ceiling, which can be interpreted as the universe. Jannot restores a fourth figure, not depicted by Dempster, for the sake of symmetry.
Bone and ivory boxes were favoured by the Etruscans, and a number of these have been found in burials. One drawn by Durm in Florence, Museo Archeologico, but unfortunately now missing, follows the format of the cinerary urns. All that remained of the bone box was one schematic Atlas figure, whose form was nearly rectangular, and the small Doric frieze that he and his pair would have supported, decorated with alternating rosettes and boucrania in the metopes. The frieze shows continuity of the idea that Atlantes were part of the decoration of the Doric order. The work is relatively crude, but details such as mutules beneath the triglyphs are represented. The bearded figure was rendered in simple outline, standing on a low plinth. Plaques from a box with similar figures were found in Tomb no. 20 of the Osteria Necropolis, Vulci. The four figures originally decorated one or two boxes, but it is difficult to pair any two as they vary considerably. They are dated by their archaeological context to the late third century BC. Another box, now in Naples, had ivory plaques depicting Caryatid figures.
A young Telamon-type figure, possibly a satyr, was found in a room of a Late Republican villa at Gozzo, Malta. Preserved from the waist up, the surface is badly worn, but the expression on the face is conserved and appears to be one of pain. The figure acted as an architectural support, as attested by the pose, the flattened top of the skull and the fact that it was attached to a pillar at the back. From a private House at Centuripe, ca. 100-75, come four terracotta figures, two Satyrs and a pair of female figures, probably Maenads; three-quarter life-size, they appear to have decorated two doorways rather than supported a cornice. Their arms are variations on the Atlantes pose, folded back to a lesser degree. Due to the nature of the material, the figures are damaged and not fully preserved. Terracotta Atlas figures from a second century BC private Villa at Fregellae were recently discovered during excavations. An early Imperial female terracotta figure in the Atlantes pose was part of the decoration of the House of Josef II at Pompeii, and shows the continuity of this form of decoration in private architecture; she wears a peplos, and is slightly archaising.
Terracotta was used for a number of small-scale figures in this pose, but due to the nature of the material they were decorative trompe l’œuil figures rather than actual structural supports. Many of the sculptures lack a context; all the examples come from Italy. The torso of a small terracotta Atlas was found in Bolsena. The back is smooth and unworked, showing that it was placed against a vertical surface. Remains show that though the arms were folded back, the head could not have served as a support. Nude, the figure is heavily muscled and close to those from the Baths at Pompeii; Jannot restores the figure as from a similar frieze, but ca. 200 BC so earlier in date.
Perhaps the most famous terracotta examples are the row of figures arranged as a frieze around the walls of the Tepidarium of the Forum Baths, Pompeii, dated by inscription to 80-75 BC. The figures support with their forearms and the pseudo-poloi they wear on their heads. There is variety within the figures, with some nude and others wearing different types of loincloths; the heights of their poloi also differ. This belies the suggestion that they were mass-produced mould-made figures. Four monumental terracotta Atlantes dated to the first century BC were found in Aquileia. They are large and flat, and would have decorated the exterior of a building, though not necessarily a theatre. Most terracotta Atlantes would have served as decoration in private architecture.
Two male Telamones come from a Hellenistic structure in Solunto; the figures are carved in relief on a slab of stone, and probably flanked a doorway or were placed between columns, rather than replacing them and acting as architectural supports. A similar male figure comes from Monte Scaglioso, ca. 300 BC; the figure is larger, better preserved, and in slightly higher relief, with more defined musculature. It was built into a wall, and could have been used as a support.
A poros Atlas figure was found beneath the northern part of the Roman agora of Thessaloniki, and is believed to come from a Hellenistic building of ca. 150-50 BC. The sculpture is preserved from the head to the upper thighs, including the shoulders, but its arms are missing. It is battered but otherwise quite well preserved, and reconstructed of three fragments: the torso, lower hips and thighs, and the buttocks. The head, which had a taenia, is set forwards with the beard strengthening the neck and relieving some of the pressure on it. The naked figure is a Telamon or Atlas figure, not a Silenos, with exaggerated musculature that was clearly not modelled from life and so fails to create an illusion of the great burden he was meant to be supporting. The back of the figure was notably less worked than the front, suggesting that it stood against a wall rather than being attached to one, and it shows no obvious points of attachment other than at the crown and forearms. The clean break below the shoulders and the presence of dowel holes indicate that the arms were pieced. The large figure, originally ca. three metres tall, was probably set against rather than into a wall.
The form originally used to represent the encumbered figure of Atlas, who laboured under the burden of supporting the universe, developed iconographically to fit the context. In tombs mourning women were depicted, whilst in theatres they became Satyrs, Silenoi and Maenads, as part of the Dionysiac retinue. Atlantes were particularly popular for the decoration of the theatres in Italy and Sicily, although they were also to be found in the East Mediterranean. There are no examples known for sure to have decorated temples: arguably Canosa was one, but otherwise there are no certain examples between Agrigento and the Hadrianic period. The figures, like Caryatids in the East, were part of Etruscan funerary iconography, and possibly to a lesser extent in southern Italy. Atlantes were later incorporated into the repertoire used for the decoration of private houses in Italy; there are a number of examples of these in terracotta, and the stone figure from Malta.
Vitruvius, writing in the last years of the Republic and the beginning of Augustus’ power, mentioned two types of figures used within the sculptural decoration of buildings. Although he did not describe them at length, it is clear that their contexts and the structures they decorated were Doric. The form of Atlantes is clear, but that of Caryatids is more controversial. Vitruvius’ text however clearly excludes the famous carved female figures who replaced columns in the Erechtheion’s south porch from being described, as they generally are, as Caryatids; the framework within which the Athenian maidens and their like were placed is Ionic, they were not represented as demeaned, and they are both chronologically and geographically inconsistent with the account. The numerous figures from the Hellenistic period, and firmly placed within the repertoire of the Doric order, are those that supported a Doric entablature with their poloi and the palms of their raised hands; these are clearly the women Vitruvius refers to as Caryatids. The pose itself was not new, being known from perirrhanteria, mirror handles, and many Achaemenid examples, but its use in monumental Greek sculpture presumably was.