SOURCE: "Pausanias and the Temple of Hera at Olympia," The Annual of the British School at Athens, No. 90, 1995, pp. 461-73.

[In the following essay, Arafat utilizes Pausanias's work in analyzing the Heraion's contents and purpose.]

In the temple of Hera there is an image of Zeus. The image of Hera is seated on a throne, and he is standing beside her wearing a beard and with a helmet on his head. The workmanship of these images is rude. Next to them are the Seasons seated on thrones, a work of Smilis of Aigina. Beside them stands an image of Themis, as mother of the Seasons: it is a work of Dorykleidas, a Lakedaimonian by birth, but a pupil of Dipoinos and Skyllis. The Hesperides, five in number, are by Theokles, also a Lakedaimonian, son of Hegylos; he, too, is said to have studied under Skyllis and Dipoinos. The image of Athena, with a helmet on her head, and carrying a spear and shield, is said to be a work of Medon, a Lakedaimonian: they say that Medon was a brother of Dorykleidas, and was taught by the same master. There are also images of the Maid and Demeter and Apollo and Artemis; the two former are seated opposite each other, and the two latter are standing opposite each other. Here, too, are Latona and Fortune and Dionysos and a winged Victory: I cannot tell who made these images, but they seem to me to be also extremely ancient. The images I have enumerated are of ivory and gold. But afterwards they dedicated other images in the Heraion: Hermes bearing the babe Dionysos, a work of Praxiteles in stone; and a bronze Aphrodite by Kleon, a Sikyonian. Kleon's master, Antiphanes by name, was of the school of Periklytos, and Periklytos was a pupil of Polykleitos the Argive. A gilded child, naked, is seated before the image of Aphrodite: the artist who fashioned it was Boethos of Chalcedon. Hither were brought from the so-called Philippeion other statues of gold and ivory: Eurydice, Philip's [. …] There is a chest made of cedar-wood, and on it are wrought figures, some of ivory, some of gold, and some of the cedarwood itself. In this chest Kypselos, who became tyrant of Corinth, was hidden by his mother when at his birth the Bacchiads made diligent search for him. As a thank-offering for his escape his descendants, the Kypselids, dedicated the chest in Olympia … There are other offerings here also: a small couch mostly adorned with ivory; the quoit of Iphitos; and the table on which the victors' crowns are displayed … And there are Hera and Zeus, and the Mother of the Gods, and Hermes, and Apollo with Artemis. Behind these is represented the celebration of the games. On the one side there are Asklepios and Health, one of his daughters, also Ares, and beside him, Contest; and on the other side there are Pluto and Dionysos, Persephone and nymphs.
(Pausanias, v. 17. 1-20. 3)

The Hesperides were removed [from the Epidamnian treasury] by the Eleans, but were still to be seen in my time in the Heraion.
(vi. 19. 8)

The people of Megara, near Attica, built a treasury, and dedicated offerings in it, consisting of small cedar-wood figures inlaid with gold and representing Herakles' fight with Acheloos. Here are represented, Zeus, Deianeira, Acheloos, and Herakles, and Ares who is helping Acheloos. Also there was formerly an image of Athena, because she was an ally of Herakles; but this image now stands beside the Hesperides in the Heraion.
(vi. 19. 12)

Some fifty years ago,¹ Dorothy Kent Hill used the evidence of Pausanias (v. 17-20. 3) to consider some of the contents of the temple of Hera at Olympia and to characterize it as a 'sort of storehouse' and a 'museum'.² In this she had been anticipated by K. Wernicke, who saw the Heraion as a 'Kunstmuseum', and believed that it was created as a compliment to Nero on the occasion of his visit to Olympia.³ Here I take their arguments a stage further, looking at the entire contents of the Heraion as described by Pausanias and as revealed by excavation, its status and function in Pausanias' day, and the implications for the earlier history of the cults of Hera and Zeus at Olympia, and the use of temples in Roman Greece.

Although the silt that covered Olympia to a depth of around 6 m has proved invaluable in preserving an exceptional quantity of art and architecture, we are still indebted to Pausanias for many insights into the embellishment of the sanctuary. The extent of modern archaeological commentaries on Pausanias at Olympia is itself testament to the accuracy of his description, on which he himself remarked (v. 25. 1, speaking of the statues of Zeus within the Altis).⁴ He is valuable not least for giving an idea of what is irretrievably lost to us—not just Pheidias' Zeus, but the ash altar of Zeus, the seventy-odd lesser altars, the more than two hundred athlete statues he notes, and so on. Discussions of Olympia must, therefore, be written at least in conjunction with Pausanias.

The contents of the Heraion

By the time Pausanias visited Olympia around AD 173,⁵ it was by any standards a crowded sanctuary. According to his account, the Altis included two major temples, that of Zeus (c.472-457) and the early sixth-century temple named by him as being dedicated to Hera, almost the last major building he describes. Pausanias occasionally tells us that he will only record a selection of what he sees—thus, at Delphi he excluded statues of athletes or musicians (x. 9. 1-2). His phraseology as he enters the Heraion implies that this is one such occasion.… He implies that what he is about to describe is not all of what he in fact saw within the temple. And indeed, more has been found in the Heraion than he mentions: for example, fragments of Ionic capitals from the Leonidaion, an early Roman female statue, and bases of three other early Roman statues.⁶

With this literally unknown quantity in mind, Pausanias' description of the contents of the Heraion may be examined in detail. He begins with a standing Zeus and seated Hera, noting that 'the workmanship of these images is rude …, suggesting an early date, since Pausanias reasonably sees technical skill as a chronological indicator.⁷ Next, the Seasons seated, by Smilis;⁸ although the number is not specified, three or four may be presumed.⁹ Themis follows, by Dorykleidas, whose approximate date is given by his being a pupil of Dipoinos and Skyllis, themselves pupils or sons of Daidalos (ii. 15. 1; dated c.580 by Pliny, NH xxxvi. 9).¹⁰ Contemporary statues of the Hesperides made by Theokles and taken, as Pausanias tells us later (vi. 19. 8), from the Epidamnian treasury,¹¹ add five to the number. A contemporary Athena by Medon is probably the one mentioned subsequently (v. 19. 12) as being from the Megarian treasury of c.520 (below). The catalogue continues with Kore and Demeter, Apollo and Artemis, Leto, Tyche, Dionysos, and a winged Nike.

This list comprises some twenty-one statues (assuming four Seasons).… [Pausanias] says that they are all chryselephantine,¹³ the use of which for statues was necessarily less common than that of other materials, conditioned by the expense of both the gold and the ivory. Twenty-one statues in the technique in one temple seems, therefore, an unlikely indulgence if they were originally made for the temple. The size of these statues is not specified: clearly, they are not as large as the most famous chryselephantine work at Olympia, Pheidias' colossal Zeus, but, for an order of magnitude, we may perhaps think of the surviving chryselephantine fragments from Delphi of about two-thirds lifesize.¹⁴

The concentration of statues is striking; so too their antiquity and material. The latter, as Hill observed in 1944,¹⁵ indicates that the well-known limestone female head found at Olympia cannot be from a cult statue of Hera for this temple; such a conclusion seems to me (as it did to Hill) inescapable from a reading of Pausanias.¹⁶ Hill re-identified it as that of a sphinx, an interpretation recently vigorously supported, although by no means universally accepted.¹⁷ Doubts have also been raised about the stratigraphical context in which it was found, which may have made it inaccessible to Pausanias.¹⁸ While its date is compatible, and its findspot (west of the Heraion, by the palaistra) inconclusive, the evidence of Pausanias, on whose account the original association was based, indicates without ambiguity that the material is different.

With the removal of the association of this head, and the acknowledgement that the cella is crowded with a large number of statues to a miscellany of gods, heroes and other figures, the necessity to think of the building as still functioning as a temple to Hera is diminished. Indeed, it raises suspicions as to whether it was functioning as a temple at all. Exactly what it might have been at which period is an issue to which I will return.

The temple was still more crowded: Pausanias next notes images dedicated 'afterwards', by implication post-archaic offerings. These comprise the Hermes and Dionysos by Praxiteles, and a bronze Aphrodite by Kleon of Sikyon, whose artistic lineage is traced to Polykleitos two generations back, placing him around the early fourth century.¹⁹ The gilded child (?Eros) associated with Aphrodite is by Boethos, and its association with Aphrodite suggests contemporaneity.

We have, then, a further three statues to add to the previous twenty-one. Of these, the Hermes and Dionysos has survived, found where Pausanias specifies. It is now generally agreed that it is not a fourth-century original.²⁰ However, this need not invalidate Pausanias' statement that it was by Praxiteles: it is possible that he, or we, or both, would think instinctively of the wrong Praxiteles, a possibility made more realistic by the practice of naming grandchildren after grandfathers, opening the possibility that the work was by a descendant of the Praxiteles, of the same name.²¹ It is very likely that if Pausanias were told that the statue was by Praxiteles he would, as we do, think first of the Praxiteles; as Martin Robertson has put it, 'it was surely to [the Praxiteles] that Pausanias thought he was ascribing the statue'.²² At the least, the fact that the statue's authenticity still divides scholars suggests that we have only a limited right to pass judgement on Pausanias in this instance.

Although there is a small textual problem,²³ it is sure that Pausanias next described in the Heraion two chryselephantine statues from the Philippeion, the circular building which he says was built by Philip II, and which may have been completed by his son Alexander after Philip's death in 336.²⁴ These are of female members of the Macedonian royal family, dating from some five hundred years before Pausanias' own time, now removed to the Heraion from their original setting, and of no cult significance whatsoever to the temple in which they were seen. Their provable association with a specific building marks them out from the other objects in the Heraion noted thus far.

There follows the chest in which the Corinthian tyrant Kypselos was supposedly hidden as an infant by his mother in order to protect him from the Bacchiads, and for which Pausanias is our only detailed source. Dio Chrysostom, the only other author to mention it (Or. xi. 45), places it in the opisthodomos of the Heraion; Pausanias is not specific, but he has already passed from the opisthodomos (where he notes a wooden column) to the cella, and it seems unlikely that he would return to the opisthodomos in the middle of listing the objects in the temple since he would, on that basis, then return to the cella. The implication follows that Pausanias saw the chest, like the other objects, in the cella. Whether the chest had been moved in the century or less since Dio visited, or whether Dio misread his notes, cannot now be ascertained.

The chest is probably of the early sixth century, contemporary with the Heraion itself, which would disqualify it from being the 'real' chest of Kypselos, tyrant c.657-627, born therefore in the early seventh century. While we may infer the date of manufacture, Pausanias tells us that it was dedicated by the Kypselids themselves, presumably therefore during their period of power, which continued to 582.²⁵ In any case, the exactitude of the date is irrelevant in the matter of mythmaking. The chest was cedarwood and adorned with figures in wood and gold and ivory; Pausanias details thirty myths on the chest, over one hundred figures, more than twenty horses, enough armies to fill one side with infantry and cavalry, and unnumbered figures such as spectators at the funeral games of Pelias, Muses, Nereids, Harpies and centaurs, and tripods, apple trees, pomegranates, a vine, a house, and so on.²⁶ The size of the chest is not specified, but it is unlikely to have been small with such extensive and elaborate decoration (despite the miniaturist tendency of archaic Corinthian art). It would have to be at least convincingly baby-sized, even if it never in fact contained a baby; as with the irreconcilable date, that is an adjustment necessary to create an effective icon.

And there is still more in the Heraion: Hippodameia's toy couch, a discus bearing the Olympic truce and said to have belonged to Iphitos, and the chryselephantine table on which the wreaths for Olympic winners were laid, made by Kolotes.²⁷ Six sculpted figures follow: Zeus, Hera, Apollo, Artemis, Hermes (all for the second time in the temple), and the Mother of the gods. A lacuna before this list robs us of certainty as to the nature of these representations: they may be decorative figures on the table of Kolotes, presumably therefore in relief; but there is nothing to prove that they need be part of the table which, being chryselephantine, was hardly a negligible object in itself, even without further decoration. Given the staccato way in which Pausanias is running through the objects in the Heraion (with the exception of the long description of the chest of Kypselos), an abrupt switch in focus from the table to unconnected statues would be quite compatible with his narrative technique.²⁸ If these are free-standing statues, we know nothing of their size or material; but this would raise the possibility that the statue of the Mother of the Gods is the cult statue originally in the Metroon which was by Pausanias' day occupied by Roman statues (v. 20. 9).²⁹ K. Hitzl³⁰ suggests that an earthquake soon after the mid-first century BC may have destroyed the cult statue from the Metroön, hence Pausanias did not mention it there, and S. Stone³¹ that Nero removed it in AD 68. However, neither theory, although credible, has a basis in the sources, and the possibility that this statue, like so much else, had been transferred by Pausanias' time to the Heraion remains worthy of serious consideration.

Next, Pausanias mentions what is presumably a relief sculpture with figures on both sides; he places this opisthe the six figures which may be from the table of Kolotes, translated by Frazer and Jones (Loeb ed.) as 'behind', but by Levi as 'on the back', with the implication that these are further relief figures on the table, an idea strengthened by the following phrase kata de heteran pleuran. Unfortunately, neither interpretation is provable, nor is any link with the table. While this interpretation is of some interest, it does not affect the number of statues, since they are in either case relief figures rather than free-standing. The latter reliefs show the diathesis of the games, translated by Frazer as 'the celebration', by Jones as 'the disposition', and by Levi as 'the management', perhaps implying the rules of the games; Pausanias details a number of personifications appropriate to the games.

Pausanias continues with a wooden Athena taken, like the Hesperides, from a treasury, in this case the Megarian of c.520 (vi. 19. 12). This is, as noted above (p. 463), in all probability the 'image of Athena' mentioned at v. 17. 2, next to the Hesperides in the cella of the Heraion.

In summary, Pausanias' description of the contents of the Heraion reveals a substantial number of statues: at least 25, and as many as 31 (allowing for uncertainty over the number of Seasons, whether Athena is mentioned twice, and whether the six figures mentioned after Kolotes' table are reliefs decorating it, or free-standing), and half a dozen other objects.

The Function of the Heraion

Whatever the exact number of objects in the cella of the Heraion (and nothing in Pausanias' account suggests that any of them is elsewhere than in the cella), it was exceptionally crowded. Indeed, it is important to stress that in his entire work Pausanias describes no other temple as being anywhere near as crowded, even if one uses the minimum number calculated above. Added to these was the statue of Poppaea Sabina dating, like the three Roman bases also found, from the second half of the first century AD, but the only one found in the cella, as the bases were found in the pronaos.

Such crowding reduces the likelihood of any one statue standing out as the prime one, the cult statue that every temple would have boasted. That there were originally cult statues of Zeus and Hera has been inferred from the wide base found at the west end of the cella, as well as from Pausanias' account; and the further inference has been drawn that the temple was originally dedicated to Zeus and Hera together, becoming a temple of Hera alone with the building of the temple of Zeus c.472-457. The idea of the Heraion as originally a joint temple of Zeus and Hera is long-established: Frazer says 'formerly it would seem to have been a joint temple of Hera and Zeus, if we may judge from the fact that the image of Zeus stood beside that of Hera in the cella',³² a straightforward assumption that the first two statues Pausanias describes are in fact the presumed, and expected, cult-statues. B. Ashmole and N. Yalouris refer, interestingly and without comment, to 'the so-called Heraion',³³ characterizing it in the period before the building of the Zeus temple as 'probably not a temple of Hera only, but of Zeus with Hera as his consort'. W. B. Dinsmoor is more definite, saying that the Heraion was 'originally a temple of Zeus and Hera together, but relegated, after the completion of the new temple of Zeus … to the worship of Hera alone'.³⁴ He states that in the fifth century, 'Zeus—though not Hera—acquired a more spacious home than the old Heraeum which had originally sheltered them both', and speculates that 'the archaic cult statue of Zeus may have been transferred for the occasion from the Heraeum',³⁵ an idea also propounded by Frazer³⁶ and Papahatzis.³⁷ But is the implication that the cult statue (identified in this case with the Zeus and Hera mentioned by Pausanias) moved back to the Hera temple, presumably on the installation of the Pheidian statue in the Zeus temple? If so, why would the Zeus statue move back into what had become a Heraion? Is it not likelier that, with the building of the Zeus temple, the old temple had simply ceased to be used as a temple? Might it have become popularly known as the Heraion since it was the only other temple beside that of Zeus (excluding the exceptionally small Metroön) and was, appropriately, smaller and less striking?

Dinsmoor does not explain his belief that the temple was originally a joint one; nor, more recently, does Jeffery Hurwit, who refers to 'the so-called Temple of Hera (it was also dedicated to Zeus)'.³⁸ However, the attribution of the archaic temple to both Zeus and Hera may be presumed to be based on a combination of Pausanias' account, the statute base, and a general feeling of inappropriateness that Zeus should be without even a half-share in a temple until the early classical period.

To take each of these objections in turn: Pausanias does not, in my view, oblige us to think in terms of a joint cult statue, since the Zeus and Hera mentioned are two among many; admittedly they are mentioned first in his account, perhaps reflecting a prominent position opposite the entrance, but not necessarily implying priority in function. Secondly, as noted (n. 17 above), Andrew Stewart estimates the dimensions of the base as unsuited to a statue of the size suggested by the extant limestone head which, in any case, we have seen grounds to dissociate. Whether or not one sees Zeus and Hera as cult statues, they are chryselephantine, and no certain inference follows concerning the form or dimensions of any base. Thirdly, I do not feel that the lack of an earlier Zeus temple, if there is indeed such a lack (and Pausanias describes the statue of Zeus in greater detail than that of Hera), is inappropriate, since the cult centre had long been the ash altar, as it continued to be long after the Zeus temple was built, as Pausanias' account of the rites carried out on the altar amply testifies (v. 13. 8-11). The Zeus temple is a showpiece, as the squeezing in a generation later of an ill-fitting cult statue of such conspicuous splendour indicates; had cult been the prime function of the temple, the more ancient the statue, the better it would have been for the purpose. Its cult function is therefore limited. It is true that Pausanias does mention an altar within the Zeus temple (v. 14. 4), but only in passing, and it clearly has less function than the ash altar and the seventy or so altars at which he saw rites being enacted; nonetheless, it is a point of difference from the Heraion, for which Pausanias records no altar. It is the ash altar that makes clear Zeus' supremacy at Olympia, and the temple is, if not incidental, at least secondary (so too at Dodona). If, then, the archaic temple were dedicated to Hera only, no impiety or inappropriateness should be inferred.³⁹ The question remains open.

There are, therefore, questions raised by Pausanias' account, in conjunction with the extant evidence, concerning the original dedication of the archaic temple. Whatever may have been its original status, the sheer number of statues found in the Heraion by the time Pausanias was writing suggests that it was no longer operating in the conventional manner with dedications offered to a single identifiable deity (or pair of deities if it was dedicated to Zeus and Hera together), but that it was, in Hill's words, a 'sort of storehouse' or a 'museum'.⁴⁰ In this it appears to have been exceptional: while comparative material is scarce, it is, as I have already emphasized, uniquely crowded in pausanias' own account. It contained objects from several buildings which were no longer fulfilling their original function. This is true of the Metroon which, as Pausanias notes and as has been confirmed by excavations, contained Roman portraits rather than the cult statue expected;⁴¹ it is clear that the Metroön was 'desanctified'⁴² at some point, and subsequently re-dedicated to the Imperial cult, a practice readily paralleled at other sites.⁴³ It is also true, at least in part, of the Philippeion which, although it still retained some of its original statues, was—significantly for present purposes—denuded of two of the statues which were its raison d'etre. It may also be true of the Epidamnian and Megarian treasuries, from which sculptures had been removed, leaving more still in situ, but suggesting that the function of the treasuries, as storage areas, had been partly transferred, along with the statues, to the Heraion.

Furthermore, the twenty-one chryselephantine statues Pausanias first saw must have been under cover originally (as those from the Philippeion were), since their material made them unsuitable for display in the open air. The evidence therefore suggests that the Heraion had become by Pausanias' day a repository for works from elsewhere, most of them valuable for their antiquity and their material. The notion of a storage area is reinforced by the presence in the temple of the discus of Iphitos and the chryselephantine table on which the wreaths of Olympic winners were displayed. The wreaths are presumably those displayed prior to being awarded (unless they had been awarded, and were subsequently dedicated by the winners), with the further implication that the table would be needed for the games; but this would mean five days use every four years, so that the placing of the table within the Heraion strongly suggests storage between Olympics. Although we do not know whether the discus of Iphitos, on which the Olympic truce was written, was displayed, it too is unlikely to have had any active role in the functioning of the sanctuary for more than five days every four years. However, both the table and, perhaps more so, the discus might well have been objects of interest to visitors all year round, particularly to the kind of cultural tourists Olympia attracted in the Roman Imperial period.⁴⁴ The same might well be said of the relief bearing the diathesis of the games if that is indeed a separate item, and not part of the table (v. 20. 3, p. 466 above).

In addition to the evidence which Pausanias' description provides that the Heraion was used as a storage area in his time, its position in his account is significant. Naturally, he begins his tour with the temple of Zeus: entering from near the south-east, the temple of Zeus would be seen first, and specifically the east end, the 'business end', and the one whose sculptures he does in fact describe first. Inside the temple he describes at length Pheidias' chryselephantine cult statue of Zeus. He continues to the Pelopeion,⁴⁵ and the ash altar of Zeus, where the first mention of the Heraion occurs, merely as a topographical indicator (v. 13. 8). Next he describes 'all the altars in Olympia' (v. 14. 4). He then leaves the Altis (v. 15. 1), to describe the workshop of Pheidias⁴⁶ and the Leonidaion. Pausanias' apparent leaving the sacred for the secular may seem strange, especially as he subsequently returns to the Altis. However, the visit to the workshop of Pheidias arises naturally from the discussion of the cult statue and, furthermore, it boasted an altar to all the gods; the more striking inclusion is that of the Leonidaion, founded in the later fourth century, but rebuilt c. AD 150, and by Pausanias' time essentially a Roman building.⁴⁷ Although it is mentioned only for its function, not its form, it may still be counted a rare exception to his general policy of not mentioning even approximately contemporary buildings, of which the omission of the Nymphaion of Herodes Atticus is the most famous example at Olympia, although one could add the buildings of Nero.⁴⁸ For present purposes, it is significant that he should mention a secular building before turning to the second major temple within the Altis.

Pausanias then returns to the Altis, and (via several altars) to the Heraion. Thus he places it at the end of his description of the Altis bar the single pillar of the house of Oinomaos, the Metroön and the Philippeion. Of these, the house of Oinomaos was by now a monument rather than part of a building (if in fact it ever had been⁴⁹); the Metroön had been rededicated and thus no longer served its original function; and the Philippeion had lost two of the statues for which it was built, and was therefore at most partly fulfilling its original purpose. Pausanias' sequence of description at Olympia (if we exclude the Leonidaion which (as noted, p. 469 above) is mentioned only for its function, not its form) is: sacred (Zeus temple, Pelopeion, ash altar, other altars); the workshop of Pheidias (sacred to the extent that it contained an altar, and perhaps by association since the cult-statue was made there); the Heraion; buildings of changed or uncertain status (Metroön, Philippeion); secular (dedicatory offerings, athlete statues, treasuries, stadium, hippodrome), with the sole exception of the Hippodameion (a sanctuary linked to the Heraion by the presence in the latter of Hippodameia's toy couch). This sequence places the description of the Heraion between that of the sacred on the one hand, and on the other the secular and the 'desanctified' (to borrow Stone's phrase (p. 468 above)). The position of the Heraion in the sequence is not determined by its topographical position, since after it Pausanias first goes east to the Metroön, then west to the Philippeion, then east to the treasuries, and beyond to the stadium and hippodrome. As Pausanias describes the altars in the order in which the local people sacrifice at them (v. 14. 4), is it not possible that the order in which he visits the buildings also reflects some form of local grading? The position of the Heraion in his narrative appears, therefore, to have been determined by a perception on Pausanias' part that it was the first of the non-sacred buildings he describes.

That the Heraion was in some sense a storeroom seems to me inescapable. Nothing in Pausanias' description of the Heraion indicates whether objects were simply stored there or were ostentatiously displayed, but the very fact that he saw all the objects in such detail suggests that they were highly accessible to visitors. This is implied particularly by the case of the chest of Kypselos, which may not have been substantially bigger than baby-sized, and yet is described in the sort of detail otherwise reserved by Pausanias for considerably larger objects such as Pheidias' Zeus or the paintings of the Knidian Lesche at Delphi. If not formal display, this does not sound like the contents of a typical 'box-room'. The valuable material of most of the objects suggests that they had an intrinsic worth made all the greater by their antiquity, and the need for careful preservation of such antiquities may well have been an important factor in the manner of their storage. It may, therefore, be that the Heraion was by Pausanias' day acting effectively as a sanctuary treasury. All temples were to some extent treasuries, and a look through Pausanias' text shows how often antiquities and valuables were displayed within a temple: the example of the Erechtheion was cited in n. 40, and one could add many more, such as the display of the hide of the Kalydonian boar and the fetters used to bind Spartan prisoners in the the battle of the fetters c.580-560 in the fourth-century temple of Athena Alea at Tegea (viii. 47. 2; Herodotos (i.66) also saw the fetters in the earlier temple at Tegea). However, both the Erechtheion and the Tegea temple are used to display votives and objects of great historical importance to the local community, thus fulfilling one of the basic functions of a temple, the supreme architectural form of investment by the community rather than by individuals. There are two significant differences between this sort of display and that evident in the Heraion: first, the objects have been moved in from other buildings and are not (with the exception of the Zeus/Hera statues) naturally associated with the Heraion; secondly, the function of acting as a container, which is a basic one to all Greek temples, has in this case taken over completely, quite possibly to the exclusion of the religious function. The uniquely crowded Heraion should perhaps have a unique function; and any sanctuary which attracted ri h offerings over many centuries would have had a storage problem. The objects in the Heraion have, in all cases where we can tell, been moved, and from only very short distances, and several are from treasuries, structures which were themselves designed for display.⁵⁰

As a further indication that the function of treasury is central here, the treasuries Pausanias saw comprised the active (e.g. the Sikyonian, vi. 19. 2-6, and Carthaginian, vi. 19. 7), the empty (Geloan, vi. 19. 10), the changed in function (the Kyrenian, containing, like the Metroon, statues of Roman emperors⁵¹), and those of cities no longer extant (Metapontum, vi. 19. 11).⁵² They originally contained valuable objects, but in some cases no longer did so. If, for whatever reason, it was felt necessary to store these objects elsewhere, the Heraion seems the likeliest candidate.

Here the motivation behind this collection of statuary may be considered. At the end of last century both Treu⁵³ and Wemicke⁵⁴ noted that the four Roman statues, of the second half of the first century AD, were female, and may have had some relevance to the Hera cult. The argument can be extended in view of the predominantly female identity of the other statues in the temple. However, striking though the statistics are, I am hesitant to see gender as a central motivation: for it to be so, one has to discount the male statues which, while few, are of significant figures, including Zeus, Apollo, Hermes, and Dionysos. Nor are they physically inconspicuous to judge from Pausanias' account. Such an argument requires a broader examination of the gender of dedications in Greek temples before a correlation, general or specific, can be proposed. Within the scope of the present article, this idea can only be mooted, and the author's hope expressed that it will be fully examined elsewhere.

In respect of the Roman statues, the neighbouririg Metroon provides an excellent reason not to infer too much from the gender of the statues dedicated: this was still known as the Metroön in Pausanias' day despite its re-dedication to the imperial cult (below), and thus retained its 'female' identity, but the Roman statues number four male and three female. Similarly, I would suggest that the 'Heraion' was such only in name, and that the female connection is, literally, nominal.

The contemporaneity of the Roman statues suggests that they were intended as a group, and it might be argued that the early chryselephantine statues, a group by material and, broadly, date (as far as we can judge), were also deliberately collected. Periodical clearing was necessarily standard sanctuary procedure; chryselephantine statues are likely to have been thought too good to destroy, but too old-fashioned to be favoured; archaic statues as a genre did not appeal to the Romans, a rule to which Pausanias is an exception. While the number of early chryselephantines is striking, I would be reluctant to see their collection as the motivation behind the crowding of the Heraion because there are objects of other materials and dates (as well as later chryselephantines). Another factor is that we know little of the likely quantity of chryselephantine one might expect at a sanctuary, and should be very cautious, therefore, in assuming that it was as rare, valuable (and collectable) as it now appears—indeed, the quantity in the Heraion is good evidence that while sculptures of this material were comparatively rare, they were by no means unfamiliar, and we cannot therefore assume that their material would be a reason for their collection.

That the Heraion was being fully exploited as storage space by the second century AD seems clear. But does this mean that it was 'desanctified'? If the criterion is that which leads Stone to apply the term to the Metroon, namely the removal of the cult-statue,⁵⁵ the answer lies in the interpretation the reader places on Pausanias' mention of statues of Zeus and Hera, and on the view taken of the limestone head discussed above. But if 'desanctified' means having no religious function, there is one factor which militates against this interpretation: Pausanias mentions a robe woven for Hera every four years for the Heraia, the women's games which, in effect, compensated the majority of women who were excluded from attending the games proper (virgins being the exception).⁵⁶ Indeed, the consequence of attending illegally could be the dire one of being thrown from Mount Typaion (v. 6. 7). These quadrennial games in honour of Hera imply some cult practice centring on a statue of Hera, presumably inside the temple since the discussion of the robe and the games comes between that of the Heraion and its contents. However, this is only implication, and there is no suggestion of any cult activity beyond this one practice every four years; the presence of representations of so many other gods suggests that at the least this is not Hera's exclusive domain. If visible religious activity within is a sine qua non for a working temple, the Heraion's position is at best an uncertain one.

If the Heraion was by Pausanias' day a museum, or a storeroom, it cannot have been built as such: apart from the improbability of the temple form being used for a simple display area, it was built at the period of the earliest treasuries (e.g. at Delphi the Corinthian, dating from the time of Kypselos according to Herodotos (i. 14. 2),⁵⁷ and the Sikyonian c.570-560), so that if a display area had been wanted, it would probably have taken the physical form associated from the beginning with the treasury as a genre. It must, rather, have been a temple dedicated in the conventional manner to a single deity (or, in this case, perhaps to a pair of deities), and receiving dedications to that deity, rather than miscellaneous items moved from their original locations as so many of the objects Pausanias saw in the Heraion were. Whether the Heraion was originally dedicated to Zeus, Hera or both is not of importance in this debate: Pausanias (like Dio Chrysostom before him) calls it the temple of Hera, and there is no reason to suppose that it was not by then universally known as such. However, this nomenclature need not mean that it was still a temple of Hera in Pausanias' day: the neighbouring Metroön, the most pertinent parallel, 'still preserves its ancient name' (v. 20. 9), despite having been deprived of its cult-statue and re-dedicated to the imperial cult. In short, there was nothing left but popular tradition to justify the name 'Metroön'; and the same may well have applied to the Heraion. But in saying that, we should not underestimate the significance of popular tradition.

The Heraion may have been converted to a 'museum' with the building of the Zeus temple; or maybe in the hellenistic period when the difference between temple and museum narrows (cf. e.g. Herodas, Mime 4). Wernicke suggested that it was converted to a 'museum' as a compliment to the visiting Nero (n. 3). It is true that Nero had a considerable impact on Olympia, in terms of the buildings which are being increasingly understood (n. 48), and the procedure of the games, introducing dramatic and musical events (Suet. Nero, 23. 1). But these were held, significantly, only at the Olympiad at which they were introduced and in which Nero himself participated, also winning the chariot race by cheating (Suet. Nero, 24. 2; cf. Dio Cassius, lxii. 14. 1). Pausanias himself informs us that this 'is the only Olympiad which is omitted in the Elean register' (10. 36. 9). The immediate cessation of Nero's innovations at the games was matched by the swift replacement of his building north of the Prytaneion with a larger structure;⁵⁸ perhaps also by the changing of the processional entrance from the south-east area to the south-west of the Altis, where it was in Pausanias' day (v. 15. 2).⁵⁹ Whether these re-arrangements were deliberately aimed at the acts of Nero, or were simply coincidental, is not recorded, but the former seems eminently probable. Although Pausanias gives us no information, we should not expect approval of Nero from him: he has more to say on Nero than on any emperor bar Augustus and Hadrian, and Nero's character forms a constant theme (representative are ix. 27. 3, x. 7. 1). Although the Roman statues in the Heraion, like those in the Metroön,⁶⁰ date from the first century AD, that does not imply that all statues were accumulated at the same period; indeed, the very range of dates suggests a longer process. It is, therefore, extremely unlikely that the creation of the 'museum' within the Heraion can be attributed to the Neronian period, much less be a compliment to him.

We cannot be sure what the temple Pausanias calls that of Hera was called when it was built, or after the Zeus temple was built, or in the early Imperial period. We know what it was called in Pausanias' time, but what it actually was must, I suggest, be assessed against the background of the 'desanctifying', 'resanctifying', or abandonment of temples in the Roman period. Thus a change in function can, I believe, be seen in the Heraion, as it can in the Metroon, as an example of a wider pattern evident in Roman Imperial Greece.

Notes

¹ I thank Dr A. W. Johnston and Dr A. J. S. Spawforth for their advice and interest, and the anonymous reader for helpful suggestions. I am also very grateful to Prof. Ulrich Sinn for periegeseis of Olympia in 1989 and 1994, the latter in particular accompanied by much stimulating discussion.

References to Pausanias are from the Teubner edition of M. H. Rocha-Pereira, vols. i-iii (2nd end; Leipzig, 1989-90). Special abbreviations:

Arafat = K. W. Arafat, 'Pausanias' attitude to antiquities', BSA 87 (1992), 387-409

Frazer = J. G. Frazer, Pausanias's Description of Greece, i-vi (London, 1898); translations are from vol. i, with modifications

Papachatzis = N. D. Papachatzis … (Athens, 1979)

² 'Hera, the sphinx', Hesp. 13 (1944), 353-60.

³ 'Olympische Beiträge II. Zur Geschichte des Heraion', JdI 9 (1894), 101-14.

⁴ Paus.'s accuracy, many times attested by excavation (as e.g. C. Habicht, Pausanias' Guide to Ancient Greece (California, 1985), 32, 149-51, observes), ironically slips when he dates the Heraion to 1096; this historical miscalculation cannot, however, invalidate the accuracy of the autopsy he records. On the Heraion and its date, A. Mallwitz, 'Das Heraion von Olympia und seine Vorgänger', Jdl 81 (1966), 310-76.

⁵ Frazer i. xv; Habicht (n. 4), 9; Papachatzis 196-7.

⁶ G. Treu, Olympia Ergebnisse, iii: Die Bildwerke von Olympia in Stein und Thon (Berlin, 1897), 251-5, 258-60; Frazer iii. 589 mentions six marble pedestals for Roman statues in the pronaos.

⁷ Arafat 392-3, on technique as a criterion of antiquity for Paus., and the word   . Papachatzis (283) says the helmet suggests Zeus Areios; against, C. Kardara, 'Olympia: Peirithoos, Apollo or Zeus Areios', A. Delt. 25 A (1970), 12-19.

⁸ At vii. 4. 4 Paus. calls Smilis 'a contemporary of Daidalos'. Frazer discusses his date (iv. 122-3). For present purposes, what matters is that Paus. sees him as early archaic, by association with the first Greek artist.

⁹LIMC v. 1. 502-10 (V. Machaira), 510-38 (L. Abad Casal).

¹⁰ Arafat 404-5.

¹¹ On the treasuries at Olympia, including the Epidamnian, K. Herrmann, 'Die Schatzhäuser in Olympia', in W. Coulson and H. Kyrieleis (eds), Proceedings of an International Symposium on the Olympic Games (Athens, 1992), 25-32. On the material of the Herakles and Hesperides group, Hill (n. 2), 354. Paus.'s interest in technique is frequently manifest (Arafat 392-3). J. J. Pollitt says of the Epidamnian treasury Hesperides that 'presumably they were reliefs' (The Art of Greece (rev. edn, Cambridge, 1990), 23 n. 6), but there is nothing in the language used, nor in the context, to indicate that technique, and much to contradict it.…

¹³Pace H. Hitzig and H. Blümner, who say that the material of the Horai, Demeter, Kore, Artemis, and Apollo (listed at 5. 17. 2-3) remains unknown ('vielleicht' wooden: Des Pausanias Beschreibung von Griechenland (Leipzig, 1901), ii, pt. 1, p. 391), although these are included in Paus.'s list of chryselephantines.

¹⁴ e.g. P. Amandry, 'Rapport préliminaire sur les statues chryséléphantines de Delphes', BCH 63 (1939), 86-119. J. Boardman, Greek Sculpture: The Archaic Period (London, 1978), fig. 127.

¹⁵ n. 2, 354-6.

¹⁶ Several of the scholars cited in n. 17 below maintain that the head is that of Hera, and others doubt the attribution, but have not faced or countered this reading of Paus. This may result from not reading Paus. beyond the sentence containing the reference to the Hera statue (to take the earliest example, this is certainly the case in Treu (n. 6), 1-4).

¹⁷ Hill (n. 2), in support, and seeing it as part of a pedimental group. U. Sinn …, AM 99 (1984), 77-87. The identification of the head with that of the presumed cult statue of Hera is long-standing and still general: Treu (n. 6), 1-4; Frazer (i. 593-4, with illustration); B. Ashmole, Architect and Sculptor in Classical Greece (New York, 1972), 6, fig. 5; Papachatzis raises no doubts (282-3, fig. 285). R. E. Wycherley says the head 'belonged very probably to the image of Hera seen by Pausanias in the temple. The high crown is appropriate to Hera; and the primitive style agrees with Pausanias' remarks—the eyes are large and triangular, the mouth is a simple curve, and the whole face is flattish' (Loeb edn, vol. v. 132-3 and pl. 57 a).… Boardman calls it Hera (n. 14, caption to fig. 73), and says 'it appears to be from the seated Hera which, with a standing Zeus, was the cult group in the Temple of Hera … the rather stark features suit its size and function' (25). C. M. Robertson says it 'probably belonged to the cult statue' (A History of Greek Art (Cambridge, 1975), 35, cf. 47), but adds that 'none of these [aforementioned] facts is conclusive for the identification … but all are entirely compatible with it and taken together make it highly probable' (48). B. S. Ridgway is sceptical of its identification as the head of Hera on sculptural grounds, as well as archaeological (The Archaic Style in Greek Sculpture (Princeton, 1977), 123-4; also n. 21 below); and by A. Stewart, who says 'the likelihood that it comes from Hera's cult statue in the Heraion is small: the base is too narrow to hold a seated statue of this size. In addition, it apparently had no right ear, it is markedly asymmetrical, and was found with what may be wing fragments; perhaps it does come from a sphinx' (Greek Sculpture: An Exploration (Yale, 1990), 113).

¹⁸ Ridgway (n. 17), 123-4, 144, citing references including A. Mallwitz, Olympia und seine Bauten (Munich, 1972), 146-7.

¹⁹ v. 21. 3 gives a date of 388: Frazer iii. 624.

²⁰ The technical arguments against a 4th-cent. date mustered by S. Adam have been widely accepted (The Technique of Greek Sculpture in the Archaic and Classical Periods (London, 1966), 124-8; counter-arguments have been put by R. E. Wycherley, 'Pausanias and Praxiteles', Hesp. suppl. 20 (1982), 182-91. B. S. Ridgway places it not earlier than the 3rd cent., saying that it 'perhaps may be better placed within the second century or later' (Hellenistic Sculpture, i: The Styles of c.331-200 BC (Bristol, 1990), 14); in the same work, she is more definite, saying that it 'has been convincingly placed within the second century BC' (80). Stewart calls it hellenistic (n. 17 above, 177, 198, 279, 296, and see next note). K. D. Morrow dates it not before hellenistic, and probably 2nd cent. (Greek Footwear and the Dating of Sculpture (Wisconsin, 1985), 83-4). It is not mentioned in R. R. R. Smith, Hellenistic Sculpture (London, 1991). Pollitt (n. 15), 258-9, gives a detailed bibliography on the controversy.

The base has been dated to the 2nd or Ist cent. BC (Frazer iii. 595; Dinsmoor, AJA 35 (1931), 296; Adam 125), which would support that date for the statue also, unless one believes, with Frazer and Papachatzis (283-6) that the statue is of 4th-cent. date, and was originally set up elsewhere and transferred to the Heraion. Frazer cites Dorpfeld's idea that the Hermes 'originally stood either against the western wall or nearer the middle of the cella, and that when the cella was remodelled by the taking down of the cross-walls the pedestal was moved to its present position between the columns' (iii. 589-90; the inner columns are here referred to). He then notes that 'other parts of the temple beside the cella were enriched with statues and votive offerings of different sorts', and goes on to detail the evidence, including the Roman statues mentioned above. If the Hermes were indeed transferred to the Heraion, the picture I believe Paus.'s account paints of objects being 'imported' into it gains credence.

²¹ First suggested by C. Blümel, Der Hermes eines Praxiteles (Baden-Baden, 1948); most recently Stewart has said, 'Pausanias saw, and we have, a statue by an imitator, perhaps one of his sons or grandsons' (n. 17, 177; for the genealogy of Praxiteles' family, Stewart 277).

²² n. 17, 386.

²³ At v. 17. 4, the words … added by Buttmann, and retained by Rocha-Pereira (Teubner edn), Papachatzis (286-7), and Jones (Loeb edn), so that the names are not certain, but the reference to two female statues is.

²⁴ E. Kunze and H. Schleif, 01. Forsch. i (Berlin, 1944), 1-52; F. Seiler, Die griechische Tholos (Mainz, 1986), 89-103; S. Miller, 'The Philippeion and Macedonian hellenistic architecture', AM 88 (1973), 189-218.

²⁵ e.g. Frazer iii. 600, 606, following H. Stuart-Jones, 'The chest of Kypselos', JHS 14 (1894), 30-80; J. B. Carter, 'The chests of Periander', AJA 93 (1989), 355-78, esp. 361-3, 374-5; J. M. Hurwit, The Art and Culture of Early Greece, 1100-480 BC (Cornell, 1985), 227-30.

²⁶ Reconstructions: Stuart-Jones (n. 25), pl. 1, opp. p. 80, reproduced by Frazer (iii. 606), and by Papachatzis (298-9 fig. 301); Papachatzis (296-7, fig. 300) also reproduces that of W. von Massow-Biese, 'Die Kypseloslade', AM 41 (1916), 117, fig. 25.

²⁷ On the problematic date of Kolotes, Pollitt (n. 11), 220.

²⁸ The translation of Frazer given above carefully conveys the uncertainty, as do those of Papachatzis (300) and W. H. S. Jones (Loeb edn, ii. 497). Pollitt translates the passage unambiguously as appertaining to the table (n. 11), 220. K. Hitzl also sees the figures as reliefs on the table (01. Forsch. xix: Die kaiserzeitliche Statuenausstattung des Metroon (Berlin, 1991), 14 n. 149).

²⁹ Treu (n. 6), 243-8, 255-8, diagram p. 225; Hitzl (n. 28); S. C. Stone, 'The imperial sculptural group in the Metroon at Olympia', AM 100 (1985), 377-91. That the cult statue had been transferred from the Metroon to the Heraion was first suggested by Mallwitz (n. 18), 160. Against, most recently, Hitzl, 14 n. 149; Stone, citing Mallwitz and saying, 'I cannot find it in Pausanias' account (386), to my mind an over-positive rejection since, whichever interpretation one supports, the passage seems to me to be undeniably open to both.

³⁰ Hitzl (n. 28), 14.

³¹ n. 29, 387; followed by Hitzl (n. 28), 14 n. 149.

³² iii. 591.

³³Olympia: The Sculptures of the Temple of Zeus (London, 1967), 5.

³⁴The Architecture of Ancient Greece (3rd edn, London and Sydney, 1950), 53.

³⁵ n. 33, 151.

³⁶ iii. 589.

³⁷ 279.

³⁸ n. 25, 182. Not everyone thinks it originally a joint temple: Robertson sees it as a Hera temple alone (n. 17), 271.

³⁹ It may be that the identification of the temple as a joint one of Zeus and Hera owes much to what has been, effectively, an article of faith that Zeus is the senior god and that, therefore, his cult must go back as early as the shrine. This wishful thinking is best shown by the identification of terracotta statuettes of the Protogeometric and Geometric period as figures of Zeus, which may be correct but is not based on attributes or other genuine methods of assessment of their identity: C. A. Morgan, Athletes and Oracles (Cambridge, 1990), 65-92; W. D. Heilmeyer, 01. Forsch. vii: Fruhe olympische Tonfiguren (Berlin, 1972).

⁴⁰ Hill (n. 2), 354, 355; Wernicke (n. 3), 110, 112, 114, calls it a 'Kunstmuseum'. The presence in temples of objects classed by Paus. as 'antiquities' can readily be paralleled: e.g. the Erechtheion contained (among other, unspecified, objects) a folding stool made by Daidalos, the breastplate of Masistios, and the sword of Mardonios (i. 27. 1; Arafat 404). However, the Erechtheion, like other temples similarly described but unlike the Heraion, is clearly a working temple; also, as noted, in no case are there anywhere near as many antiquities as in the Heraion.

⁴¹ Treu (n. 6), 232-5, 243-8, 255-8; Hitzl (n. 28); Stone (n. 29); S. Price, Rituals and Power: The Roman Imperial Cult in Asia Minor (Cambridge, 1984), 160-1; Frazer iii. 622; Papachatzis 303.

⁴² Stone (n. 29), 386-7.

⁴³ This is also shown by an inscribed architrave block. Stone (n. 29) and Hitzl (n. 28) reach different conclusions about the date and nature of the Imperial cult in the Metroon, but both place it, and the statues, well before Paus.'s day, which is the significant point for present purposes.

⁴⁴ F. Eckstein says that the kline and discus 'waren wohl in der Zella verwahrt, wahrend der Tisch bei der Zeremonie, für die er eigens geschaffen war, wohl am Eingang, im Pronaos seine Aufstellung fand, die übrige Zeit ebenfalls im Tempelinnern verwahrt war' (Pausanias Reisen in Griechenland (1986-7), ii. 247).

⁴⁵ H. Kyrieleis, 'Neue Ausgrabungen in Olympia', in Coulson and Kyrieleis (n. 11), 22-4; AR 1990-1, 31; 1989-90, 30. H. Kyrieleis, 'Neue Ausgrabungen in Olympia', Antike Welt, 21 (1990), 181-8; A. Mallwitz, 'Cult and competition locations at Olympia', in W.J. Raschke (ed.), The Archaeology of the Olympics (Wisconsin, 1988), 79-89.

⁴⁶ Most recently, W. Schiering, 01. Forsch. xviii: Die Werkstatt des Pheidias in Olympia, 2: Werkstattfunde (Berlin, 1991).

⁴⁷AR 1992-3, 28-9; Mallwitz, in Raschke (n. 45), 82, caption to fig. 6: 2; V. Heermann, 'Banketträume im Leonidaion', AM 99 (1984), 243-50; Frazer iii. 568-9.

⁴⁸ The project to investigate Roman imperial Olympia, under the direction of Ulrich Sinn, continues to produce revisions of our view of the period at Olympia: e.g. AR 1993-4, 26; 1992-3, 28-9; 1991-2, 25-6; 1990-1, 31; U. Sinn … in A. D. Rizakis (ed.), … Achaia und Elis in der Antike (Athens, 1991), 365-71.

⁴⁹ It has recently been suggested that it was in fact the turning-post of the early stadium: E. L. Brulotte, AJA 98 (1994), 53-64.

⁵⁰ There are parallels in the Roman period for moving religious monuments into central places (as museums?), most famously the 'itinerant temples' (most recently, S. E. Alcock, Graecia capta (Cambridge, 1993), 191-6), but this is a different order of activity.

⁵¹ Hitzl (n. 28), 119-22.

⁵² At Delphi, he says he saw no treasure in any of the treasuries (x. 11. 1); the picture is less consistent at Olympia.

⁵³ Treu (n. 6), 252-3.

⁵⁴ Wernicke (n. 3), 110.

⁵⁵ Stone (n. 29), 387. Here might be added temples which seem no longer to fulfil the role of a temple; most obviously, those with no roof, on which Paus. remarks on several occasions, e.g. the temple of the Mother of the Gods at the source of the Alpheios (vii. 44. 3). Alcock (n. 50), 200, 207-8, 257 n. 43 ('to Pausanias an empty temple appears to form as significant a landmark as any functioning sanctuary'); ead., 'Abandoned temples, abandoned towns: Pausanias and the sacred landscape of Roman Greece' (AJA 96 (1992), 349). There are also temples which have been re-consecrated, such as the Metroön at Olympia (p. 466 above), or the roofless temple in the Agora of Elis, which had been re-consecrated to the Roman emperors (vi. 25. 1). However, re-dedication to the Roman emperors is a very different matter from re-dedication to another deity.

⁵⁶ On the Heraia, T. F. Scanlon, 'The footrace of the Heraia at Olympia', Ancient World, 9 (1984), 77-90; further refs. in Raschke (n. 45), 215 n. 69.

⁵⁷ Also LSAG² 102-4, no. 21.

⁵⁸ Sinn, in Rizakis (n. 48), 370.

⁵⁹ Frazer iii. 492, citing Dörpfeld, refers to a possible triumphal arch erected by Nero; in fact, as Sinn points out to me, there is no archaeological evidence whatever to justify identification as an arch, never mind its postulated demolition after the death of Nero (pace, most recently, Alcock (n. 50), 190).

⁶⁰ Stone (n. 29), 378, not citing Wernicke's theory, despite arguing for the removal by Nero of the cult statue from the Metroon.