• r2-20-10
    Stele with Artemis, Cybele, and two worshippers, front. (©Archaeological Exploration of Sardis/President and Fellows of Harvard College)
  • r2-20-20
    Stele with Artemis, Cybele, and two worshippers, front. (©Archaeological Exploration of Sardis/President and Fellows of Harvard College)
  • r2-20-30
    Stele with Artemis, Cybele, and two worshippers, three quarter view. (©Archaeological Exploration of Sardis/President and Fellows of Harvard College)
  • r2-20-40
    Stele with Artemis, Cybele, and two worshippers, detail of Artemis. (©Archaeological Exploration of Sardis/President and Fellows of Harvard College)
  • r2-20-40
    Stele with Artemis, Cybele, and two worshippers, reconstruction drawing of naiskos. (©Archaeological Exploration of Sardis/President and Fellows of Harvard College)
  • r2-20-50
    Stele with Artemis, Cybele, and two worshippers, detail of Cybele. (©Archaeological Exploration of Sardis/President and Fellows of Harvard College)
  • r2-20-60
    Stele with Artemis, Cybele, and two worshippers, detail of two worshippers. (©Archaeological Exploration of Sardis/President and Fellows of Harvard College)

Stele with Artemis, Cybele, and Two Worshippers

Ca. 400 BC, Late Lydian (Persian)
Manisa, Archaeological and Ethnographic Museum, 3937
Museum Inventory No.
Sardis or Museum Inv. No.
Marble, Stone
Object Type
Sculpture Type
Naiskos, Relief, Human Figure, Votive Relief
Syn FC 68
Syn FC Spolia
B-Grid Coordinates
E110.68 - E111.39 / N5.91 - N6.53 *96.51
Reused face down in the stylobate of the peristyle of Syn Fc built “not before the late 3rd C., nor likely after the 5th” (A. Seager, by letter Jan. 30, 1976), 95.51, top.

On both sides remain parts of the top of the triangular pediment (W. 0.49). Part of the interior of the pediment is preserved on the l. (H. 0.24; D. 0.65). Its corner ends just above the inner edge of the l. pilaster capital. It must have had acroteria. Two pilasters which supported the pediment stand on simple rectangular plinths; one profiled Ionic capital is preserved.

At first glance, the wide form of vertical “naiskos” stele with Ionic pilasters and triangular pediment (reconstruction Fig. 79) seems closest to Attic sepulchral stelai of around 400 B.C. (e.g. Hegeso, cf. Diepolder, 17, 25ff., pls. 10f., 17, 19f., 22, with dates from ca. 420-390 B.C.; Ampharete, Friis-Johansen, 14ff., figs. 1, 4). Later Attic naiskos stelai are more slender (e.g. Diepolder, Attischen Grabreliefs, 37, pl. 31). However, the Sardian form is somewhat simpler, lacking the intermediate band of the entablature which appears on the Hegeso stele.

Two frontal goddesses occupy two thirds of the niche space. They are approached from the r. by two worshippers shown in profile. The taller goddess, Artemis, holds a hind, the shorter, Cybele, a lion. Her tympanum is hanging on the “wall” in the background. The worshippers raise their r. hands in adoration.

Artemis wears a low polos over a veil which falls on her shoulders (Fig. 81). Her hair is parted in the middle, as is indicated by incisions. She wears a chiton with wide, buttoned, half-length sleeve showing on her r. arm. It descends in thin, vertical folds over her feet with outer folds flaring. Her cloak, wrapped around the lower part of the body, ascends diagonally from the r. hip to under the l. arm. Two long vertical folds with “swallow tail” ends drop from her l. elbow to her ankle. There are traces of folds under her r. forearm. Her battered face is broad, her neck thick. Her feet show no indication of either shoes or toes. Her r. arm supports the deer, which she grasps by the forelegs with her foreshortened l. arm. The tail of the deer is outlined against her r. inner elbow.

Cybele’s polos had a rectangular attachment over the center of her forehead (Fig. 82), and her veil falls in triangular folds. Her chiton has pointed sleeves and falls over her feet in the same way as that of Artemis. Her cloak has a diagonal roll of fabric which crosses from her r. hip to and over her l. arm. A long vertical fold with zigzag ends hangs down on her l. side. A vertical crease bisects her cloak. She holds a small lion, her r. hand grasping his hindquarters and her l. the r. shoulder. The fingers of both hands are slightly spread. The lion extends his paws horizontally in front.

The two worshippers (Fig. 83) have rather awkwardly elongated bodies and upper legs. The man in the back seems to wear a long cloak draped in a roll around the midriff. Curving folds are shown going down to his knee over his leg. The woman seems to wear a chiton like the goddess, but it appears only at the bottom since she is completely draped in a cloak which apparently also conceals her l. arm. Long incised folds go from her r. shoulder to l. hip and down to the curving edges above her feet. Both worshippers’ r. arms are bent, fingers pointing upward. The man is beardless and has a full head of hair. The woman has a short, round roll, as Krito on the Krito-Timariste stele (Lippold, Griechische Plastik, 206, pl. 64:3, dated 420 B.C.; Dohrn, Attische Plastik, 35f., 38, 400-390 B.C., surely too late). In their long, unstable proportions and gestures with slightly awkward, stumpy fingers they might reflect a model similar in style to the Krito-Timariste group.

For a votive relief with worshippers, the relief from Sardis is unusual in having a vertical rather than a horizontal format and in giving over so much space to two large, austerely frontal divinities. The general disposition and the type of a large frontal goddess with small but frontal worshipper recur on a relief from Eleusis (Lehmann-Hartleben, Zwei Meisterwerke, 88f., fig. 2, dated 460 B.C., too early). The juxtaposition of frontal Demeter and Kore on a relief, dated by inscription to 421-420 B.C., shows a freer arrangement and rhythm (Binneboessel, Urkundenreliefs, no. 5; Hausmann, Weihreliefs, 41, fig. 20; Lippold, Griechische Plastik,191, 198, pl. 73:3). An incomplete stele from the sanctuary of Aphrodite at Daphne, ca. 415-410 B.C., is a step further toward the mannerism of the Nike balustrade, ca. 415-410 B.C. (Svoronos, Athener Nationalmuseum, no. 1597, pl. 129; A. Linfert, Xenokrateiareliefs, 156, n.11, Beil. 84:1). It includes a small worshipper. Nearer in overall composition, and especially in the size and gestures of the worshippers, is a relief in Brocklesby Park (Michaelis, Marbles in Great Britain, 228, no. 10; B. Ashmole in Amelung Antike Plastik, 13-15, pl. 2) with a large god and goddess in profile and father, mother, daughter as worshippers. It is dated by Ashmole around 415 B.C., by Dohrn around 400 (Attische Plastik, 35).

While a dating around 400 B.C. fits the Attic evidence, especially of architecture, it is possible to date the relief to ca. 430-420 on the basis of figure style. In this case it would become a new link in the argument that the architectural form of the naiskos stele was not created by the Hegeso Master (Dohrn) but was developed in Eastern Greece whence it was introduced to Athens around 420 B.C.

The emphatic stiffness and frontality may have been intended to convey an icon-like quality or represent a style earlier than the time of the relief. The gesture with which the animal is held is that of mid-6th C. archaic images, such as the perikalles agalma, possibly “Aphrodite” of Cheramyes of Samos (Buschor, Altsamische Standbilder, 83, fig. 344; Richter, Korai, figs. 188-189). The poloi and veils, too, are archaic-traditional. What a 4th C. B.C. interpretation of such an image with veil and polos looks like is shown by the relief from Halikarnassos at the American Academy in Rome (A.W. Van Buren in Amelung Antike Plastik, 50-53, fig. 1). However, as G.M.A. Richter first observed (by letter), the goddesses wear clothes of 5th C. One statuary type which seems particularly close is seen in the Kore (?) relief from Corinth. The original of this type should fall between 450 and 430 B.C. It is assigned by Lippold to the Corinthian school, by others to Argive or early Phidias. The Kore’s cloak is enriched by large classic radiating folds, and the enlivening element is is even more evident in the Phidian “Kore Albani” (Lippold, Griechische Plastik, 154, pl. 56:1). If we translate Artemis and Cybele into statuary, we should envisage their images as more austere than the Kore from Corinth, on the borderline from Severe to High Classical style (450 B.C.?). In costume (especially sleeve) and proportion, one may compare an Attic votive relief dated by Friis-Johansen (Attic Grave-Reliefs, 139, fig. 70) to ca. 470-450 B.C. For the image of Artemis of Ephesus, the type holding the hind in her arm was hitherto not known. Fleischer (Artemis von Ephesos, 112-114) lists only images flanked by two standing stags.

Because Herodotus mentioned “the native goddess Cybele” as having a shrine at Sardis, which was burned by the Greeks in 499 B.C., and because the huge temple and most inscriptions in Lydian and Greek mentioned Artemis and not Cybele, many scholars believed that the two goddesses were amalgamated at Sardis. Our relief proves that they were different and that Artemis (presumably Artemis of Ephesus) was the bigger sister, but it also seems to imply that they could be worshipped jointly. This raises the question: was there a temple and/or shrine in which Artemis and Cybele were worshipped jointly, perhaps while the burnt temple or Cybele was being rebuilt? The preserved inscribed parastades from the Metroon may well belong to the late 5th or to the 4th C. B.C. Were there, after the burning of Sardis in 499 B.C., two images made in Severe style (470-450 B.C.), one of Cybele and the other of Artemis? The traditional type for Cybele at Sardis was that of a standing goddess (cf. Cat. 7 Figs. 20, 27) and only after Agorakritos’ immensely influential image became known, did the representation at Sardis change (Cat. 21 Figs. 84-85). Finally, why is Cybele smaller? Is it because she was a real epichorie theos, descended from old Near Eastern/Anatolian tradition, who was displaced in popular favor by Lydian-Greek Artemis?


Coarse-grained white marble. Most of the surface covered with calcium-carbonate accretion. There were large patches of mortar above the head of Artemis and between the heads of worshipper and lion.

Original triangular pediment top broken when stele was prepared for reuse in the Syn stylobate. Top of r. pilaster damaged. Part of the back broken off on r. (Figs. 79, 80). Lower r. corner broken and reattached in front. L. side of stele smoothly finished but much weathered. Roughened area above, from pilaster capital to the beginning of sloping surface of pediment. A second roughened area, from bottom of stele to 0.13 above base of l. jamb. Similar roughened area on r. side of stele at top. Heads of goddesses intentionally defaced; all surfaces much worn. Because the surface of the stele was already battered and damaged, no attempt was made to clean off all incrustation.

Stele: H. 0.99; max. Th. at bottom 0.29; W. 0.667. Niche: H. 0.642; W. 0.545. Pilaster plinths: H. 0.06; W. 0.10; D. 0.045. Pilaster shafts: H. 0.64: W. at bottom 0.065, at top 0.057. Capital: H. 0.03; W. at bottom 0.065, at top 0.07. H. of Artemis 0.63, of Cybele 0.61, of male worshipper 0.45, of female 0.38. Hind: L. 0.18; H. 0.105. Lion: L. 0.165; H. 0.09. Tympanum: H. 0.13; W. 0.105.
On the Corinthian Kore see Lippold, Griechische Plastik, 174, n.8, pl. 63:3, copy in Rome (?); V.H. Poulsen, Strenge Stil, 132; Homann-Wedeking, Meisterwerken, 55, 210; detailed discussion, Johnson, Corinth IX, no. 5.
See Also
See also: LATW Cat. 35
Published: BASOR 199, 48ff., fig. 39; Hanfmann-Waldbaum, “Kybele and Artemis”, 265f., 4 ill., with discussion of the development of Cybele.