Amaseia (Byzantium), Aynalı Mağara, Wall Painting

1. The painting decoration of the church

The church situated within 2 km from Amaseia of Pontos, was given the Turkish name “Aynalı Mağara”, meaning “the cave that looks like a mirror”, possibly due to the smooth surface of its walls. Τhe monument was cut in the rock during the Hellenistic period in order to serve as a grave.1 Its arched facade is 9.73 m wide and 8.61 m high. Access is through a rectangular door 4.43 m from the floor, which is 1.43 m high and 1.32 m wide.

Internally, a small wall divides the cave into two rectangular compartments. The large one was used by the Christians as the nave of the church and the small as the Bema with a north-east direction. The saint to whom the church was dedicated remains unknown. The facade and the interior of the church include wall paintings belonging to two different layers, while nothing has been preserved from the third and latest layer, apart from some surfaces with mortar that were used as underlayer of the paintings.2

1.1. Layer A

This layer includes red crosses painted directly on the walls of the Bema and the nave. The ceiling of the sanctuary is adorned with two medallions, with a band showing continuous triangles running around their internal peripheries; their centre is occupied by a cross, the forked arms of which are decorated with circular motifs imitating precious stones. Another jewelled cross is represented on the southeastern wall of the same space; the traces of red colour preserved on the horizontal arm possibly belonged to a non-preserved inscription.

On the small wall separating the Bema from the nave there are outlines of plain crosses beneath the extensively damaged subsequent decorative layer, while three additional crosses in medals are represented at the eastern end of the northeastern wall of the nave.

Finally, two crosses, one with thin arms and forked edges and a second of an indeterminate shape can be seen on the southwestern wall of the same compartment, beneath a damaged representation belonging to Layer B.

1.2. Layer B

This layer includes the representation on the facade of the monument as well as most of the representations in the interior. The front representation is divided into two zones defined by vertical and horizontal red bands, which form a frame. The lower zone extends on either side of the church entrance, while the upper zone is above the entrance. The decoration of the lower zone has been preserved only in the left part of the facade and shows a procession of standing figures, six or eight in total, two of which, on the left side, are the best preserved. On the upper zone, at the centre, immediately above the entrance, Christ is represented sitting on the throne. To the left there are four full-length standing figures, one of which is female, while to the right there are barely discernible traces of two halos. The figure at the left end is better preserved, wearing a purple tunic with crosswise decorative bands and white dots in each rhombus created among them. This two-zone representation, in its present form, with Christ at the centre of the upper zone and the rest of the figures around him, possibly shows the Last Judgement.

In the interior of the church, on the northwestern wall of the sanctuary, there are two unidentified figures; the first is dressed in a red chiton and a yellow himation and the second in a red chiton.

On the barrel-vaulted ceiling of the nave there are representations of twelve standing figures divided into groups of six, each group surrounded by a red frame. Since until the early 20th century beside two of these figures there were inscriptions with the names of Andrew and Philip, the figures may be easily identified with the twelve apostles.3 The apostles are wearing tunics and himatia in various colours, which end on the front, above their left shoulder. Black bands also decorate some of the tunics.4 Τhree apostles of each group are holding folded scrolls and three closed codices (probably Gospels). Two of these codices preserve their decoration: the one codex has an equilateral cross and the other codex has two crossing crosses, in diagonal positions, on yellow background. The only thing visible on the apostles’ faces is a yellowish surface, the proplasmos. Furthermore, although the halos surrounding their heads must have initially been yellow with white borders, they are now in a dim green colour. Finally, the today background of the figures is black. However, a lower green or yellow zone is preserved, probably indicating the earth, while the upper zone was possibly painted in dark blue or light grey colour, indicating the sky.

On the northwestern wall of the same compartment, which ends in a semicircle, the decoration is divided into two zones: the lower is occupied by four full-length standing figures, while the upper by the representation of the Deesis. In the lower zone, from the left to the right, there is an officer-saint wearing a red dress and holding a round shield in his left hand. Behind him there is a female figure wearing long black clothes, a figure in dark blue clothes and, in the end, on a larger scale, a male figure in red shoes, wearing a black tunic with red hem. The tunic is adorned with crossing bands, white on the black surface and black on the red surface. At the centre of the upper zone Christ is represented frontally, holding an open Gospel in his left hand and wearing a dark blue tunic. His hair falls loose on his shoulders and the back, while the halo with the cross surrounds the head. To the left of Christ is Virgin Mary, turning in three-quarter profile towards him, with her hands lifted as if she is praying. Symmetrically opposite her and in the same position, John the Baptist is turning in three-quarter profile towards the centre. Only a part of his left hand is visible.

In the lower part of the northeastern wall, higher than the two above-mentioned crosses of Layer A, there is a black surface with red crosswise bands, while only the red outline of the base of a throne, which was part of an enthroned figure, has been preserved at the northern end of the wall.

Finally, a representation of two military saints in a red-bordered panel has been preserved in the western part of the southwestern wall. They seem to be in a rush, while striding with their red himatia fluttering behind them. Τheir bodies and faces are turning in three-quarter profiles towards the centre and downwards, where a snake is painted; the saints have stabbed their spears to the snake. A larger shield, decorated with a peripheral band including floral motifs, is fastened on the back of each saint.

1.3. Layer C

This layer includes five square plaster surfaces, whose decoration has not been preserved. The three of these surfaces are in the Bema, one on each wall, while the other two are in the nave, one on the representation with the officers-saints and the snake on the southwestern wall and the second on the enthroned figure of the northeastern wall.

2. Style and Chronology of the Decoration

From the three painting layers the church initially had, as shown above, only the two earlier have been preserved, which help the chronology of the decoration.5 The repetition of the decorative motifs of the first layer in other churches of Cappadocia and Trebizond dated to the period of Iconoclasm6 suggests that they were made during that period. However, their presence in 12th c. cappadocian churches7 in the valley of Göreme as well broadens the period in which the paintings of Layer A at Aynalı Mağara were made, thus incorporating them in the longer period between the 8th and the 12th century.8

In an attempt to more accurately date the paintings of the first, earlier layer, it is quite important to determine, as terminus ante quem, the date the paintings of the overlying second layer were created. This chronology mainly studies the representation of the apostles on the dome of the nave, which has been preserved in a satisfactory condition and may be studied in parallel. The representation was originally dated by Jerphanion9 to the third quarter of the 11th century judging from the stylistic similarity between the apostles and figures of the church of Karabaş Kilise, dated accurately to 1060/1.10 Αlthough this comparison was not fully accepted by Winfield, the latter thought that the representation of Aynalı Mağara should be dated to the 11th century, as the limited published comparative material is more convincing.11

However, the most detailed study on the style of the particular representation so far, conducted by Demus, led to slightly different conclusions.12 Demus saw in apostles’ himatia subtle but significant stylistic differences that made him identify two distinct types of figures: A and B. In the himatia of the type A apostle the folds are vertical and fall freely without outlining the body, while the himatia of the type B apostle form slanting and bending lines crossing the clothes, thus giving the impression of movement. That means that type A figures represent a more schematic style, while those of type B a more classical style. The close stylistic association between type A figures and the mosaic representations at the Church of Hagia Sophia in Kiev (1043-46)13 dates type A to the mid-11th century. Likewise, the stylistic associations between type B figures and the mosaics in Cefalu, Sicily (1148),14 date this type to the first half of the 12th century.15

According to these conclusions, the second half of the 11th century may be considered terminus ante quem with respect to Layer A of Aynalı Mağara. Although the moment when the Hellenistic grave was converted into a Christian church remains unknown, it must have been internally decorated with paintings for the first time at some moment before the first half of the 11th century. Some decades later new paintings were created in the interior and on the front part of the monument, while some of them were finally covered with plaster that served as the underlayer of the third and latest (completely damaged today) iconographic layer of the church.

3. The inscriptions of the facade

Until 1953 the researchers that had investigated the monument had read on the front side two inscriptions in capital letters:16 the letters ΤΗΣ / ΑΡΧΙ / ΙΕΡΕΥΣ on the first inscription, above the opening of the entrance to the church, and the letters ΚΛ . . . . . ΟΕΟΣ / Χ . . . . 'ΙΟΥ on the second inscription, below the entrance.17 These inscriptions were written in different periods and concerned different persons (although in the same grave18). Only the first inscription was preserved in 1971, according to the front design included in Winfield’s last publication concerning the monument.19




1. About the architecture of the church, see Winfield, D., “Aynalı Mağara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), pp. 282‑83.

2. About the fresco layers of the church, see Winfield, D., “Aynali Magara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), pp. 283-291.

3. Jerphanion, G. de, Mélanges d'archéologie anatolienne. Monuments préhelleniques, gréco-romains, byzantins et musulmans de Pont [=Mélanges de l’ Université Saint-Joseph 13] (Beyrouth 1928) p. 14.

4. However, it is very likely that the original colour of the bands was different, probably dark red, which became black in the course of the centuries and due to exposure to the sun; see Winfield, D., “Aynali Magara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), p. 288.

5. About the chronology of the first fresco layer, see Winfield, D., “Aynalı Mağara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), pp. 291-93; about the second layer, see Demus, O., “Zu den fresken von Aynalı Mağara”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), pp. 295-297.

6. Winfield, D., “Aynalı Mağara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), p. 292.

7. Winfield, D., “Aynalı Mağara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), p. 292.

8. Winfield, D., “Aynalı Mağara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), pp. 291-92.

9. Jerphanion, G. de, Mélanges d'archéologie anatolienne. Monuments préhelleniques, gréco-romains, byzantins et musulmans de Pont [=Mélanges de l’ Université Saint-Joseph 13] (Beyrouth 1928) p. 14.

10. Restle, M., “Zum Datum der Karabaş Kilise im Soğanli Dere”, Jahrbuch der Österreichischen Byzantinistik 19 (1970), pp. 261-66.

11. Winfield, D., “Aynalı Mağara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), p. 293.

12. Demus, O., “Zu den fresken von Aynalı Mağara”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), pp. 295-297.

13. Powstenko, O., The Cathedral of St. Sophia in Kiev (New Υork 1954), pl. 88, 95, 128, and Lazarev, V., Old Russian Murals and Mosaics (London 1966), fig. 25.

14. Demus, O., The Mosaics of Norman Sicily (London 1949), pl. 4a.

15. Demus, O., “Zu den fresken von Aynalı Mağara”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), pp. 296-297.

16. About the inscriptions and their transcriptions suggested at times, see mainly Jerphanion, G. de, Mélanges d'archéologie anatolienne. Monuments préhelleniques, gréco-romains, byzantins et musulmans de Pont [=Mélanges de l’ Université Saint-Joseph 13] (Beyrouth 1928) pp. 12-13, and Bean, G. E., “Inscriptions from Pontus”, Belleten 17 (1953), p. 169, no. 5.

17. The inscriptions are given here as they were transcribed by Bean, G. E., “Inscriptions from Pontus”, Belleten 17 (1953), p. 169, no. 5.

18. Bean, G. E., “Inscriptions from Pontus”, Belleten 17 (1953), p. 169, no. 5.

19. Winfield, D., “Aynalı Mağara. Amasya”, Jahrbuch der Österreichischen Byzantinistik 20 (1971), fig. 1 on p. 282.