1. History of the structure

The church of the Holy Apostles was situated on the 4th hill inside the Constantinian wall of the city, in the area of Constantiana, near to the “Aqueduct of Valens”, at the site now occupied by Fatih Camii, which was erected shortly after the Ottoman conquest of 1453. Holy Apostles was the second most important, and probably second largest church after Hagia Sophia, in Byzantine Constantinople. Erected immediately west of, and adjoining to, the rotunda mausoleum of Constantine the Great (336), it was sometimes mistakenly attributed to Constantine I. The church was probably consecrated under Constantius II, who saw that the relics of St. Timotheus and of the Apostles Andrew and Luke were translated there. However, according to C. Mango, the relics were put to the Mausoleum of Constantine, which was thus becoming a martyrion of the Apostles; as for the church, it may have been consecrated even later, in 370 according to the Chronicon Paschale.¹ Another, cruciform, mausoleum was added probably in the 6^th century and is attributed to Justinian I. The mausoleums and the church were used for imperial burials until 1028, Constantine VIII being the last emperor buried there.²

The church played such a significant role in imperial ceremonies, and Porphyrogennetos's De cerimoniis names the Holy Apostles as a fixed point along the processional route that began at the Golden Gate.³ Evidence of 9^th-century repairs to the church, come from the Vita Basilii, with descriptions of Basil I strengthening the church with buttresses and reconstruction damaged sections.⁴ Probably around the same time a cycle of mosaics was added for the decoration of the church, which is partly described by Constantine of Rhodes ca. 940. At some point between the 10^th and the 12^th century the church was refurbished again and its mosaics partly redone; they were described once again by Nicolas Mesarites.⁵ The church may have suffered damages during the capture of the city by the Crusaders in 1204. The latest account giving details of the building is in the 15^th-century Choždenie inoka Zosimy (The Journey of the Deacon Zosima).⁶

After the Ottoman conquest in 1453, the church was given to Gennadios II Scholarios as the seat of the Patriarchate, but it was quickly found to be unsuitable, and the seat of the Patriarchate was moved to the Church of the Virgin Pammakaristos.⁷ The Ottoman mosque of Fatih Camii, along with its imaret, (“the mosque of the conqueror”) was built on the site by Mehmed II in 1463 and was extensively rebuilt by Mustafa III after having suffered damages from an earthquake of 1766.

2. Evidence provided by the sources

Until recently, it was thought that no physical evidence of the Church of the Holy Apostles remained. However, several reconstructions of the church and mausolea have been attempted on the grounds of descriptions in the written sources, a few – largely stylised – depictions in manuscripts, and comparison with other structures that copied the Holy Apostles, such as the church of St John in Ephesus and of St Mark in Venice.

Eusebius’s Vita Constantini provides a detailed description of the Mausoleum of Constantine. The building was a freestanding rotunda, resembling the extant 4^th-century imperial mausolea in Rome, within a rectilinear porticoed enclosure, surrounded by subsidiary buildings. It had a gilded dome and bronze grilles at openings around marble-reveted walls. Within it were stone sarcophagi containing imperial burials. Much less evidence is available for the 4^th-century church, although it was a large cruciform structure like some other major 4^th-century churches in both the East and in Italy. ⁸

The Justinianic church, dedicated on June 28, 550 and demolished by order of Mehmed II in the mid-15^th century, is described by Prokopios.⁹ This second church followed the cruciform plan of its predecessor, with an elevation of five domes, one over each “arm” and the fifth over the crossing, where the altar was placed. There was an atrium that prolonged the west arm, which was thus projecting longer than the other three. An internal colonnade was running parallel to the church’s cruciform plan and all arms had galleries.¹⁰ The central dome, on a drum pierced by windows, supported on four piers, stood higher than the other four domes of the church.¹¹ The church was rebuilt under Basil I, but its plan probably remained the same, so it is presumed that the building depicted in the “Menologion of Basil I” is the same as the Justinianic church.¹²

The internal decoration of the building has been described by Constantine of Rhodes between 931 and 944. The mosaics described are believed to belong to the period of the refurbishing under Basil I. A later account by Nicolas Mesarites, written between 1198 and 1203 while he was skeuophylax of the churches of the Great Palace, presents various differences that seem to hint a further, post 10^th c. redecoration.¹³ Mesarites also describes the marble revetment of the walls and the opus sectile pavements. The altar, under the central dome at the crossing, was beneath a pyramidal canopy and to its east there was a synthronon, which had probably survived from the Justinianic period of the structure.

3. Archaeological evidence and attempted reconstruction

Actual remains of the church have been long considered lost, except for some fragmentary pieces of sculpture found in the mosque courtyard in 1953, re-used Byzantine columns in the mosque and nearby Ottoman buildings, an underground colonnaded cistern in the north of the Ottoman precinct, and the imperial sarcophagi (including those currently displayed outside Istanbul Archaeological Museum).¹⁴

However, in 2001 Ken Dark, co-director of the Istanbul Rescue Archaeological Survey along Feründün Özgümüş, examined Fatih Camii and the site of its surrounding külliye, and found evidence of a very large structure that dates earlier than the surviving portions of the original 15^th-century mosque. This structure, apparently cruciform in plan, can be identified by courses of light whitish-grey limestone ashlar blocks that are clearly more eroded than the stonework above them. The blocks visibly pre-date 15^th-century features and serve no structural purpose in the 15^th-century mosque, while they might well belong to the Byzantine church of Holy Apostles that previously occupied the site.¹⁵

Based on this new evidence, Dark attempts the reconstruction of a structure ca. 57 m wide and 38 m long, with transepts ca. 35 m long and projecting ca. 6.5 m. This means that the ground plan of the building would be cruciform, but not arranged in a Greek cross pattern as Krautheimer proposed in his earlier reconstruction, based only on scriptural evidence.¹⁶ Dark further proposes an alternative reconstruction of the church, with a western narthex and side porticoes, which would give a plan closer to those of St John in Ephesus and St Mark in Venice, the two churches that are known to have been modelled on the Church of the Holy Apostles.¹⁷ In any case, provided that the structural evidence found in 2001 are indeed the remains of the Justinianic Holy Apostles, they provide a new and more solid basis for reconstruction of the church, which was long considered lost beyond rediscovery.