In the Pentadaktylos Mountains at 700m a.s.l., in the middle ward of Saint Hilarion castle (Kyrenia district) overlooking the north coast, the Mesaoria plain and the strategic mountain pass linking them. The site (and the church) is often assumed to have been monastic before its conversion to military use (before the late 12th century), although there is no secure textual or other evidence for a monastery [see Papacostas (1999a), vol. 1, 46ff.].
Description: The now ruinous domed octagon was built in rubble masonry with brick in its superstructure. The dome stood over six engaged cylindrical piers (two on each of the the north, south and west walls) and two free-standing piers with engaged columns in alternating brick (in thick mortar beds giving the appearance of recessed brick masonry) and stone courses; over these supports and the round brick arches they carry there was a very irregular pendentive zone creating an elliptical dome base. The main apse is horse-shoe shaped, lit by a triple window (with brick arches), and flanked by inscribed lateral apses. There are blind recessed niches on the north and west façades (behind the east wall of the later narthex). Traces of two layers of fresco decoration survive on the south wall, the latest dating perhaps to the 12th century The masonry of the south façade is covered by plaster imitating ashlar and brick masonry [cf. Asinou]. The trapezoїdal narthex and the north annexes were added in the 12th century [Megaw (1963) 5, Megaw (1985a) 218; MKE 6, 88; Papageorgiou (1985a) 329; Wharton (1988) 73-74].
Dating: A late 11th / early 12th century date is suggested by the architectural type, the use of brick in the masonry, the fresco decoration and the history of the site [Papageorgiou (1985a) 329]. Soteriou had proposed a 13th-century date for all Cypriot domed octagons [Soteriou (1942) 429-31, although see also Soteriou (1931a) 485 for an 11th-century date].
Modern repairs: The castle precinct was enclosed in 1904 to prevent removal of stone blocks [Jeffery (1907a) 14] and the debris was removed from the interior of the church and the masonry was treated in 1913 [RCAM 1914, 1]. The bema was restored in the late 1950s, and the east piers (standing up to c.1 m), arches, squinches and vaults were rebuilt [ARDA 1958, 16, 1959, 16]. Work in the narthex revealed a tomb in the south part and fresco decoration on the west wall of the naos [ARDA 1963, 11].
Early literature: The church was first described by Mas Latrie in a report of 1846 by which time the dome had already collapsed [Mas Latrie (1997) 357], then by Rey in 1860, who dated the church to the late medieval period [Rey (1871) 240]; Enlart in 1896 reported a now lost (?) Annunciation fresco in what must be the small domed room north of the prothesis [Enlart/Hunt (1987) 433]. Jeffery comments on the use of bricks in the masonry, dating the church to the 10th / 11th century on historical grounds [Jeffery (1907a) 15-16, Jeffery (1918) 266, Jeffery (1935) 3].
Views: Soteriou (1935) pl. 23b [before the 1950s restoration]; ARDA 1959, fig. 26 and Müller-Wiener (1966) pl. 131 [apses after the restoration]; Megaw (1963) 11 and Megaw (1974) fig. 33 [bema after the restoration], Megaw (1985a) 221 [northwest squinch; also in Vocotopoulos (1981) fig. 7].
Plan / section: Jeffery (1915/16) 114 [misleading regular plan: the east piers with engaged columns are shown as round columns]; Soteriou (1935) 21 [the narthex is not shown]; Papageorgiou (1966a) 226 [there is no distinction between the various building phases], Papageorgiou (1982b) 440 [identical to previous, but the different phases are shown; the triple apse window has been omitted]; Megaw (1963) 6-7 [plan of the castle]; Department of Antiquities Archive C.19.663.