¹ The words ‘reconstruction’ and ‘restoration’ arouse such alarm in Britain that I prefer to use here a more innocent term. Anastylosis implies the re-erection of those fallen elements (walls as well as columns) the positions and elevations of which are certain and the materials largely surviving.

² F. W. and H. W. Beechey, Proceedings of the Expedition to explore the northern coast of Africa . . . in 1821-22 (London, 1828).

³ R. Murdoch Smith and E. A. Porcher even stated (History of the recent discoveries at Cyrene (London, 1864) p. 58) that some crosses engraved on marble columns at Apollonia were ‘almost the only decided relics of Christianity we met with in the Cyrenaica’.

⁴ P. Romanelli, ‘La basilica cristiana nell’Africa Settentrionale Italiana’, Atti del IV Congresso Internazionale di Archeologia Cristiana (Rome, 1940), I, pp. 245-89.

⁵ For a provisional account, see Illustrated London News, 14 December, 1957.

⁶ By the 6th century the Dux was in effect the governor of the province, and not merely the military commander as in earlier times. Our palace is surely too elaborate to have belonged to the metropolitan Bishop, and it is even less likely to have been a monastery. The absence of any altar-fittings in front of the principal apse shows that it cannot have been a basilica discoperta.

⁶ Note on previous page.

⁷ E. Honigmann, Le Synekdemos d’Hierokles et Georges de Chypre (Brussels, 1939), pp. 47 and 61.

⁸ Procopius, De Aedificiis, VI, 2.

⁹ The most complete copy of the Decree, found at Ptolemais, is published as Suppl. Epig. Graec. IX, 356. The marble fragments of the Apollonia copy were found in 1954 by the French expedition on the seaside cliff immediately north of our palace, from the walls of which they may well have come. A third copy has been found at Teuchira (SEG. IX, 414).

¹⁰ Procopius, Anecdota, IX, 27 ; XII, 30.

¹¹ Beechey, op. cit., p. 515. The building is marked (e) on their plan of Apollonia.

¹² The fortresses and fortified country churches of the Byzantine period in Cyrenaica are generally well built with large blocks, and have resisted the weather. Cf. JRS, 43 (1953), pl. VI-VII.

¹³ I must here express my indebtedness to my colleagues in the Department of Antiquities who assisted in supervising this work, and notably to Essayed Abdulhamid Abdussaid who has made preliminary plans and drawings (figs. 1 and 2). The foreman Abdulkerim Hussein and mason Yusef Ramadan bore the main burden of the work for over a year.

¹⁴ E. Ghislanzoni (Notiziario Archeologico, I (1916), p. 105), observed these remains before the fort was built. Nothing is visible today.

¹⁵ This wall is the only one of the palace which corresponds more or less, to the axis of the earlier town-plan.

¹⁶ After the hall’s collapse the apse was left standing to buttress the south wall, which was already leaning perilously. We found remains of additional late buttresses in this room, possibly of the early Islamic period when the upper storey of the palace was still in use.

¹⁷ After the vault had fallen the ground-floor of this room was filled with soil, and a rough stone floor laid at the level of the top of the cupboards. This, too, is probably of early Islamic date. Although the invading Arabs were to make their own capital of Cyrenaica at Barca, they stayed at Apollonia long enough to overturn, in systematic fashion, the columns of all its churches.

¹⁸ At the moment of writing this room has not yet been completely cleared of soil.

¹⁹ Syrian churches frequently contain reliquaries provided with both a filling and emptying orifice, and J. Lassus (Sanctuaires Chrétiens de Syrie (Paris, 1947), pp. 163-7), shows this to be a means of providing pilgrims with ‘holy oil’ sanctified by contact with the relics. The presence in the Apollonia reliquary, of an upper orifice only, shows that it was not used in this manner.

²⁰ R. G. Goodchild, Cyrene and Apollonia—An Historical Guide (Cyrene, 1959), p. 68. ‘... Visible from the south door is an arched recess (arcosolium) which probably originally contained an important tomb, later moved elsewhere’. (This was written before the excavation of the Apollonia palace began.) Mention should here be made of the Byzantine palace of Qasr ibn Wardan in Syria which has a detached martyrium-chapel (not ‘built-in’ as at Apollonia). Cf. Lassus, op. cit., p. 146, fig. 62.

²¹ Triclinia, as we shall see, existed in the palaces of both Constantinople and Ravenna, and the term seems to have become accepted by the Byzantines in a political rather than social sense.

²² A blocked doorway communicating between the south cloister and room 15 is still visible, though partly masked by a very late ( ? early Arabic) flight of stone steps leading to the upper floor.

²³ In the early 5th century tribesmen from the interior wishing to pass into Roman territory had to take an oath of good conduct, sworn in either pagan or Christian fashion. It is interesting to observe that by the 6th century A.D. the Byzantine government was itself obliged to give guarantees.

²⁴ Procopius, Wars, IV, 21.

²⁵ E. Dyggve, Ravennatium Palatium Sacrum—la basilica ipetrale per cerimonie (Copenhagen, 1941).

²⁶ The only plan (fig. 4) of this building is that published by Ricci, L’Architettura Romanica in Italia (Stuttgart, 1925), p. XII. It is not at all clear which elements are certain and which conjectural.

²⁷ G. Ghirardini, ‘Gli Scavi del Palazzo di Teodorico a Ravenna’, Monumenti Antichi, XXIV (1916), pp. 738-838.

²⁸ D. Levi, Antioch Mosaic Pavements (Princeton, 1948), 1, p. 461, note 235.

²⁹ For ceremonial halls in Islamic palaces, of. Sergio Bettini, ‘Il Castello di Mschatta in Transgiordama nell’ambito dell’ “ Arte di Potenza “ tardoantica’, Anthemos (Scritti in onore di Carlo Anti), Firenze, 1955,

³⁰ K. M. Swoboda, Rómisene und Romanische Paläste (Vienna, 1924).

³¹ Dura-Europos, Prelim. Report of 9th season, Part III, The Palace of the Dux Ripae and the Dolicheneum (Yale, 1952), p. 72. In pp. 83-92 of this same report Rostovtzeff gives a valuable discussion of palaces and praetoria.

³² G. V. Gentili, La Villa Imperiale di Piazza Armerina (Itinerari dei musei a monumenti d’Italia), Rome, 1956. The hall (room 30 on Gentili’s plan) is called a ‘basilica’.

³³ A brief description and bibliography of the Gortyna praetorium is given in L. Pernier and L. Banti Guida degli Scavi Italiani in Creta (Roma, 1947), pp. 18-21 and pl. 27.

³⁴ G. Pesce, Il ‘Palazzo delle Colonne’ in Tolemaide di Cirenaica (Rome, 1950), pl. 5.

³⁵ Here, however, the lofty colonnade of the peristyle in front of the triclinium is not interrupted as at Ptolemais. The building has recently been studied, and will be published by Professor P. Mingazzini.